0:05
[music] I see a large IP on the horizon.
Search this video
0:12
>> You're [music] surrounded by journalists. Hold your position.
0:24
[music] Overnight success.
0:31
The place of 30,000 [music] double right [music] there information.
1:03
Mark clearing order inbound.
1:07
[music] >> Let's just roll.
1:12
>> We are surrounded by journalists. Hold your position. Come on. Get up. Trust the experts here. Five. We are founder Mel 5.
1:46
I see multiple journalists on the horizon. Stand by. UAV online. [music] Double blaze. Triple blaze. [music] Double kill. Five kill.
2:33
Coke is w team deathmatch. We are experts. [music] Triple blades.
3:02
[music] That's just wrong. Right.
3:10
Quart [music] clearing order inbound.
3:24
We are surrounded by journalists. Hold your position. >> Strike one. >> Strike two.
3:42
Activate golden retriever mode.
3:55
[music] [music] >> Market clearing order inbound. [music] >> 5.
4:22
I see multiple journalists on the horizon. Stand by.
4:35
[music] You're watching TVPN. >> We're bigger today.
4:42
>> And it's Wednesday, January 28th, 2026.
4:46
We are live from the TVP Ultradome, the Temple of Technology, the fortress of finance, the capital of capital. >> Ramp. com, baby. Time is money. Say both.
4:54
These used corporate cards, bill pay, accounting, and a whole lot more all in one place. >> That's right.
4:58
Will >> I am wearing a sample today new >> of this new ramp quarter zip. Mhm.
5:05
>> And uh you'll notice here I included a little pocket here for your ramp card. It's just right here. Always close. Right over your heart.
5:15
>> Right over your heart.
5:16
>> Uh excited to get this one out in the world very soon.
5:20
>> Yeah, this will be fun.
5:20
Uh I wrote about Netflix.
5:24
There was a funny very brief interaction between Ben Thompson of Strategy and Netflix co-CEO Greg Peters on last Thursday's strategy interview and they go back re the only mention of AI in like an hourong interview or something.
5:41
It's just two little exchanges.
5:42
Ben says, "Is AI slop going to save you if it overwhelms the UGC platforms?"
5:48
And basically it's like you're a refuge.
5:50
So this is actual this is real.
5:53
And Greg Peters just says, "I think it's credible.
5:55
I don't know if that's the reality."
5:57
So I can't say with certainty that's where we're going to land, but it's a credible possibility.
6:02
And so he's like, "Maybe that's a bull case for me.
6:03
Maybe that's a bull case." It is interesting.
6:05
I mean, Netflix has been trading down uh over the last couple months, but in general, it's up I think it's 4x up since the launch of Chat GBT and is generally like near all-time highs.
6:16
Like the business is doing very well.
6:17
Um but every CEO needs to contend with the AI uh question, the AI issue.
6:22
How will AI change their platform?
6:24
And AI has already been changing Hollywood.
6:27
I mean, I was I was reflecting on the Avengers.
6:29
When did the when did Infinity War come out? Uh Infinity War. That was uh what 2018.
6:33
Um I just remember seeing maybe he was even in in one of the first ones.
6:40
Um but uh the whole the whole pro CGI process for Thanos, he has this like very distinct large chin.
6:48
So Josh Brolan is the actor that plays Thanos. Is he a mogger? >> He is a mogger. He has this huge chin.
6:55
This m It's actually like >> he's kind of like the OG. >> I don't know.
6:57
He looks like chin implants.
6:59
It's kind of crazy, but it has these like cracks in it and it it has this like very distinct look, Thanos.
7:05
And normally the way the VFX pipeline works is that you go and you put these black dots all over your face and then you wear a helmet that has a camera pointing at your face.
7:15
I think it's a I don't I don't know what type of camera, but it tracks all the points.
7:18
So when you smile like it sees that the that the that the actor that's driving the performance capture is smiling and then that facial movement is transferred.
7:28
So they're recording the lines, they're acting it out, they're giving their facial performance and then that's transferred all the little subtleties of how their how their eyebrows move, all of that is transferred to the CGI character.
7:41
It can look a little flat though.
7:43
Um, so what they did with this is they still have all the points on the face, but then they interpolate from the small points that are on the face into a higher res model. Yes, don't read that. Don't read that.
7:56
Don't [laughter] read that.
7:59
But but it is a good point.
7:59
Um, >> I wasn't even reading the chat.
8:01
I didn't even see that you said that.
8:02
Course I was just looking at this absurd picture.
8:06
>> Oh yeah, it is an absurd.
8:06
So So all of those are tracking markers.
8:07
And then the question is like you have a much higher resolution CGI model if you just transfer with 50 points or 20 points.
8:17
You're not getting all the detail of what a human face actually looks like and the way it moves.
8:20
And so uh digital domain which was one of the many VFX studios that worked on the Marvel series.
8:26
Uh they built a straightup machine learning pipeline. Like they used AI.
8:30
It wasn't a diffusion model.
8:30
It wasn't, you know, an LLM, but they used an a machine learning model to basically translate from the low resolution, just a few dots, to a much higher resolution mesh that then became the performance of Thanos on the screen.
8:44
And I don't know if you remember 2018, the movies, obviously, you didn't see any of these movies, but um I I don't remember like AI back then.
8:54
>> My my prefrontal cortex wasn't fully developed.
8:57
>> [laughter] >> But but but truly like I mean people people did make the oh it's too CGI the build the explosions are too crazy it's too over the top but in general people weren't uh up in arms about like a use of AI or use of too everyone was just like this is a CGI epic this is a crazy
9:14
you know Marvel movie uh like we're fine with all this and there wasn't backlash to that and I don't think that there would be backlash to this type of like AI tool um now obviously Marvel's Avengers that's Disney property but the the same VFX pipeline is being used all over the industry and it will continue to be used. Uh, interestingly, I talked
9:29
Uh, interestingly, I talked to Jason Carman about >> the Carminator.
9:35
>> The Carminator, [laughter] the Carminator about uh about using AI tools in film making because he's obviously making movies and doing VFX and stuff and and I was like certainly if you need to rotoscope out a background.
9:49
So, rotoscoping is where you are basically using like you're you're cutting out like a subject from the background and then just doing like a background replacement.
9:57
That's an example of rotoscoping, but it's over motion. So it's moving.
10:00
So you need to track the hand here, move it over here, track it again, track it again.
10:03
And it can be very very timeconuming.
10:05
Typically this is offshored to like a bo and then they have a whole team of people that are all aiming it and they have some software that's used.
10:12
But I was like this feels like something AI could just one-shot.
10:15
He said that AI was not there at the level that he wanted to deliver.
10:19
He wanted to deliver in 4K.
10:19
And so he went to a team.
10:21
They I think he paid him a fortune. They did it.
10:24
and uh and uh and when they when they rotoscope a head, they actually like draw new hairs on to like kind of create this.
10:31
It's very like artisal still.
10:33
Um but obviously AI can rotoscope.
10:35
You see it in the uh the cap cut edits.
10:41
>> Where's the rotoscoping neolab?
10:43
>> It's uh it's actually runway.
10:43
We've had Crystal Ball on the show and pre-Chat GBT Runway had a fantastic AI rotoscoping tool where you could basically load up a video, put a couple dots on what you wanted to keep and then add and then flip over to the red dots and put those in the background and be like cancel all this out.
11:03
And it would sort of use that as uh like an intuition for the model to to drop out the background.
11:10
background. you could and you could do a really really clear like cut this person out of this video uh just in runway and now that's available in cap cut and edits and that's where you're seeing all those crazy hype reels where you know you'll see the F1 driver like standing up and then it'll like it'll drop out
11:26
the background and cut in a different background and then the F1 driver drops out and it's like this very like skitso edit it's a really cool technique uh yeah you can see is this runway is edit editing tool so it it basically draws a mask around it you just highlight light like what do you want to actually rotocope? This is an example of like uh
11:41
This is an example of like uh motion adding motion to uh video but like you're you're putting in a little bit of your own uh aesthetic taste.
11:50
>> Um quickly let me tell you about reream one live stream 30 plus destinations.
11:56
If you want to multiream, go to reream. com. Um, so goat.
11:57
Um, so, uh, like I don't think Netflix should take a hardline stance on AI broadly because they want to use AI tools.
12:09
Obviously, they've been using AI for recommendations forever.
12:12
Uh, the original collaborative filtering algorithms were machine learning models and that's how you open up Netflix and says we think this would be good for you based on what you watched.
12:22
And it is much more nuanced than just if you like, you know, K-pop demon hunters, we're going to recommend Squid Game next.
12:30
Like, it is it is machine learning.
12:30
Um, and so many of these wrote tasks will be AI enabled, and they already are.
12:36
Um, and there's not going to be I don't think there'll be a crazy uh push back here.
12:42
here. Although it's possible that there's some sort of uh you know comms mishap if especially if a if a director comes out and is like we didn't use any AI in this film because they don't think they used any AI but there's a VFX house that in the motion capture stage
13:00
>> did use AI to upres motion capture data like you could see AI being used in Matt painting for the background and the director doesn't even know because they just said like yeah the background just make the forest a little bit bushier and they think that they're hand painting it. After that, they were using CGI and
13:15
After that, they were using CGI and 3D modeling it and now they're using AI and that sneaks in and then all of a sudden they face some backlash.
13:24
>> But I don't think that's I think that's manageable.
13:26
I don't think that's that that big of a deal.
13:27
Um >> the bigger question is like how does Netflix position itself against YouTube and the UGC platforms is what we talking about.
13:36
So Neil Mohan, the CEO of YouTube has taken a very open stance on AI.
13:38
Uh, and I liked his stance.
13:41
He was like, "We're not going to throw an AI tag on everything.
13:45
Uh, we're gonna let you use AI tools right in the shorts creator. We have VO3.
13:50
We're we're great at this."
13:52
Like, we're going to lean into this and and the algorithm will sort out if you like it.
13:58
And if you think it's slop and you don't want to see slop, the algorithm will learn that and not show you that stuff.
14:02
But for the people that like that, they will be served it.
14:04
Um, but it's a it it is it is sort of a balancing act and there's definitely this like stated preference for like I don't want any AI on my platform. Now, how real is that? We'll see. Um, >> yeah.
14:17
And and and what we were getting into earlier before the show started is is Netflix's decision the bigger like >> the decision that's bigger than just like are we going to in sort of lean into AI or not will just really be do we ever does Netflix ever lean into UDC, right?
14:34
They've been signing some bigger >> podcasts and they obviously work with independent media companies, but >> there's a very big difference between just allowing people to like will they ever have an upload button? It feels like no.
14:45
And I feel like that could end up being their advantage where where there is >> part of what made has made YouTube magical since the very beginning was that anybody could go and put YouTube, you know, upload a video, anybody could be a creator.
14:57
And I think like as the amount of content that gets created, >> Yeah.
15:02
10, 100 x's, thousand x's because of AI. >> It's going to be it. Yeah.
15:07
Netflix could be this like refuge where you're like, "Okay, at least if I go here, I know that there was some filtering process.
15:13
I know that uh this isn't just a total free-for-all." >> Yeah. Yeah.
15:18
The upload button is probably a bigger deal than like AI on Netflix. Exactly.
15:23
That's I think that's a very good thesis.
15:24
Quickly, let me tell you about Console.
15:26
console builds AI agents that automates 70% of IT, HR, and finance support, giving employees instant resolution for access requests and password resets.
15:34
And uh someone was asking in the chat about the linear lineup, of course, linear is the system for modern software development.
15:41
7% 70% of enterprise workspaces on linear are using agents.
15:45
And we have a great show.
15:47
Mark Gurman's coming in person. Miles in person.
15:50
Uh and then we have an amazing uh lightning round coming up.
15:52
And I think we got a surprise guest coming on too from >> fund >> over at Voyager and then >> Mitchell and Gabriel >> Gabe from Rogo.
16:03
Super excited for that one. >> That's fun.
16:05
>> Um so back to Netflix and YouTube.
16:05
So uh YouTube has been on an absolute tear in terms of uh watch time on TVs.
16:12
According to Neielson, YouTube has been number one in streaming watch time in the US for nearly three years.
16:19
And so this has been the backbone of the case for like let Netflix buy Warner Brothers even though they'll get HBO Max like you're merging two seemingly big streaming platforms but the combined watch time will still be lower than YouTube.
16:32
So [snorts] um should be fine from a regulatory p perspective.
16:38
[snorts] Um but the bigger question is like the there's this gap between Netflix and YouTube and at the start back in I mean Netflix is almost 30 years old.
16:46
YouTube's over 20 at this point.
16:48
um back in 2005 like these were seen as like wildly different platforms.
16:53
One was DVDs in the mail and the other one was like a video of a guy going to the zoo on his on his like VHS camera.
16:59
They felt extremely separate and they felt extremely separate for for years and years and years.
17:04
Uh now they are starting to converge especially around um around video podcasts.
17:09
I feel like I feel like YouTube really drove a big boom in video podcasting because the podcast was I think invented by Apple or >> Yeah.
17:21
When I started doing any work with YouTube channels and podcast back in the day, there wasn't a lot of overlap.
17:26
It was very clearly like this these were just different types of creators. Totally.
17:32
And then there was a big shift where it was like wait I'm leaving >> a ton of attention on the table if I'm not uploading to YouTube.
17:38
And that forced a lot of creators to actually get into video. >> Totally. Totally. Yeah.
17:42
There was the pivot to video and then Spotify went really big into video podcasting.
17:46
They went on a I I didn't realize how big of a push Daniel Eker really did around podcasting.
17:54
So, seven years ago in 2019, they acquired three companies.
17:56
Gimlet Media, Ankor, and Parkcast. >> Anchors.
18:00
Uh Mike from Lightseed's company. >> Yeah.
18:03
Uh I mean all three uh very interesting.
18:06
Anchor is more of like a product.
18:07
Parkcast had a bunch of >> Yeah.
18:09
Anchor was like, "We'll make it easy for you to create a podcast because it's still like a lot of people were there.
18:15
It was >> It wasn't like it was impossible to figure out, but there was quite a bit of friction." >> Yeah.
18:20
And and they knew that everyone who wanted to distribute a podcast wanted to distribute it everywhere.
18:24
Um, but they could sort of default you to getting into Spotify as well. So, that did very well.
18:30
And then they signed exclusive deals with Joe Rogan and a bunch of other people.
18:34
Even uh like uh some of the royals also signed a deal.
18:38
I forget [laughter] who they are.
18:39
But but but but they really they really spent a ton of money trying to get into podcasting.
18:43
I think they were successful.
18:44
I mean we we we see a ton of audience on on Spotify.
18:47
>> And interesting the timing.
18:47
Apparently there was uh somebody sent me a chart.
18:52
There was over a million podcasts launched in 2020.
18:55
>> Oh yeah, you saw that chart.
18:57
>> Jeremy shared this with us andart.
18:59
actually fully retraced to now there's like sub sub 50,000 a year.
19:01
That feels like >> you would really have to may maybe they're being extremely explicit about defining what a podcast is.
19:09
But either way, there was if you look at the chart, it looks like there was like an insane bull market in podcast and then a huge correction.
19:19
>> Yeah, it does feel like there was a 2020 boom during COVID.
19:22
I mean, we talked to the folks at Acquired about that.
19:23
Uh, and they said that during COVID, they saw a huge spike in people uh, like just going for walks, throwing on air airpods.
19:33
All of that sort of hit around the same time.
19:35
Airpods were getting to like mass adoption.
19:37
So, people were just throwing on podcasts constantly and it really really grew.
19:41
Uh, anyway, graphite code review for the age of AI.
19:43
Graphite helps teams on GitHub ship higher quality software faster.
19:46
Uh so um there is this chance that with YouTube with the trough at YouTube getting sloppier by the day uh that Netflix carves out more of a unique value prop and you wind up seeing more space between the two.
19:59
So I think of Netflix and YouTube as starting out extremely separate then sort of like coalescing with like Joe Rogan experience and I mean Portoi and uh and Bill Simmons uh both are YouTube dominant and now have Netflix deals, right? >> Yeah.
20:18
Isn't their deal set up so that the video can only be on Netflix.
20:19
Audio is still elsewhere which is >> why why do you remember that?
20:23
Because Porty drilled it into your head.
20:25
said, "If you want video, Netflix, Netflix, Netflix, you know, video, Netflix, video, Netflix, bar stool, video, Netflix."
20:32
And uh and it's true like that.
20:34
And so I would think of bar stool like if I'm trying to watch a bar stool YouTube video or a podcast, I would just go to YouTube, right?
20:40
Uh now Netflix is getting into that.
20:42
So the so the gap between Netflix and and YouTube is getting like pretty narrow like they're they're becoming competitors.
20:47
But there's this question on the AI issue.
20:50
Do they diverge more and is that valuable?
20:52
um the pure AI feeds like Sor and Meta, they haven't really been able to hang on to the top spots in the charts.
20:58
I think so is around 60 70 point uh ranking.
21:02
Um but like it's still too early to call that because the quality will get better and better.
21:07
The audio is still very like clockable when you uh when you hear it.
21:12
Uh and uh there's a lot to be done there.
21:15
But um the struggle will be cons to create like unifying conversations around particular pieces of content that are AI generated.
21:24
I I still feel like the the K-pop demon hunters moment, the Squid Game moment, the Alex Honold Taipei 101 moment, these live events, these like key things that everyone talks about are really really valuable.
21:36
And that's a lot of what's driving the uh the Warner Brothers acquisition is uh you know people still dress up as Batman around Halloween.
21:42
And if you can be the place there, that's way easier than Well, I did I I've been getting AI generated content about a superhero, but my superhero is different than your superhero.
21:55
And so if I wear a t-shirt with that superhero on it, you're like, what is that like [laughter] AI slot man? I'm not a fan of that. I like that.
22:05
So there are there are some things that will like remain remain true.
22:09
So, uh, yeah, I mean, uh, about the upload button, it's almost better to think about Netflix less as being AI free and more about being UGC free.
22:16
Uh, they're paying for curation and a quality bar that's backed up by a brand that's going on 30 years in the business.
22:24
Uh, the AI tools will come, some fully AI generated content will come, but true slop will be filtered out by the Netflix team.
22:30
And I think that that's important for a lot of viewers.
22:35
Anyway, >> let me tell you about Turbo Puffer serverless vector in fulltech search B built from first principles and object storage.
22:43
Fast, 10x cheaper, and extremely scalable.
22:46
>> Should we talk about Saudi Arabia?
22:48
>> Yes, let's talk about Saudi Arabia.
22:50
What's going on in the Middle East?
22:50
Uh, according to Bloomberg, Saudi Arabia is widening its search for capital, turning to some of the kingdom's wealthiest families as the government looks to ease pressure on public finances and fund the next phase of the crown prince's economic overhaul. Okay.
23:07
Uh, I saw this headline and >> was deeply concerned.
23:13
I was like, why uh aren't aren't you guys supposed to be funding the whole buildout?
23:17
like we were kind of counting on you guys to be >> uh you know quite liquid while the rest of us are are over here in America are levering up.
23:28
So uh some seem see seemingly somewhat of a liquidity squeeze.
23:32
This had been reported since uh back in October >> um and we can get into it a little bit.
23:39
They're also raising from >> uh they they've been raising tried to raise money from the Qataris.
23:43
They apparently asked for something like 10 billion and the Qataris in the UAE.
23:50
Qataris threw in 10b allegedly the UAE did not and and there frustration around that.
23:58
>> I thought you were going to say they threw in the Qataris threw in 10b and Saudi Arabia was like and we deployed it.
24:02
[laughter] >> We need another >> time to reup.
24:04
So, as part of these efforts, the PIF gathered about a dozen prominent families on the Red Sea last month to assess their appetite for participating in future opportunities.
24:15
At the summit, which also included others from the private sector, the $1 trillion wealth fund called for more collaboration on deals, the people said, asking not to be identified.
24:22
Government entities, including the Ministry of Investment, have also stepped up outreach to family offices, wealth managers, domestic businesses.
24:30
According to some of the people, local families are being sought after to play a bigger role in partnering with global investors to draw more money to the kingdom, they added.
24:39
Um, so >> years of excess expenditure and subdued oil revenues alongside a tighter lending environment have challenged the Gulf nation's ability to bankroll expans expansive projects planned under the $2 trillion vision 2030 agenda.
24:53
Officials this week said they would postpone the 2029 Asian Winter Games and the government had previously paired back spending on other elements of Saudi Arabia's economic rejig.
25:05
It seems hard to host a winter game that seems extremely expensive to host a winter games in Saudi Arabia if that's what's going on.
25:12
Um >> uh they have don't they have some mountains?
25:16
>> They have mountains there. I don't know.
25:18
Uh against that backdrop, Riad has been stepping up efforts to Yeah, I know that they have some stuff indoors and I guess you can just do everything indoors, but that again that feels expensive. >> Yeah.
25:26
So >> as opposed to like just go to Russia and walk anywhere and you can ski. >> Yeah.
25:31
So they've been developing something with NEOM. >> Oh yeah. Right.
25:36
>> Which is also I guess in the process of >> uh pivoting.
25:40
>> Before we move on, let me tell you about the New York Stock Exchange.
25:41
Want to change the world?
25:43
Raise capital at the New York Stock Exchange.
25:48
Um, against that backdrop, Riyad has been stepping up efforts to look for alternative sources of financing, including a rare loan deal.
25:53
Now, a range of local entities have begun to sharpen their focus on Saudi Arabia's family offices and businesses, which collectively control assets worth hundreds of billions of dollars.
26:02
So, the number of family offices in the Middle East in 2019, 250, 2024, 290.
26:07
Now we're up to 310 and the projection for 2030 is 350 family offices.
26:15
There are big portfolios. The wealth is sizable.
26:18
So the chief executive officer of the national center of for family business in Riyad.
26:24
These entities have long dominated the Saudi economy and close to 95% of private businesses in the kingdom are family-owned. Interesting.
26:30
They're not doing a lot of IPOs over there.
26:33
uh of these many groups are only just starting to form family offices.
26:39
As they grow in size and look to formalize strategies to help spread the wealth across multiple generations, that makes both established and new family offices a prime target for more investment.
26:48
They're naturally looking looking to diversify and want to contribute in areas where we've just scratched the surface.
26:53
In addition, more complex areas of finance are also beginning to emerge, drawing the attention of family offices.
26:59
That includes private credit and industry in its infancy in the kingdom as overstretched banks struggle to meet more explosive needs for financing.
27:04
I like [laughter] that the financing needs are getting explosive at this moment.
27:11
Not expansive, not not expanding, exploding.
27:16
Uh lenders for years have been the primary financeers for individuals, businesses, and government entities looking to drive investment into Saudi diversification agenda.
27:23
But they are starting to pull back as liquidity tightens, leaving many local firms scrambling to find new sources of financing.
27:31
>> Well, we got to bring somebody on to learn more about this.
27:35
>> And I will tell you about public.
27:35
com investing for those who take it seriously.
27:39
Stocks, options, bonds, cryptos, treasuries, and more with damn great customer service.
27:45
>> Shako says, "Yes, >> you're bearish on the US dollar, the token used to buy AKI products." Bold.
27:51
[laughter] >> That's really good. Yes.
27:54
Uh, the US dollar is all over the place.
27:57
Uh, Tether is shaking up the gold market with massive metal hoorde. I did not see this. This is interesting.
28:02
There are roughly three 370,000 nuclear bunkers in Switzerland. That's so many.
28:11
I've seen a video about one of them, but I didn't realize there were so many.
28:14
Uh, the legacy of the Cold War that are now reg rarely used.
28:16
One of them, though, is a hive of activity.
28:21
Every week, more than a ton of gold is hauled into the highsecurity vault owned by crypto giant Tether Holdings SA, which is now the world's largest known horde of bullion outside of banks and nation states.
28:33
Over the past year, Tether has quietly become one of the biggest players in the global global gold market.
28:39
The embodiment of a meeting of the crypto and gold worlds whose shared distrust in government debt is a major factor behind the surge in prices to never before seen highs above 5,200.
28:52
Now it was 5,000 yesterday on the cover of the journal.
28:55
Um, gold is on an absolute tear and yet relative uh relatively little is known about its inner workings or its gold strategy.
29:04
When two of the most senior gold traders quit leading boolean bank HSBC holdings last year, the industry was a buzz about gossip whether about where they would head next.
29:13
Few guessed that the answer was Tether.
29:16
In an interview with Bloomberg, chief executive Paulo Adorno said described the company's role in the gold markets as similar to that of a central bank and predicted that Washington's geopolitical rivals would launch a goldbacked alternative to the dollar. Interesting.
29:32
He revealed that it plans to keep plowing its enormous profits into gold while also beginning to compete with banks in trading the metal.
29:38
We are soon becoming basically one of the biggest, let's say, gold central banks in the world. Interesting.
29:46
This is like the original crypto narrative, right?
29:48
E-old even before Bitcoin.
29:51
Um, >> well, was E-old actually goldbacked?
29:56
>> I think that was the whole pitch was Yeah, it was it was trying to be digital gold and I don't think it ever really got adoption.
30:00
It was very early, late 90s.
30:02
>> Yeah, Tether Tether has a scale now.
30:04
They they also launched their US focused stable coin this week to compete with USDC.
30:09
Yeah, >> the new token known is known by its ticker USA >> uh is being issued by Anchorage.
30:17
>> Uh we I think we had the CEO of Anchorage on at one point.
30:19
Maybe it was really quick.
30:22
>> Um >> Caner Fitzgerald, which already manages the reserves of Tether's mainstay 16 186 billion USDT stable coin, will do the same for the new coin as its designated reserve custodian and preferred primary dealer.
30:37
So USA is already available for trading as of uh yesterday. Uh so we'll see. We'll see.
30:44
Uh it would it be it'd be interesting if like at what point does USDT dep. >> Yeah.
30:55
>> If gold keeps ripping, right?
30:58
>> Does I mean does USDT have a claim on the overall assets of Tether?
31:01
I think that's not what the product is.
31:03
I think I think like the Tether stock would own the treasury. >> Sure.
31:10
But historically when you when you saw a stable coin like DPEG, it was based around the reserves. >> Yeah.
31:18
But I think that the contract is that they will they'll never give you more than a dollar. So it >> No, I know. I know.
31:22
But I'm just saying like stranger things have happened in crypto where >> No, it's always been a very profitable company.
31:28
Uh so certainly certainly bullish for them.
31:30
Uh Bant automate compliance and security.
31:33
Vanta is the leading AI trust management platform.
31:35
Uh there's one funny quote in here.
31:38
So Tether makes his money from its dollar stable coin that is the giant of the sector with 186 billion in circulation.
31:47
The company takes in real dollars in exchange for that USDT token and invests them in treasuries or other assets such as gold, raking in billions in interest and trading profits.
31:56
Processing the physical metal is crucial, Adorno said.
32:01
Uh so much so that the company has taken the unusual step of storing the bullion itself in the former nuclear bunker in Switzerland guarded by multiple layers of thick steel doors.
32:12
And he says it's a James Bond kind of place. It's [laughter] crazy.
32:17
That's a great quote to give Bloomberg.
32:19
It's just like James Bond.
32:21
Uh the secretive nature of >> another another is a positive this is a positive reference. It's okay.
32:26
If you're building a secret bunker to hold all of your gold, I think you can safely use the James Bond analogy.
32:32
Uh the secretive nature of the gold market means that while it's easy to describe broad drivers of investment, it can be hard to pinpoint who exactly is behind the buying.
32:42
China, for example, officially disclosed just 27 tons of purchases last year, but many traders believe it bought much more.
32:49
Such is the scale of Tether's disclosed purchases uh that some market watchers have pointed to their role in shifting global prices.
32:56
The purchases likely contributed to gold's 65% rally last year, Jeff said, uh describing Tether as a significant new buyer, which could drive sustained gold demand.
33:09
Still, Tether is only a small part of a much larger rush from investors into gold with central banks and ETF investors collectively buying more than 150 uh 1,500 tons of metal.
33:19
I wonder if this is moving the gold watch market.
33:21
Do you think uh you know a uh the the Texas Timex is booming on the back of gold spiking? >> We'll see.
33:29
Yeah, I think a lot of prices are still down pretty dramatically in the 2020 2021 era, but uh I definitely I mean you got to be a little bit scared right now if you're in the business of manufacturing >> totally >> gold watches and other precious metals just given that uh your input costs are are going >> uh you know I'm sure they have like one or >> you could imagine one or two years worth of of of supply.
33:53
So they're not they can be somewhat insulated, >> but uh price prices will go up or at least costs will go up. >> Yeah. Well, Finn.
34:03
ai the number one AI agent for customer service.
34:06
If you want AI to handle your customer support, go to Finn. ai.
34:09
Um I heard a very funny uh very funny interview with an actor who was talking about why he uh always wears a gold Rolex and he was calling it a helicopter watch. Did I send this to you? Yeah.
34:22
>> And he's saying that like >> this was Army Army.
34:23
I think it's Army Hammer.
34:24
Uh, and he's saying uh >> most like obscure such such a funny situation, but he's like if you're ever in a crisis, you're on the top of a building, the zombie apocalypse is upon you, someone shows up with a helicopter and they're going to save a few people.
34:40
Uh, everyone knows what a gold Rolex is and they know that that's valuable.
34:42
But if you're >> you're not going to have time to be like, "No, >> it's an FP.
34:46
[laughter] It's a it's a PC."
34:48
You know, like let me explain high herology.
34:52
Here here's here's aftermarket prices. >> Exactly.
34:56
It's like gold Rolex is a store of value.
34:59
It's always going to trade and now it's probably going to trade even higher. Yeah.
35:03
>> Anyway, uh Cisco on February 3rd, the Cisco AI Summit brings together leaders from NVIDIA, OpenAI, AWS, and more to discuss the future of AI and the economy.
35:13
The whole thing will be live streamed and we'll be there for a gig.
35:16
Uh Paula says SF escape room called the permanent underclass and it's just a room with a laptop and claude code installed.
35:25
>> How do [laughter] you get there's something here that's funny? >> So good.
35:29
>> Have you ever done an escape room?
35:31
>> No, I have never once in my life thought that that would be fun.
35:41
[laughter] nor thought that I would I this is how I want to kill time.
35:47
>> Yeah, it's like an hour.
35:47
Well, if you're good, you get out faster, right? >> Yeah.
35:51
Have you Are you >> I'm not into them, but I've done them and they're fun sometimes.
35:55
It depends on like if it's a well structured one. But I like puzzles. Like it's fun.
36:00
It's like a fun puzzle to like figure out, but you can very quickly get caught in like just like overthinking what's going on and being like, "Oh, it must be some like complex math thing."
36:09
And it's like, "No, you actually just needed to like pre like press this lever instead of like analyze the situation or something."
36:15
>> Are some people in there using their phones to try to figure stuff out?
36:17
Cuz I imagine you could just say >> I imagine there's only a picture.
36:22
>> There's probably only a series of rooms that you're >> Yeah.
36:27
Well, usually you you start in one room.
36:29
They're all different, but uh oftentimes you'll start in one room and then you'll there'll be a series of puzzles, locks and keys and whatnot.
36:35
And then oftentimes like you'll unlock something and then you'll progress to another room and then you'll progress to a third room and then you'll finally get out of like a series room because it's a lot of like they're pretty easy to set up in sort of like a defunct office space or like you know storefront that's just kind of like you know going in between.
36:54
It's like the spirit Halloween of commercial real estate.
36:56
Like you just like anyone can come in and just say like oh we'll be in there for like a couple months.
37:02
It's not super permanent.
37:04
The buildout's pretty simple.
37:04
It's mostly just like some some walls and decorations and like some creativity on the on the on the uh on the puzzle side.
37:12
>> Says, "Can you guys put Tyler in an escape room and see if he gets out before the show ends?"
37:17
>> I mean, it'd be very hard to find a three-hour long escape room.
37:19
I think most of them they aim for like 45 minutes.
37:23
>> Well, maybe we need to make one.
37:24
>> One of our guests had to leave the Ultradome and go straight to an escape room.
37:28
So, you know, there's there are escape room fans among the TVPN army all over the place.
37:34
Have you done one, Tyler? >> I have not.
37:36
>> Are you interested in it?
37:36
You're a speed cuber, so you know. >> Yeah. I mean, could be fun. I don't know.
37:41
I'm kind of in the Jordy camp, though.
37:42
I've just like never >> Yeah.
37:44
>> been uh super >> It's It's very like I don't know.
37:46
Maybe it's great with It's probably great with like uh kids who are maybe like 8 to 10 who can do the puzzles.
37:54
Like family event would be fun.
37:56
Uh, and then it was it was a common thing like uh when when they came out there was definitely like a boom in like uh in the in escape room.
38:04
>> Bobby in the chat says escape room would be a good benchmark for AI.
38:07
Certainly certainly a humanoid.
38:12
>> No, I mean you could go in there, take pictures and be like, "Help me solve this. Help me solve this. Help me solve this."
38:16
Like >> I know, but I want to I want to see a humanoid run through it. >> Let's call OneX. See what they can do. Call any of them. >> Hit them up.
38:23
>> Figma Figma make isn't your average vibe coding tool.
38:25
It lives in Figma, so outputs look good, feel real, and stay connected to how teams build, create codeback prototypes and apps fast. >> Uh, moving on.
38:33
Uh, lots of reaction to Claudebot.
38:36
Very much enjoyed our uh our interview with the creator of Claudebot, now Moltbot. Um, uh, yesterday. Uh, >> one of >> Yeah. >> Yeah.
38:46
Incredibly, he had some wild lines, but incredibly refreshing kind of conversation and viewpoint. >> Totally.
38:53
uh is totally counter to the entire philosophy that I feel like a lot of people at least on the west coast are in the way that they're approaching AI right now and the way that America is approaching AI.
39:05
He's like >> we asked you know you think somebody will like you know fork what you're doing or clone it and he's like yeah I'm sure they will like I don't care.
39:15
>> Yeah, I'm building this for myself. >> Yeah.
39:16
and he's like, "I have enough I have enough money." >> Yeah.
39:19
>> Uh and I'm I'm super >> It's going to be really fun to to follow along.
39:24
Uh I have a feeling that he's just going to keep launching a bunch of random projects because he clearly is just in it for the love of the game, but I do hope that uh I do hope that he can get a lot more resources and really scale up >> I don't know that he needs more resources.
39:39
Like he has more agents and whatnot.
39:41
whatnot. Like the the interesting thing about Claudebot as a product is you you you download it from GitHub, it installs and it's it it it has all these different integrations and it does something that's very complex and it has you know all this safety text and
39:58
different there's a website and there's a community page and all of that feels like okay yeah this is like a 10erson startup they probably worked on this for a year but it's like no it's one person and it's like three months because the guy is you know using agents. You would
40:09
You would think that he would build he his his big complaint was the security like inbound security researchers asking him stuff. >> Yeah.
40:20
>> It's like >> bought >> bought it.
40:23
>> I I I think I think he will. I think he should.
40:25
But yeah, >> he said um >> uh he he posted uh uh December 26 he had his like codeex uh like dashboard up.
40:30
So he said he he's done um 250 billion tokens >> which is like probably top 10 users only from from >> So so I I I don't know that he needs way more resources.
40:43
I mean uh he should be able to to to get donations if he needs them get credits or something.
40:49
them get credits or something. I don't know like the that the the financial strain on that business does not seem it seems like it's more constrained by his ideas uh you know how he's thinking about designing the system integrating things rolling it out buddy buddy of mine Michael uh watched the interview
41:06
and and he said that he he's had Cloudbot set up he set it up but right at the beginning of January and was initially like just kind of >> got a little frustrated using it but has now got it uh he's got it set up so that able to make phone calls on his behalf specifically wants to get uh Hillstone reservations. Yeah. So, he just like Yeah.
41:25
So, he just like basically >> Wait, what's Hillstone? >> The restaurant. >> Okay. >> Yeah. Yeah. Yeah.
41:29
So, he's just like going to use it to start getting like reservations at different restaurants. >> It's so funny.
41:34
Google has a product for that that does an AI phone call, but for some reason it just hasn't really rolled out that or it just hasn't gotten to like adoption. I don't know.
41:42
there's some sort of there's some sort of like mimemetic like I think people seeing the clip where he's like explaining how he's feeling the AGI really really hit people.
41:51
Uh >> obviously he's uh has it make him a daily brief >> that feeds into an RSS like on a podcast.
41:59
So in the morning he can just listen to like a five minute podcast on like what his day is like, things that he should be responding to.
42:05
kind of like uh it seems like you could have >> worst podcast ever.
42:09
Your enemy texted you something [laughter] and you will be you'll be deeply upset when you see what happened.
42:16
Everyone's praying on your downfall. [laughter] No, no.
42:20
Security is important with with Claudebot, now Moldbot.
42:23
Uh Crowd Strike is also important. Your business is AI.
42:29
Their business is securing it.
42:29
Crowd Strike secures AI and stops breaches.
42:31
a 10 a random 10person team in Paris just dropped what looks like it looks like an e something superior to clawbot according to uh chubby on x it's called twin the AI company builder >> mercier yeah they raised a $10 million seed round >> they have over a hund agents deployed >> okay >> so Tyler give it a spin check it out twin labs >> there's a community note on here this is undisclosed advertisement.
43:03
Um, but Doug over at semi analysis fabricated knowledge says, "I think that Claudebot is going to be a moment and yeah, someone's going to do this."
43:12
someone's going to do this." And I I'm still wondering about like how quickly a business can actually scale when you're getting constantly hammered with toos violations from every mag 7 legal department constantly like hey yeah we noticed that you built a CLI for the web
43:35
interface on WhatsApp and we don't want you to do that and and it's one thing if it's like an open- source repo that people are running themselves and can change and can edit and fix and tweak versus like, hey, you're a corporation that we could potentially sue or tie up in litigation or go on a press tour around like there. It's just
43:53
It's just a very different dynamic. >> Yeah.
43:56
And the prompt injection risk as well. >> Yeah, totally.
43:58
>> Yeah, totally. It's like does that does that liability fall back on the company uh that you know and I I'm excited to talk to Mark German about how he's processing this with the uh with what will happen with Siri the timelines there what will be integrated because Apple and the iOS ecosystem should have
44:15
a lot of the same functional hooks into these products but >> we talked about this Monday the there was a team that built a product which sold to Apple and became shortcuts and then they built software ware applications incorporated in a product called Sky and they were pre-launch and OpenAI acquired it. It was focused on
44:32
It was focused on integrating AI into the operating system and so you can imagine OpenAI behind the scenes has been cooking on a lot of stuff like this but >> uh this isn't the kind of thing that you can just at OpenAI scale just say hey we're going to just ship this and see what happens. Yeah.
44:48
>> Whereas a startup or an open source project. >> Yeah.
44:50
Uh let me tell you about Lambda.
44:53
Lambda is the super intelligence cloud, building AI supercomputers for training and inference that scale from one GPU to hundreds of thousands.
44:59
Um there there are a lot of people building stuff on Moltbot around Moltbot around the concept.
45:06
Uh another there's actually someone in the chat that's working on a Claudebot for the cloud.
45:10
Uh he says he has a name.
45:13
I'm very excited to uh to uh learn more about his project.
45:17
Uh I think his name is Jeff in the chat.
45:19
Um, Brexton says, "Yes, that was fast." Yep.
45:23
Software is cooked because uh Kayn Y says, "Introducing Moltbot for Teams.
45:28
One-click connections to all the apps.
45:30
Client side encrypted and builds your team's memory."
45:32
Uh, get your Momo is the is the uh product.
45:35
I wonder like this feels like almost something where if you've been building a product that's like this, you should potentially reintroduce it on the back of all the multbot hype and and you know if you're a if you're like a you know multi-entled agent that can touch a lot of different systems.
45:56
Well, we'll put it in terms that people understand and say, "Hey, like this is, you know, this is clawbot or moltbot for the enterprise and it has these, you know, functions and these benefits and this security approach."
46:10
Um, but clearly a lot of people will be uh focused on this.
46:12
Let's uh click over to re-watching the interview with the Clawbot creator Peter Steinberger.
46:17
Um, because I want to hear about his AGI moment, the AGI achieved moment.
46:23
First, I'm going to tell you about Gusto, the unified platform for payroll benefits and HR built to evolve with modern small and mediumsiz businesses.
46:31
And then we will play the video from Claudebot Creator. I want to see.
46:37
>> So, in November, >> yeah, >> I I don't know, you know, I wake up every day.
46:43
I'm like, okay, what do I want to work on now? What would be cool?
46:44
And then they was like, "Okay, I I want to chat with my computer on WhatsApp because because if my agents are not running, if they're running and then I go to the kitchen, I want to check up on them or like I want to like do little prompts." >> Yeah.
46:57
>> So I I just hack together some WhatsApp integration that literally receives a message, calls cloud code, and then returns what cloud code returns. One shot. >> Yeah.
47:11
>> And it took like 1 hour and it worked.
47:14
like, "Well, okay, that's kind of cool."
47:16
>> But I usually use prompts, like a little text and an image, cuz images are like they often give you so much context and you don't have to type so much. >> Mhm.
47:26
>> So, I feel like this is like one of the hacks where you can prompt faster, just like make a screenshot.
47:30
So, the agents are really good at figuring out what you want.
47:33
Um, so I hacked together images and then I I was on a trip in Marrakesh with like a weekend birthday trip >> and I [clears throat] I found myself using this like way more than I than I saw but not for not for programming.
47:48
It's more like, hey, there like there's like restaurants.
47:49
Um, because it it had Google in it and it it could figure out stuff and it's like especially when you're on the go, it is like super useful.
48:00
And then and I wasn't thinking I was just sending it a voice message, you know, but I didn't build that.
48:04
There was no support for voice messages in there.
48:09
So So the reading indicator came and I'm like, "Oh, I'm really curious what's what's happening now."
48:12
And then after 10 seconds, my agent replied as if nothing happened.
48:17
I'm like, "How the f did you do that?"
48:20
[laughter] And it replied, "Yeah, you sent me you sent me a message, but there was only a link to a file with no file ending.
48:27
So I looked at the file header.
48:30
I found out that it's Opus, so I used FFmpeg on your Mac to convert it to to Wave.
48:34
And then I wanted to use Visper, but didn't have it installed, and there was an installed error.
48:39
But then I looked around and found the OpenAI key in your environment."
48:42
So I sent it via curl to OpenAI.
48:44
OpenAI. um got the translation back and then I unresponded >> and that was like the moment where like >> wow >> you know it's like that's where it clicked these things are like >> damn smart resourceful >> beasts if you actually give them the power beasts >> app loving [laughter] >> the ad at the end I love it yeah people were reacting to this people are not are
49:09
genuinely not ready says Vtorio lots Well, that that's the um >> one way one way the reason that I I feel like that is such a powerful moment and I'm glad that he shared it is if you give somebody if you give if you're talking with a model and you give it a task >> and then it just hits a dead end >> it's just incredibly like that's sort of like people are very used to that right now. >> Yeah. >> Yeah.
49:38
>> And it's not that it needs to be that way but it's just kind of like the steady state.
49:42
Yeah, people are used to like, okay, I I know what models can do.
49:46
>> That's effectively the agent having real agency. That makes it an agent. Yeah.
49:50
It's saying like, well, I didn't know how to do this or I was confused and I tried a number of things until I did I did what you wanted, right?
49:56
And that's like what you want out of >> that's what you want out of a team member, right?
50:01
If you're working with somebody on a project and they have a task, >> Yeah.
50:05
>> you don't they don't just try one thing and come back and or just say like, I actually can't handle this because the file type is wrong.
50:10
It's like convert it, figure it out. >> Yeah.
50:13
>> What was your reaction, Tyler? >> Yeah.
50:15
I mean, it's like pretty insane. Definitely.
50:16
It definitely raised my uh you know, chance of permanent underclass. [laughter] >> Oh, no.
50:23
>> You know, it's making me a little worried. >> Yeah.
50:24
Will Will says, "It's over. It's over. We need to move."
50:27
And there are a lot of people quoting this.
50:29
G Fodor has the Eleazar Udicowski uh meme and uh lots of people here.
50:35
Um they uh uh lots of people were you know interesting that uh that that he uses uh codeex here. Uh Run said Codex 5.
50:46
2 is really amazing but using it from my personal and not work account over the weekend taught me some user empathy. Lol. It's a bit slow.
50:53
And Euchin Jin says, "Every time I ask my OpenAI friend, when will you beat Claude at coding?"
50:59
They say, "We already beat them."
51:01
I think Sam is realizing speed, not intelligence, is Codeex's blocker to a Claude code moment.
51:06
Cerebrus chips, Cerebris chips might unlock a Codeex moment.
51:11
Yeah, that would be very interesting.
51:13
Uh and that's and that's one of the like the models are so powerful now that there's there's uh that there's so many moments where uh you know it's going to deliver it can deliver uh because it can work around all these problems.
51:26
But if you're like I need this done in five minutes I'll just do it myself. >> Yeah. Like I uh 5.
51:32
2 Pro is like an incredible model but it just takes like >> five six 10 minutes every time you prompt it totally because it's kind of like thinking or you know >> deep research type model. Yeah. Yeah.
51:44
>> Well, let me tell you about Labelbox reinforcement learning environments, voice, robotics, evals, and expert human data.
51:50
Labelbox is the data factory behind the world's leading AI teams. I like the fireworks.
51:55
>> Do some fireworks for that. >> I like the fireworks.
51:56
Um, what else is in here?
51:59
Uh, so you already mentioned this, Tyler, but uh, Peter's probably in the top 10 users of Codeex at the moment.
52:05
Over 250 billion tokens in a a few months is a lot.
52:09
So, how much does that actually cost?
52:11
That feels That feels like you're up in the hundreds of thousands of dollars. >> That says it here. >> Yeah. How much does it say? >> 51,000.
52:22
>> That's [laughter] not much. Wow. Yeah.
52:23
Usage cost and he has a streak going of 74 days. No days off. Yeah.
52:26
So, $51,000 for this type of result. Pretty remarkable.
52:33
Uh, so where do we move the goalpost now if >> I'm so I'm so what what are you unhappy?
52:40
I know you're un I know there's gonna be something that you're unhappy with about AI.
52:43
Like you can't be satisfied. >> Me? >> Yeah. >> Me? >> Yeah.
52:49
>> Permanently dissatisfied.
52:50
>> Well Well, part of it's a bit >> Oh, yeah. I know.
52:51
Obviously, but >> but I have to I have to we got to move the goal.
52:56
>> I have to pro I have to provide like an alter opinion.
52:57
But of course, we have the goalpost here.
53:00
We have been sitting in one place. Yes.
53:03
>> I guess I guess what I'd be interested in is Peter and his team >> built a hit product, >> an entire, you know, this novel >> experience. >> Yep.
53:15
>> Uh that has taken the world by storm.
53:17
And how much how much did we know they only spent, you know, few months on it?
53:22
>> How much did they actually how many people actually worked on it >> and what was their total how much did they spend in total on tokens, right?
53:29
because they're spending some with >> you sort of see that from the GitHub commits, right?
53:34
>> Probably for some sort of average. >> Yeah. Yeah.
53:36
I'm sure you can do that.
53:36
I mean, just based off that it's like, okay, $51,000 on 250 billion tokens >> and a lot of those tokens are probably using CloudBot to do things, right?
53:44
Not just not just building it, right? >> Yeah.
53:49
I mean, that was over how many months?
53:51
Like all those tokens were not just on Claudebot. Yeah.
53:54
>> That was just all of his like >> hundreds of And he says he talks to Opus45 all the time, too.
53:57
So, I wonder what he's what numbers he putting up there.
54:00
But still, it's interesting that it's not in the millions.
54:01
Like, it's it's it's pretty accessible. Um I don't know.
54:05
We need to we need to find a new AGI benchmark. >> Okay.
54:08
So, I I think um maybe it's something like this, right?
54:10
So, I think the main one of the main takeaways I have from this is just that um >> like big model companies can't really release an equivalent product just because of the integration thing, right?
54:21
Magum companies are not going to like talk to each other.
54:24
you're not going to get open in WhatsApp like this.
54:28
>> Um, >> so it really >> But do you think they would go so far as to say >> we are not going to allow you to navigate our applications using using something like >> I mean that's literally happening with like the New York Times does not allow OpenAI to browse it with a bot like you can go >> Yeah.
54:50
But it's different having your local device being effectively used by >> Yeah.
54:58
No, it it is >> a computer just using itself >> at at the same time.
55:01
If if OpenAI has the Chatt app right now, if you go to the Chat PT website and you say, >> "Hey, uh, I found this I'm a subscriber to the New York Times and uh there's an article in the business section that I want you to summarize for me. Here's a link.
55:19
can you open the link and uh turn into some bullet points for me?
55:22
It'll just say like, "No, no, no. I can't go over there."
55:25
Like they they said, "Don't go. I'm not going."
55:26
Right now, if I download the Chat GPT app and it's able to run that and maybe the New York Times can't block it, the New York Times still has the ability to like sue OpenAI and like be a headache.
55:39
So then they can say, "Hey, enforce the same thing."
55:43
talking about a situation where the user, let's say they have a Mac Mini, they're running Moldbot, >> they on the computer, they've given Moldbot their New York Times login. Yes.
55:53
>> So, it's fully logged in. >> Yes.
55:54
>> And they're paying New York Times subscriber.
55:57
>> I I would be I think users would be upset that you're saying >> if the New York Times says we're not going to allow we're going to figure out a way to try to Yeah.
56:04
Yeah, >> I don't even know how you would go about uh trying to actually stop that activity.
56:09
But I'm just saying even >> even I can imagine users are going to users are basically going to demand I if my data I don't care if my data is stored in the cloud.
56:21
Yeah, >> I want to be able to access it on any device that I'm logged into even if I'm not physically present with the device. Right.
56:29
I I >> and so I just I just think it's going to be a really really really tough argument >> for some of these larger companies to say you can't you can only access your data if you are physically moving the mouse yourself. >> Yes.
56:44
I want to get Tyler's response, but first I'm going to tell you about cognition.
56:47
They're the makers of Devon the AI software engineer.
56:49
Crush your backlog with your personal AI engineering team.
56:52
Tyler, what are you thinking? >> Okay. Yeah. So so I agree.
56:54
I I think basically what you're describing is that all these companies need incredible deal guys, right?
56:59
This is the whole thesis like >> the dealer with open source with this like hacker culture you can kind of always get around these rules.
57:05
Is it like maybe breaking terms service like def yes definitely these big companies can't do it.
57:11
They need more deal guys, right?
57:13
But >> it's like hard to find these deal guys.
57:16
You need the AI to become the deal guy.
57:17
So you need a Claudebot deal guy.
57:17
That's the new benchmark.
57:20
>> That's the new benchmark.
57:21
>> You're going to be good at making deals. >> Okay.
57:23
quad AI that can do a deal between two mag seven companies. Here we go. >> There we go. There we go.
57:29
Move it to the other side.
57:31
>> Time to move the bench. >> See you.
57:34
>> I want to see >> uh Tyler, why do you I kind of want to take this a little bit further just because >> I don't know like h how do you actually enforce this?
57:48
Is is Google Drive going to say you need to have your camera on and we need to see that you're sitting in front of your computer actually using it?
57:57
>> I mean, you can always do these like bot detection stuff with like whatever it's coming from the same IP these things.
58:03
>> I think it's like kind of a constant race.
58:06
>> Um, but like I think the open source stuff will if you like really grind it out, you can kind of always get around this stuff.
58:12
>> And so like the only only real solution is to have like actual I think deals between these like different companies.
58:18
Yeah, I I think that if you if you have an app from a big company that enables you to do something like like like if if Chrome ship started shipping with a torrent browser, something that would allow you to manage torrents, which are not illegal, like that's just software.
58:40
You can just download torrancets that are not piracy.
58:42
Like I can just be like, "Hey, I put up a torrent of TBPN content.
58:47
and anyone can download it.
58:47
Uh but just by virtue of the fact that it's so prevalently used to pirate media, they would face pressure from media owners like Warner Brothers and Disney and they would get angry letters and they would strike a deal and say, "Okay, yes, like we're not going to enable this."
59:06
And so even though so so if if any of the big tech companies launch an app that enables behavior that other apps don't like, they will have leverage and they all have deals together and they all have legal teams and there's a bunch of different pressures that they can apply and it's this like constant negotiation.
59:25
Oh well, you know, maybe we won't renew this deal or this contract.
59:30
Maybe we'll go with someone else.
59:30
If you're the bad actor, you'll get pushed out, right? So I don't know.
59:32
I think it's going to be a big a big debate.
59:35
I I I I don't I don't I would be surprised if I if if any of the big labs just launch something that can actually just go and do anywhere and do anything because uh the push back.
59:49
>> But do you think do you think this is something that will like why would this just not happen at the OS layer? >> Yeah, I don't know.
1:00:00
>> I mean then then you need Apple to like actually lock in.
1:00:02
you do, >> which is like >> Yeah. Or Satia. >> Yeah.
1:00:06
And that's why we're having Mark German on the show to break down Apple's strategy and what they're doing. Anyway, interesting.
1:00:11
Uh, MongoDB, choose a database built for flexibility and scale with best-in-class embedding models and reankers.
1:00:18
MongoDB has what you need to build what's next. Uh, Kimmy 2.
1:00:23
5 is out and, uh, Alex Chima is running it on his desk.
1:00:27
It runs at 24 tokens per second uh with two M3 Ultra Mac studios connected via Thunderbolt.
1:00:36
He went much bigger than the Mac Mini.
1:00:38
And uh Jo, >> by the way, apparently Kimmy likes to be refers to itself as clawed. [laughter] >> Yeah. What happened here?
1:00:45
Tyler, can you break down what's going on with Kimmy? >> Yeah.
1:00:48
I mean, it's always kind of unclear with the Chinese open source models like what they're actually doing to train the models. So there's like Yeah.
1:00:55
So, so when you ask the models like introduce yourself, it's like hello, I'm Claude. >> Yeah.
1:00:59
>> So, there's probably like two scenarios.
1:01:01
One is that they they trained on the outputs of like Opus or or anthropic model, right?
1:01:06
>> You'd think that they would just find and replace in the training data though, you know, like that would be pretty easy to just be like, okay, we're going to scrape a ton of a ton of responses from Claude.
1:01:16
Let's make sure we do a find and replace on Claude so that if it says, hi, I'm I'm Claude. I'm helpful assistant. It just changes that.
1:01:23
That's just like an >> Yeah.
1:01:25
I mean, you would expect that that Kimmy knows that it's going to call itself Claude. >> Yeah.
1:01:29
>> Um so it's like why would they allow that?
1:01:31
Maybe they're just basically like mogging >> anthropic like oh yeah we have Claude.
1:01:35
>> What are you going to do about it? >> Yeah.
1:01:36
Like what are you going to do? >> We stole Claude. >> Yeah.
1:01:38
And then there are also some rumors like very unclear if this is just like completely fake headlines but like maybe there was some >> leaked checkpoint that somehow got out to China and then they they did a fine tune on it or something. >> Yeah.
1:01:51
Um >> that would be pretty pretty crazy if uh >> yeah people have said it before for other models.
1:01:55
It's I'm not really sure how how reliable they are. Yeah.
1:01:59
>> But it seems like fairly very likely that they just trained on Opus outputs. >> Yeah. Yeah.
1:02:04
>> Cuz you kind of people said uh DeepSc did that originally with uh with chat GBT. Um >> yeah. >> Yeah.
1:02:11
I talked to one uh one technologist about like what it takes to reverse engineer like the GPT4 API and it's it's a surprisingly low amount of outputs to sort of interpolate all the weights or something that approximates it.
1:02:24
So it's clearly sort of a sort of a game back and forth a uh yeah just like a constant war.
1:02:30
There's something I think it is quite um >> I think it's probably a good sign for US AI labs that um the Chinese models are essentially just like completely based off the US ones.
1:02:41
Like they're not actually training these models from scratch. >> Yeah.
1:02:45
>> They're not doing these incredible uh you know training runs for for much cheaper like people say.
1:02:48
Um they're actually just they're basically just like copying off off the US.
1:02:53
>> Do you think that's because of just chip restrictions or or actual like architectural hurdles?
1:02:58
Yeah, it's probably all of them.
1:03:01
Um, like it's hard to to figure out the architecture.
1:03:03
You don't have the chips to to try these like different, you know, science projects, training runs. >> Wow.
1:03:10
The the the Chinese models lag by exactly how long it takes to to to scrape the API basically and do a do a fine tune. What a coincidence.
1:03:17
That's >> Dean Paul has been on a tear the last 24 hours.
1:03:24
>> The post >> he said people significantly underrate the current margins of AI labs.
1:03:27
It's yet another way in which pattern matching to the technology and business trends of the 2010s has become a key ingredient in the manufacturing of AI. Copium >> copium.
1:03:39
>> And he says >> this Derek Mohler uh post was very informative here.
1:03:42
Just look at the market clearing prices on inference from open source models and you can tell the big labs pricing has plenty of margin. Uh deep infra has GLM4. 7 at 43 cents in $1. 75 out.
1:03:54
Sonnet is at $3 in $15 out.
1:03:59
How could anyone think Anthropic isn't printing money per marginal token?
1:04:01
And Dean Ball says the reason they think labs lose money is because 10 years ago some companies in an entirely unrelated part of the economy lost money on office rentals.
1:04:12
We work and taxis Uber and everyone thought they would go bankrupt because at that time another company that made overhyped blood tests did go back did go bankrupt.
1:04:20
That is literally the level of ape-like pattern matching going on here.
1:04:25
The machines must look at our chattering classes and feel great appetite. >> Yeah.
1:04:31
You you could also have uh like seen that the margins from the whole open code thing with anthropic. Oh yeah. Right.
1:04:36
Where there was um >> open code.
1:04:38
So anthropic they have cloud code right there.
1:04:41
You can get the cloud subscription and you get like free uh cloud code tokens >> and then you could use those to O for open code and they they removed that. Yeah.
1:04:50
Um, and the whole reason you would want to do that is because the like uh cloud code tokens versus like actual cloud API, yeah, >> is like 10x difference.
1:04:57
So there's like a massive markup to the actual API. Mhm.
1:05:00
>> So you assume that like unless Anthropic is losing just insane amounts of money on cloud code, which maybe they're losing some amounts of money, it's like >> well well you look back at the I forget which interview Dario was doing where he was trying to get people to think about like you think of each model as a company where you spend all this money on training which is capex.
1:05:16
And then when you're actually running the model, it's very profitable.
1:05:19
But if you look at the business as a whole, you have massive massive losses from training and and stockbased comp and and you know hiring,400 you know of the best engineers in the world.
1:05:32
So if you actually look at at on a company companywide you have you know continued scaling massive losses. >> Yeah.
1:05:41
>> But the important thing is at the effectively at the product level when they're selling the product they actually are making money.
1:05:45
So in the S1 good to you and to me looks like training clearly broken out as capex solid gross margins above 60% something like that on inference and then you know all of the AI copium can probably subside if openi or anthropic go out with an S1 that shows really solid inference margins >> and and if it comes out that it's Oh, their inference margin is like 10% or something, then people are going to be panicking.
1:06:19
>> To be clear, I think these will be some of the most special S1s >> that have ever >> graced the financial >> the capital markets. >> Be great.
1:06:27
Let me tell you about Sentry.
1:06:29
Century shows developers what's broken and helps them fix it fast.
1:06:31
That's why 150,000 organizations use it to keep their apps working.
1:06:35
And without further ado, we have Mark German, the manager, managing editor of Bloomberg >> in the Ultradome. Welcome to the show.
1:06:46
Thank you so much for taking the time to come on down there. >> Yeah, right here.
1:06:51
>> Uh we'll have this microphone for you.
1:06:53
You have a diet coke, so you're locked in. >> We're ready to go. >> How have you been?
1:06:55
How's your new year going so far? >> Oh, freaking great. >> Yeah. >> Yeah.
1:07:01
Had a wife and I had a baby at the end last year. [applause] >> Gong.
1:07:09
I need to get one of those uh things that you guys make on the Twitter.
1:07:11
just me >> the card >> baby.
1:07:15
We'd be happy to do that. We're happy to do it.
1:07:17
We never We never Yeah, we never We never know.
1:07:19
Some people Some people want to be, you know, more lowkey about >> Oh, I am lowkey about it.
1:07:23
I just figured, you know, >> but but how do you how do you rate parenthood?
1:07:26
How does it uh how does it feel being in it versus, you know, all the expectations that that people have?
1:07:34
just taking care of this >> individual and they they're being so reliant on us and getting to to teach them.
1:07:44
But >> uh shout out to my wife for being the one >> really >> pulling the strings there and taking care. So it's great.
1:07:52
>> Any feel like I feel like reporting and parenthood are especially hard to balance in some ways.
1:08:02
>> I would imag Well, I would Yeah.
1:08:02
No, I I would I would just like obviously it's completely possible and you're you're clearly doing fine, but the nature of the work means that like a story is happening and you want to be the first to provide the best coverage and sometimes, you know, crying on you >> and you might not be able to pick up the phone, >> but you get through. >> You know what?
1:08:23
People have been hearing crying babies over the phone for uh >> It's true. >> for for forever.
1:08:28
So, they'll have to deal with it.
1:08:30
I just can't wait till we go on an airplane. >> Oh yeah.
1:08:33
>> Like I already told my wife like we're gonna get like banned from American Airlines.
1:08:37
Like >> No, I've said this on the show before.
1:08:38
I like the funniest thing is growing up I assume that when a baby was crying on the airplane, it means that the parents were bad parents.
1:08:46
Like they were doing something wrong. Totally. Totally.
1:08:47
And in reality, babies just cry for a million reasons.
1:08:51
And it's totally possible that there's no solution.
1:08:55
Sometimes they just need a >> Yeah.
1:08:56
I I saw like an Instagram reel of of a mother talking about her kid and the kid was crying in the background.
1:09:00
I was like, why would you take advice from her?
1:09:03
Like that's such a bad moment.
1:09:04
>> Yeah, it has nothing to do with it at all.
1:09:06
Like now like the baby's always crying. >> Hungry. They have gas.
1:09:09
They don't like the sweater they're wearing. >> A million reasons.
1:09:14
>> You know, they didn't like my article.
1:09:16
You know, [laughter] who knows?
1:09:17
But anyways, anyway, let's talk about Apple.
1:09:19
>> Let's talk about Apple.
1:09:19
Uh I mean, we've been talking about new Siri expectations there.
1:09:24
The big news yesterday was uh Claudebot.
1:09:27
Uh have you been uh do do you think there's anything about have you been how have you been processing the cloudbot now moltbot story?
1:09:34
Uh what's possible there?
1:09:37
Expectations around AI assistance feel skyhigh in the open source community.
1:09:41
Uh what do you do you think like anyone at Apple's like updating on cloudbot and molbot and what's possible?
1:09:49
>> Well, let me just take a step you know with Apple and AI. >> Yeah.
1:09:52
So in 20 what was it 2018 they hired John Gene Andrea.
1:09:55
He was this highf flyier at Google.
1:09:58
He ran AI and search and Apple thought they had a coup here.
1:10:03
Apple thought they would hire this guy and really just hit the ground running and be at the forefront of artificial intelligence.
1:10:09
Just seven years earlier they announced Siri in 2011.
1:10:14
There was absolutely nothing like it. It was breakthrough. >> Yeah.
1:10:17
>> But then it just became utter junk. Right.
1:10:19
Google Assistant lapped it. Alexa lapped it.
1:10:21
So they thought they were gonna bring this guy in and it'd be a game changer.
1:10:25
Turns out, and maybe this will be Tim Cook's fade to Comply, but this was the biggest mistake this hire of Tim Cook's tenure.
1:10:33
I think it's easy to say >> Apple is so behind in AI.
1:10:40
>> There's been so much ink spilled on this and so many conversations on this and I've written about and talked about it half a million times.
1:10:46
Uh, I think it does, you haven't even scratched the surface about how big of a problem this is for Apple, >> right?
1:10:53
They've completely screwed up AI in every which way and it comes down to just hiring the wrong people and entrusting the wrong people.
1:11:03
>> But it but is it a do nothing win scenario because I think we're seeing a situation now where the Mac Mini might end up selling out >> boxing is fine or is seeing like financial >> the stock that's the problem, right?
1:11:15
When you've had no real negative hit, Yeah.
1:11:19
it's hard to >> go wartime. >> Go wartime, right?
1:11:23
>> And acknowledge that there is failure because all the numbers are >> It is wartime, right? >> Yeah.
1:11:30
>> The numbers are great.
1:11:30
Tomorrow, >> they might report their first 135 billion to$140 billion quarter, right? Let's go.
1:11:39
>> Like I I don't you know the problem is how can you put $140 billion and Tim Cook needs to go retire in the same sentence, >> right? It doesn't make sense.
1:11:51
But when you think about the long term, you think about the future, >> these things are are going to need to get rectified.
1:11:58
And I guess the good news is they are on a path to rectifying it to some extent.
1:12:03
M >> this Google Gemini deal is a breakthrough for Apple. It's embarrassing.
1:12:09
I mean, it's absolutely crazy that >> you poach their guy >> and then years later you're paying them billions. >> He screws you up.
1:12:17
Okay, you pay him 25 million a year for eight years, right?
1:12:22
And then now you're paying basically 4x that, 8x that, maybe a little bit more in order to get, you know, the new good stuff.
1:12:29
And so they'll announce a new series next month. Mhm.
1:12:32
>> But the thing with this new series is basically a replay of everything they announced in June 2024.
1:12:35
So it's basically everything they announced 2 years ago coming very late.
1:12:39
Things like using what's on your screen to fulfill Siri queries, uh being able to control your apps.
1:12:47
And then in June is when the good stuff launches.
1:12:49
That's when Apple launches its first chatbot.
1:12:52
>> It's interesting because they've spent the last two years saying nobody likes chat bots. Everyone hates chatgbt. It's terrible. It's ruining the world. Okay.
1:12:59
And then you have CHP with nearly a billion, right, active users.
1:13:00
So they're like, "Okay, we kind of got to do this or we're screwed."
1:13:05
>> Uh, so they're doing it. >> Yeah.
1:13:07
>> And it's actually going to run on Google servers, Google Cloud Platform, Google TPUs.
1:13:13
>> This is great for Google, too.
1:13:14
>> And I think some people at Apple think this is going to be like a short-term thing.
1:13:18
You know, we're going to partner with Google till we get our act together.
1:13:22
No, I definitely think that this is more of a long-term play unless these models continue to get commoditized, which they will eventually, but this is not a 12, 24, 36-month thing.
1:13:32
This is this is longer term, I think, than people expect.
1:13:36
>> And Apple can't just go wrap llama because of the terms of service that Meta has put around put that around.
1:13:42
They said you can't >> I don't think they want to use llama.
1:13:44
I don't think they want to work open source models.
1:13:48
So, there there aren't that many games in town.
1:13:49
Well, let's take a step back here.
1:13:52
>> So, they're using Google >> uh in the US, right?
1:13:55
But they're going to have to do something in China, okay?
1:14:00
>> And so, you'll see them use a combination of Alibaba >> uh and and way uh I think it's Waybo or Tencent or one of those >> different AIs for different features.
1:14:11
And so, they'll use them.
1:14:11
And just because they're partnering with Google Gemini on on Siri in this chatbot >> doesn't mean they're not using other players.
1:14:19
There's a lot of open AI and a lot of applications.
1:14:21
They launched their creative cloud competitor uh today and a lot of those AI features are powered by chat GPT like some image generation stuff. >> Okay.
1:14:31
>> I still think on an image generation side >> you're getting a little bit better on OpenAI than you're getting from Google.
1:14:37
>> Um and then a lot of stuff internally like Apple runs on Enthropic at this point.
1:14:43
Enthropic is powering a lot of the stuff Apple's doing internally in terms of product development.
1:14:48
a lot of their internal tools.
1:14:49
So, that's a big one to watch.
1:14:51
Um, they have custom versions of Claude running on um their own servers internally, too, because this Google deal, this just came together a few months ago.
1:14:59
They were not going to use Google.
1:15:02
Apple actually was going to rebuild Siri around Claude.
1:15:06
>> Uh, but Enthropic, they were holding them over a barrel.
1:15:08
They wanted a crap ton of money from them, several billion dollars a year, and at a price that doubled on an annual basis as well for the next three years or so from what I understand.
1:15:21
>> At the time, Google was really an afterthought because they were in the middle of the trial with the Department of Justice.
1:15:27
Um, and then for some reason, the judge ruled that Apple and Google's deal was kosher, even though everyone knows it's a huge issue and a monopoly and all that.
1:15:36
I'm not here to be a judge. which I I write.
1:15:38
I don't make uh judgments for for for legal proceedings, but anyways, Apple and Google get off scot-free.
1:15:44
They can do whatever they want now, and you know, >> they're not being held back at all.
1:15:50
>> Uh and then obviously Open AI, that is a real firestorm for Apple right now.
1:15:53
Like OpenAI, obviously, they're working on these AirPods competitors that Chad GPT built, Sweet Pe is, you know, running the show and design there.
1:16:05
>> Uh that's a big freaking deal. >> Okay.
1:16:08
He raided the whole design team.
1:16:10
All of the Apple designers that were there under Johnny IV, not all of them, 95% of them are gone.
1:16:14
You know, a ton of them are now working at Love from and OpenAI and other companies.
1:16:18
Some of them have retired.
1:16:21
>> Uh but the crux of it is like how do you partner with a company that's trying to put you out of business, right?
1:16:25
So there's no way that they were going to work with with open AI uh no matter how good the >> What do you think about some of uh the projections that you've seen uh or or kind of estimates around uh I forget uh was it Foxcon was saying >> yeah preparing to be able to produce something like that in the first >> year it feels like a massive number but at the same time you have a billion users you know talking about the open AI earbooks. Uh, sweet.
1:16:56
>> You know what's so interesting >> and and one more thing there. >> Yeah.
1:17:00
>> The ear the the earbuds the benefit of them versus some of these other AI hardware is like if you can just take calls and listen to music like already you have some functionality that people are using all day long.
1:17:11
So it's not the same as like wearing a pendant or having some other >> uh little gadget that like actually serves no real use case or at least not a powerful use case.
1:17:20
if I can just listen to calls or sorry take calls, listen to music, at least there's like some like base level functionality and then you layer on the AI and maybe it turns into this amazing, you know, super differentiated experience.
1:17:34
>> I just don't see it >> I just don't see them selling 45 million units.
1:17:37
I just don't see it being a success.
1:17:39
The barrier to entry for a new hardware company, as you know, is extremely high.
1:17:42
Like, can you even think of one hardware company that just came into being and became an immediate success?
1:17:48
Like, I just don't see it.
1:17:48
And I also think the reason they're doing earbuds is because it's more lowhanging fruit to your point and it's easier to accomplish.
1:17:57
Isn't that what they wanted to do?
1:17:59
What OpenAI wanted to do is they wanted to create an iPhone killer.
1:18:02
Instead, now they're trying to create an AirPods killer.
1:18:04
And I don't think they're going to be able to accomplish that.
1:18:06
Will they look nicer than the AirPods? Probably.
1:18:08
Will they have better AI than the AirPods?
1:18:10
That's an open question, right?
1:18:12
What stops Apple from expanding this Gemini deal to its AirPods and just basically adding a bunch of AI functionality to the AirPods?
1:18:20
They have very fast processors, very tight iPhone integration.
1:18:22
Apple can replicate whatever OpenAI is going to do pretty quickly um by relying on Google and Gemini and whatever they can cook up functionalitywise there.
1:18:33
Do you think that AI specifically products uh you know that we had the creator of uh Maltbot on yesterday and he was saying like there's so many applications and he was referencing like My Fitness Pal.
1:18:45
Do I really need My Fitness Pal if I can just take a picture and just send it in to my agent and it'll track it.
1:18:50
And so what I was thinking there is like does that [clears throat] make the app store broadly potentially less of a hurdle with OpenAI is launching eventually launches a new phone and they have a new OS and they don't have an app store.
1:19:07
They don't have an app store on day one.
1:19:08
>> The equivalence of apps on the fly.
1:19:10
>> Well, if you ask me, you know, we're already in territory where iOS and the app store uh are legacy features, >> right?
1:19:18
Like the app store is a legacy world.
1:19:20
the iOS user experience, the Mac OS user experience where you're jumping between applications, uh where you're uh going into something to get information, uh where you're going back to your home screen to launch another app, apps are the past.
1:19:33
AI agents uh are already here and that's the move forward.
1:19:38
And whenever OpenAI comes out with a phone, and I do eventually anticipate them coming out with a phone, and when I say phone, I'm not talking necessarily about something you put to your ear like a classic phone, like an iPhone.
1:19:49
I'm talking about I feel like people are always going to have some sort of slab in their pocket, >> right? Because it's convenient.
1:19:55
You get the display, you get the sensors, you get the battery, you get the cameras.
1:20:01
It's not a replicable experience no matter how many different gadgets put on your body.
1:20:04
>> Make a like a banana, like a plastic banana, smart banana.
1:20:07
Yeah, [laughter] that would be fun. Yeah.
1:20:09
So, yeah, AI agents are where the the world is going and I totally expect Apple to move in this direction.
1:20:17
This new Siri Campo that they're launching at the end of this year, >> uh that is a huge step towards the AI agentification of uh different features on the phone >> and being able to for instance tell your phone uh pull up this photo that I I I I took at the studio.
1:20:36
uh find photos where I have bottles of water and remove the bottle water from the from the from the photo and email her to text it to Mark, right?
1:20:47
Like the AI agentification of iOS is happening. >> Sure. >> Yeah.
1:20:51
And theoretically, Siri should have killed like the weather app a long time ago because I don't like when you think about like if you're just going to navigate to the weather app and you're traveling somewhere and you're like, "Oh, where is this city?
1:21:02
Oh, I don't have it saved.
1:21:04
I'm going to search and add it."
1:21:06
and then you're like finally looking at it and it should just be a prompt but then when you try and click a level deeper when in the current Siri experience you just can't get that.
1:21:16
>> That is the big problem with Siri and the big difference [clears throat] from chat GPT and Gemini is it has the going a level deeper problem.
1:21:21
There's no back and forth.
1:21:23
It forgets context very easily. It has no memory.
1:21:25
Uh it doesn't have that tight integration with applications because the app developers they know Siri is so terrible.
1:21:33
You know here's a question.
1:21:35
So, what if I told you that you could call an Uber with Siri, right?
1:21:39
Like, you guys probably know that, but like, does anyone use it?
1:21:42
I don't have the data, but I can tell you that they announced uh support for that and they actually rolled out support. No, 10 years ago.
1:21:49
>> 10 years [laughter] ago. >> Okay. It doesn't work.
1:21:50
And no, because nobody uses it because it's too cumbersome. Totally. Okay.
1:21:55
Like, I'm very tech forward as anyone watching this probably knows.
1:21:58
You know, it's right up my alley to have Siri call my Uber for me, but like I've never done it because you know what?
1:22:05
I don't freaking trust it. >> Trust it. Yep.
1:22:08
>> It's not I don't think it's going to work and it's just going to be a time suck.
1:22:11
>> And people didn't trust chat GPT with 3. 5, but now on 5.
1:22:14
2 Pro, >> I would trust Chat GPT with ordering.
1:22:17
I would trust CG GPT with my life. [laughter] >> Okay.
1:22:23
But Siri, >> okay, so on the Siri roll out, um it seems like there's a couple milestones that we're going to be tracking throughout this year.
1:22:30
When will when do you think I'll be able to go to Siri and just say a general Google question, you know, uh give me the history of the Roman Empire.
1:22:41
The type of thing that any LLM can just pull up a couple paragraphs on, but current Siri does not have that ability.
1:22:49
>> Okay, I can't tell you that you're going to be able to do it.
1:22:51
I can tell you that you're supposed to be able to do it. How's that? I'll speak I'll speak.
1:22:56
>> I just mean I just mean is that like what's the goal? >> No, no.
1:22:58
I just mean like there's knowledge retrieval end of March >> and then and then there's and then there's like agentification like making interacting with the apps doing all the things like pulling what's in your camera roll together with your iMessage and all that.
1:23:10
And that feels harder for both of those. End of March.
1:23:12
They're supposed to be end of March, >> but the latter in terms of the app manipulation and the AI agent, that's going to be split probably between >> because that seems harder than just integrate Gemini and just give people the Gemini answers when they when they ask Siri for a question.
1:23:28
>> You know, it's so funny.
1:23:28
You I look back to WWC 2024 and obviously I made a big deal about all the delays because of course I did.
1:23:35
>> But then you think about it, it's like >> was anyone going to use these any of these features anyways?
1:23:39
Like we're talking about a few features that were delayed and you look back at the announcement of these features at WWDC 2024.
1:23:46
It's like Apple totally downplayed them too.
1:23:49
Like they gave a few cool demos, but it's like this is some gamechanging stuff if it's marketed correctly. Yeah.
1:23:55
So, not only did they pre-announce it, not only did they delay it, but like they didn't even market what they had in the hand correctly, which I guess in hindsight, like you should have realized that it wasn't that compelling or or ready to go if they weren't going to market it.
1:24:09
Cuz I'm telling you, they can market anything.
1:24:11
>> How does how do you see search >> fitting into Siri?
1:24:14
Because so much of what people are using LMS for is search now, and you're seeing >> more commerce integrations.
1:24:22
And I've always thought the lab's like obsession with the browser was interesting because in many ways like I feel like using LLMs.
1:24:32
It already feels like you're using a browser.
1:24:35
It's just kind of a new experience.
1:24:37
And and a big question is like okay so Google's powering the new Siri.
1:24:40
What happens when people start doing searches with high intent in Siri?
1:24:47
>> Do you Google search anymore? >> I do. >> I don't.
1:24:49
I I specifically still Google search when doing like product research, shopping, etc.
1:24:56
>> I don't do you >> I I still go to Google if it's something that I know is fastest in Google.
1:24:58
So today I wanted to know what year the Avengers movie came out and I know that that's half a millisecond in Google and I know that that's 5 seconds in any LLM and so I'll control T to go there.
1:25:12
But for any level deeper, I also wanted to know about the VFX house that worked on it and what technology they used and when that happened. That was an LM query. >> Yeah. Um, yeah.
1:25:24
I don't really use Google anymore.
1:25:27
I have my action button on my phone sent to Chad GPT and I'm just like living in Chad GPT. >> Sure. Sure. Sure.
1:25:32
>> You can't ask it the weather. >> Yeah.
1:25:34
>> Can't ask it to do anything on your phone. >> Sure. Sure.
1:25:36
>> That's I guess the the differentiator for what the new series is going to be.
1:25:40
>> Um, they've built a feature called World Knowledge Answers.
1:25:42
the internal name and it's basically a perplexity ripoff or a web search ripoff in Jad GPT which is to to bullet out summaries and information and give you citations and context.
1:25:54
>> Is it using Google directly for that or is it some other?
1:25:58
>> It's an Apple solution.
1:25:58
Okay, it's an Apple solution.
1:26:00
Um but the the what they've been trying to do now is under the hood change that to the Gemini model.
1:26:08
Again, all this stuff was supposed to come out >> cuz Gemini is very good at Google search.
1:26:13
>> Gemini is very good at Google search, but there's no Google services in this new Siri. It's literally a model.
1:26:16
It's basically like they hired the Google DeepMind team to develop the model to power its AI in Siri. >> Yeah.
1:26:27
So, I guess what I'm getting at is is will I ever get if I'm doing using Siri to research a product, >> will I ever get an ad powered by like a Apple ad network because it is a high intent search >> like ever.
1:26:42
>> I think the endgame for all these guys is to put ads in this stuff. >> Yeah.
1:26:46
>> You know, one thing that I haven't talked about in a while >> and and I just feel like that's been >> somewhat under discussed because people are so obsessed with just make Siri work. >> Yeah.
1:26:56
Yeah, >> that if it does end up working really well, then you would have a effectively a search engine.
1:27:03
>> If it well, yeah, if it works, they have all the tools to do a search engine, and if it does work well, um, yeah, they'll probably eventually put ads in it.
1:27:09
A couple things haven't discussed in a while.
1:27:12
One is they're, you know, working on a new Safari web browser, and you're going to see AI search the forefront of that.
1:27:18
Uh the other thing is the adsification, the advertising of uh of the Apple operating systems that starts this year.
1:27:30
>> You're going to see them up the ad slots in the app store in search in search in particular.
1:27:36
Uh you're >> give it up for more ads. >> Yeah, ads.
1:27:40
[laughter] Any sponsors we want to give a shout out to over here?
1:27:44
Um, and then >> don't temp.
1:27:49
>> Um, and then you're going to see ads in Apple Maps. >> Oh, interesting.
1:27:52
You know, that's been one of the main differentiators in the different apps.
1:27:58
>> You know, it'll be all AI and it'll be targeted to based on things that you're searching.
1:28:02
Like if you see a sushi restaurant, search for sushi, whatever, you may see some search results get elevated.
1:28:08
It's kind of like what Yelp does. Yeah. Right.
1:28:09
By the way, it's so funny on on on Siri like um there's this uh Japanese restaurant in the valley I like and it's named after a city in Japan or province in Japan.
1:28:18
And it's like take me to Chiba. >> Yeah. >> Right.
1:28:23
And I've been that's the name of the place. Shout out to Chiba.
1:28:26
>> Okay boss, you're going to need to get in a boat. >> Oh my god.
1:28:28
It's like immediately pops up 7,652 miles away.
1:28:33
I'm like I've been to this place 50 damn times. You should know by now.
1:28:38
talking about the restaurant 20 miles away. Yep.
1:28:42
>> Not the city [laughter] 8,000 miles away. >> That's crazy. >> It's really amazing.
1:28:46
>> Hiring chartering an aircraft, sir. >> Yeah.
1:28:48
[laughter] I I I got a bunch of other stuff in the Apple ecosystem.
1:28:50
Um first, uh Starlink is reportedly planning to integrate uh Apple's reportedly planning to integrate Starlink connectivity into the iPhone 18 Pro.
1:29:01
>> I saw this you saw this stuffing on Twitter the last couple days.
1:29:03
There's been no reporting on this recently, by the way.
1:29:06
This is another See, this is one of the downsides to AI.
1:29:07
People just spew BS on Twitter. >> Um, okay.
1:29:12
So, Apple already does work with Starlink, right?
1:29:14
You can hook in if you have a T-Mobile iPhone, you can get on to to Starlink. So, that already exists. >> Yeah. >> Yeah.
1:29:23
I have >> I've experienced I experienced I think during the fires last year.
1:29:26
No, all the cell service was down.
1:29:29
If you want if you want satellite satellite on the 17, but it's not what >> you don't want to be on Verizon.
1:29:35
I I should get off of it, but uh but it's not it's not uh Starlink.
1:29:38
It's the old network >> on which on the Verizon one you're talking about on the the Apple network. >> Yeah. Yeah.
1:29:44
Like like if I'm in an emergency situation, I can get out like one text message.
1:29:46
It's really really slow and it's clearly going with like I think via >> the Apple network is just super legacy. Yeah. It's just older.
1:29:53
>> They've got to >> they've got to strip that down and partner and rebuild or something.
1:29:57
Anyways, the iPhone 18 is going to have enhanced satellite connectivity.
1:29:59
Uh, and they'll, you know, obviously work with Starlink and, you know, whoever, whomever. So, that's coming.
1:30:05
What else we got on our plate?
1:30:07
You want to talk about John Turnis? >> Absolutely.
1:30:10
>> Let's talk about Turn >> Turn us around, John. >> Yes, >> John. Okay. >> Yes.
1:30:14
I So, I have a bunch of questions.
1:30:16
Let's start with like uh how like what is going on behind the scenes of this campaign?
1:30:21
It feels like there's potent it feels like it feels like internally politics there could be like people like pushing for him, pushing against him.
1:30:29
There's this weird quote that keeps going out that people say he's never made a decision in his life.
1:30:32
>> Everybody keeps making the most underhanded compliments that I've ever seen. >> Yeah.
1:30:37
>> It's like >> it's like Yeah, he's shipped a lot of products, but >> he's never had to make a hard decision. That's the one.
1:30:45
>> Just wait for my profile in Bloomberg. Okay.
1:30:46
You'll get the real story soon.
1:30:48
Not going to give it all away today.
1:30:50
>> Of course, >> but I'm here. Yes. >> For the company.
1:30:53
>> Got to give the shout out.
1:30:53
Stay tuned for the article, >> of course.
1:30:55
Uh, subscribe to Power On. Yes. In our tech bundle. Yeah. Uh, it's great value. >> Okay. Turn us. He's 50.
1:31:04
>> Everyone else on the Apple executive team, late 50s through their mid60s, turning 66 this year in the case of Tim Cook. >> You're Apple's board. You like continuity. Uh, you like an insider.
1:31:12
You like people who know what they're doing, have been there for a while.
1:31:16
They know where the bodies are buried.
1:31:20
>> Okay, these guys are all have hundreds of millions of dollars, if not more.
1:31:25
Uh yeah, at 50 he's the only one who is if let's say Tim Cook hangs out another three to five years, >> you're not going to point another CEO who's 65, 70 years old. Yeah. >> He's the only guy.
1:31:40
>> Apple, they get vast majority of the revenue from hardware. He's the hardware guy. >> Yep.
1:31:45
>> Have they screwed up any hardware since he's been in charge? >> No. He's a steady hand.
1:31:48
He knows what he's doing.
1:31:50
He's really the only choice.
1:31:53
You know, there was this New York Times report a few weeks ago basically saying that could be Greg Jazuak, could be Eddie Q, could be Dearra O'Brien, could be Craig Federigi.
1:32:03
>> It's for sure not going to be Craig.
1:32:06
>> It's not going to be Deardra.
1:32:06
It's not going to be Eddie.
1:32:07
It's not going to be Jaws.
1:32:08
The only category that makes sense is an operations person.
1:32:11
Because you look at the current CE CEO, Tim obviously comes out of the ops world.
1:32:17
You look at the guy who would have been CEO if Tim Cook didn't stay so long.
1:32:19
I'm not saying he shouldn't have stayed so long.
1:32:23
He's done obviously a fantastic job for shareholders and the employees and what have you.
1:32:26
Would have been Jeff Williams. He was the COO.
1:32:30
>> Uh so Sabi Khan, he was named COO, you know, a few months ago, but he's really been in that job for the last half a decade, I would say.
1:32:37
>> Uh so anyways, it'll be Turnis or Sabi or or someone completely out of left field.
1:32:41
I don't think this is imminent.
1:32:43
Um so we'll see what ultimately happens, but all signs are turning towards Turnis. Mhm.
1:32:49
>> Everyone has an opinion that Turnis is going to be the next CEO. Fine.
1:32:51
I've been shouting this from rooftops for the last two years.
1:32:56
But no one has given evidence like what is this based on? Right.
1:33:00
[clears throat] Has there ever been a baton handoff?
1:33:01
Is he getting more responsibility?
1:33:03
Well, >> they have a like a big baton.
1:33:06
>> They got one of these. You know what you have?
1:33:08
You have white smoke coming out of Well, actually, no.
1:33:10
There's no there's no smoke.
1:33:13
>> It's very environmentally friendly. friendly.
1:33:15
>> Um, so you know, maybe out of the >> reflection out of the sun solar panels or something, a glint off the solar panels. >> Exactly.
1:33:21
Uh, so what you want evidence, you want to hear that he's been getting more responsibility. Okay.
1:33:28
>> One, a few months ago took full control of the Apple Watch engineering team >> that was co-run with the old COO till he retired.
1:33:37
Okay, there's something for you.
1:33:39
uh when they started uh softfiring the head of AI, the guy we were talking about earlier, >> they took the robotic stuff away from him.
1:33:50
They gave that to Turtis.
1:33:53
>> And then the real news, as I broke last week, is that even though on paper uh Tim Cook is running the Apple design teams, it's not. It's Turtis.
1:34:00
He took over at the end of last year >> uh for a variety of reasons, but you look at who's run design at Apple over the course of history.
1:34:08
Well, like Steve Jobs, Tim Cook himself between 15 and 17, Jeff Williams who was the number two in a parent for a long time and then obviously all heard of Johnny IV. Yep.
1:34:18
>> Uh now you add John Turnis to that list. It's it's a big sign.
1:34:21
It's a big indicator because >> you look at what Apple is known for as a company and its design, right?
1:34:27
That design function and it's not just hardware, it's hardware and software that he's, you know, overseeing as the manager of both of those teams.
1:34:34
So, >> I would say that is your your first piece of evidence that he's getting some more material.
1:34:41
>> What does he have to do since he's the hardware guy to communicate that he has a steady hand on the tiller as Apple transitions into the AI age?
1:34:50
Like the narrative is that they have delivered on hardware.
1:34:53
They have not delivered on AI. No, they haven't.
1:34:57
>> And so, how is he going to communicate that he's forward thinking and can be the one to keep them on the cut? >> Taking a step back.
1:35:03
you think about like succession at all of these um big tech companies, right?
1:35:07
Like who's going to who's going to take over Google one day?
1:35:11
Who's going to take over Microsoft one day?
1:35:12
Who's going to take over Amazon one day, right?
1:35:14
>> Like it's probably going to be the AI guys at all these companies, right?
1:35:18
Apple doesn't really have an AI guy.
1:35:21
They're trying to make Craig Federigi the AI guy, but he's not CEO material.
1:35:25
He's he's voiced this himself, so don't get mad at me, Craig.
1:35:29
>> Uh but >> yeah, Turnis, can he become the AI guy at Apple?
1:35:33
I I think it's too early to tell and I think that is a that is a big question there like do you put a hardware person in charge of the AI era?
1:35:41
>> People have been speculating that Apple would do some big M&A deal to bring some AI native talent in house.
1:35:50
>> They wanted to buy Perplexity. >> Yeah.
1:35:52
>> So the origin story >> they were really talking about it like Eddie Q was seriously considering Adrien Pri other head of corp development M&A reports to Tim Cook.
1:36:02
They were looking at this pretty closely uh than they pulled out.
1:36:05
Why were they going to buy Perplexity to power the search stuff we were talking about earlier became less important after the Google uh deal was allowed to live.
1:36:16
>> And do you think they would have done it if Perplexity hadn't been marked up to oblivion?
1:36:23
>> Uh I think they would have done Perplexity at a reasonable price. Yeah, that's the thing. Low single digit.
1:36:29
I think they would have done it over up to five or six billion. Yeah. >> Yeah.
1:36:33
But like at this 1520 >> and then Perplexity started, you know, trying to say they're going to buy Google Chrome if it got divested and all that.
1:36:42
So, >> uh, and I think they would have done it if the Google search deal was torn apart. >> Mhm. >> Yeah.
1:36:48
>> In order to bring a search product to market faster, they like to buy companies to they I I'm not trying to do Apple corporate speak here, but this is like legit.
1:36:55
They buy companies to accelerate their roadmap.
1:36:57
You'll hear Tim Cook use those exact words tomorrow.
1:37:01
Okay, by the way, but literally [laughter] >> tomorrow. >> Apple earnings. >> This guy. >> Apple earnings. >> This guy. >> My god. >> Come on. Hey, head in the game.
1:37:10
>> But but is is there specifically a deal that he'll be talking about tomorrow? Is that what you think? >> No. No. Just in general. Like he will be asked.
1:37:17
>> I bet he'll mention the Gemini >> because he was asked that in the last earnings and and everyone was like, "Why haven't you done a mega deal?
1:37:22
Why haven't you done deals?" Okay.
1:37:23
So, he just says it's just going to be more the same.
1:37:25
Got You know what it is fair to say is their pace of deal making has decelerated significantly. >> Interesting. >> Significantly.
1:37:33
>> And so so he will be I mean in the last earnings he was sort of like managing that and saying like oh we're still doing deals but we're selective and we only do it to >> well they also don't care to announce like 95% of >> they're big enough they have to you know like the Golden State Warriors they want Giannis but they don't want to make a big trade.
1:37:48
They don't want to give up all their draft picks right they're like yeah this the same with the Lakers.
1:37:51
This is like my existential crisis.
1:37:53
The Lakers like, "Yeah, we'll trade the picks >> uh if the right guy becomes available." Yeah.
1:37:59
And they know the right guy's not going to become available, right?
1:38:01
Just like Apple knows these companies are always going to be out of the price range they want to pay. >> Sure. Sure. Yeah.
1:38:07
>> Thing with Apple is like they're very frugal moneywise. Extremely frugal.
1:38:09
The other thing is is they've been burned countless times with acquisitions.
1:38:13
Like that Beats deal, >> terrible process for integrating that company. >> Sure.
1:38:20
>> Sure. uh in >> well and even then I think it was like a 3x revenue multiple weren't they doing wasn't beats doing like a billion dollar >> oh from a financial standpoint it was just like a home run people criticized that deal but they made it back in months >> oh interesting made the money but it was
1:38:35
a nightmare from a but it's worth it's worth noting because like even paying for perplexity >> at 56 billion >> it's monetized that probably >> yeah I mean Beats was monetized from day one obviously you're selling the headphones which are terrible, by the way, but everyone loves them. And then
1:38:51
And then Apple Music, they, you know, >> Beats started the whole subscription services business at Apple.
1:38:57
And so, if you look at it, Beats is one of the most wildly successful technology acquisitions of all time.
1:39:04
And then you compare it to how much criticism it got because it's Dr.
1:39:07
Dre and Jimmy Ivy and whatever, whatever, whatever, and the headphones are crap. >> You know what?
1:39:12
From a financial standpoint, home run. >> Home run. >> Yeah.
1:39:17
you know, uh, but integrating into Apple's culture is not easy. >> Yeah.
1:39:22
Uh, speaking of hardware, what what what's going on with the robotic arm that will live in your kitchen? >> Yeah, the Pixar lamp. >> It's the Pixar lamp.
1:39:29
As you said, it's uh you got this 9 in display.
1:39:31
It's like an iPad display on a robotic arm.
1:39:33
It can float around your desk.
1:39:35
It can twirl and turn around.
1:39:38
Like, >> you know what we'll do? >> Mhm.
1:39:41
In a few years when this thing comes out, you'll have me back on and instead of me actually being here, you'll have this on here and maybe, you know, it'll be my head floating around. >> Yeah.
1:39:49
I mean, Meta tried to do something like that with the >> Meta tried to do something like that.
1:39:52
You know, these things are big in China. They are.
1:39:54
>> It's a big thing in China right now. >> Interesting.
1:39:56
>> Uh, no one really talks about them, but it's uh it's a category that has some potential. >> Okay. Yeah.
1:40:01
But it's years away just because >> I think it's uh stay tuned. >> Yeah.
1:40:06
What about the foldable phone?
1:40:07
>> That's not years away. >> That's that's closer. >> Yeah. No, that's decades away.
1:40:09
No, no, that's [laughter] uh that's coming out in the fall.
1:40:13
>> Okay, that'll be fine.
1:40:14
>> I can't wait for that.
1:40:15
>> Yeah, that'll be Are you going to be buy $2,200 on it now?
1:40:18
>> Oh, is that going to be Is it >> It'll be at least It has to be >> Is it going to be the highest tier, the biggest, the most powerful >> for Apple?
1:40:24
Yeah, that's going to it's going to sit at the top of the lineup.
1:40:27
>> It's going to sit at the top of the lineup. Okay. New status symbol. >> Sorry. What were you like that?
1:40:30
>> Uh we had Ben Thompson on maybe before the end of the year. >> Feel bad for Ben.
1:40:34
You know, he's a big Milwaukee Bucks fan and they're about to >> ditch you on us. >> The Milwaukee Bucks. That's sports team. >> Really? >> I don't know.
1:40:42
[laughter] I do know it's a basketball team and I do know that it's Ben Thompson's favorite team.
1:40:47
Uh he was talking about the >> Yesterday we didn't know what the when the Super Bowl was.
1:40:51
>> He he he was talking about the Apple Vision Pro.
1:40:52
Uh I wanted to demo the the you know, you can watch the NBA live. You liked it.
1:40:59
>> Pitch was like screw the >> screw the the Apple produc highly produced.
1:41:04
You know, they're cutting around all the time.
1:41:06
And he's like just invest the money to set up like the actual hardware in every single stadium >> cameras >> the cameras so that anybody can just drop in.
1:41:17
>> So you can just sit there watch the game like you're need you don't need the the announcer because you can just look at the scoreboard you can listen what's going on.
1:41:25
>> So he's saying strip the guys out of it.
1:41:27
I mean >> look I watched it. Yep. It was great.
1:41:30
>> You watched the whole thing? >> Yeah. >> Really? >> Yeah. Yeah.
1:41:32
Trust me people at home were not happy that night.
1:41:34
you know, you put that thing on crying baby crying baby and then you're completely isolated.
1:41:38
I mean, I'm telling you, VR does not work for for families. >> It's true. Yeah.
1:41:45
>> It just doesn't >> It's really Yeah, it is one of the major cuz then I mean, the family people have the disposable income more likely.
1:41:49
So, then they buy it, but they can't use it. >> Doesn't work. Yeah. >> Uh, okay. Here's the problem. Yeah.
1:41:55
>> It's like once a month.
1:41:55
They've got like five games on the calendar. Cool.
1:41:57
And that and that was Ben's point is like just set up the infrastructure and sell me a pass so that I can drop into any game and sit courtside >> because right now they have to pull up with a production truck. They need editors. They need voiceovers.
1:42:08
They need pre there's you know basically one time fixed cost of like he he did the math all the Yeah.
1:42:16
He said it's like 40 grand.
1:42:18
It's like not like a huge amount of money >> and you have no incremental costs per show.
1:42:23
Oh, you're not dealing with producers and running a live, you know, product.
1:42:28
>> But they need to decide if this hardware is going to continue to exist before they keep investing in the content strategy, right?
1:42:33
It's a chicken and the egg problem, >> right?
1:42:36
How do you sell this thing if there's no content?
1:42:39
>> But why invest millions in the content if you're not planning to sell this thing anyways and you're pivoting to smart glasses?
1:42:45
>> Don't forget, they were supposed to come out with the Visiona Air in 27 product called N100.
1:42:50
They uh I forgot when my article came out. It's times a blur.
1:42:52
A few months ago, six months ago.
1:42:54
Anyways, they killed that thing.
1:42:57
>> They killed it entirely. >> Killed it.
1:42:59
>> And the smart glasses >> shocked the vision team on that.
1:43:00
By the way, >> the smart glasses that was that just completely reactionary to to Meta. >> Oh yeah.
1:43:09
They started toying with the smart glasses in terms of like this non AR smart glasses, right?
1:43:13
First of all, AR smart glasses that has been the vision from day one for a decade plus.
1:43:20
But in terms of like this non-dis smart glasses, that is a concept that Meta has really popularized and you know when these things started to gain a little bit of steam in 22 23 is when they started taking a very hard look at it and they're going to do it >> and I think they're going to destroy Meta with them. I'll be honest. >> Yeah.
1:43:36
Is it the styling, the pricing, the features, the integrations? integration.
1:43:40
To me, to me, to me, the challenge is >> if you can't deliver me iMessage. >> Yeah.
1:43:50
>> Apple has the ingredients >> because of their login to destroy any company in any hardware.
1:43:54
It's just about them figuring out how to do it and get it done and not waiting too long.
1:43:59
You know, the biggest problem with them is they just take too long and overengineer everything. Yeah.
1:44:05
>> Like the Vision Pro is the most overengineered device ever, right?
1:44:06
They could have got that out 3 years earlier with a little bit less fit and finish and maybe it would be more successful today. >> Yeah. Yeah.
1:44:14
It does feel like there was a cycle where the iPhone was ahead of the curve on so many things touchscreen and then you and then in the like that middle decade period.
1:44:25
You had the Android folks being like, "Oh, we've had this Apple feature for a year. We've had this."
1:44:30
Now it's like five years.
1:44:33
>> There's, you know, many people wouldn't be caught dead with a non iPhone, you know, like it doesn't matter.
1:44:36
But AI changes that equation. >> Sure.
1:44:40
>> That changes the equation.
1:44:41
>> Especially if it's a Johnny IV product and it's expensive.
1:44:44
>> Well, you know, Johnny IV obviously he's done amazing things.
1:44:47
>> Um, and the new headphones or whatever they come out with are going to look amazing and probably work amazing.
1:44:51
It's just the barrier to entry on hardware is so high.
1:44:55
There's so much risk there. >> Yeah.
1:44:58
I I I just I still wonder if the uh like the Claudebot, these like open source agents, I don't know that everyone's going to adopt those, but something like that that sort of opens up the ecosystem just by brute forcing it.
1:45:09
We were debating this earlier like uh you can't get iMessage notifications on the meta rated band displays, which is >> I have them. Do you guys have them?
1:45:16
>> Uh I think we do have >> I've got one at home.
1:45:19
We've used the the the the non-displays a bunch and then we demoed the displays and have used them a fair amount and it's it's just a hard sell if you're an iMessage user.
1:45:28
>> A hard sell if you're an iMessage user, but the potential is just oozing with potential.
1:45:32
Like they've got a they've got a solid product there. Yeah.
1:45:36
>> And like a couple iterations on that, make them a little lighter, get the display resolution up, work better outdoors. >> Yeah. >> It's compelling.
1:45:43
And and it and in a world where you have some sort of agent running on your Mac Mini, scraping all your IME messages and then putting them into WhatsApp or something, >> you know, maybe I'll do that.
1:45:54
Maybe you guys will do that. >> It's rare. Yeah, >> it's rare.
1:45:57
Nobody wants to deal with that. Yeah.
1:45:59
You know, nobody wants to deal with that.
1:46:01
>> Even if it's just an app that you download like Napster. You don't think so?
1:46:06
>> People don't have time for that. >> Probably not. Probably not. >> What?
1:46:09
Uh >> I feel like nobody cares about anything anymore, [laughter] >> you know?
1:46:12
I just feel like back in my day, we used to care.
1:46:14
The >> world has changed. >> Yeah.
1:46:16
You know, now the >> world has changed.
1:46:17
They just want everything in front of them.
1:46:18
They want everything in their eyeballs.
1:46:20
They all want it set up from the get- go.
1:46:21
They don't want to put any work in.
1:46:22
They just want it to start working, right?
1:46:23
And I think that's been the Apple ethos from the beginning, which is just like give people what they need.
1:46:28
Let it get up and running and not deal with any of the BS.
1:46:32
>> What's going on in China?
1:46:34
>> Lots going on in China >> with Apple.
1:46:35
Are they shutting down more stores?
1:46:37
>> Shutting down more stores? No, not that I know of.
1:46:39
I don't see Apple just shutting stores at this point.
1:46:42
I think the retail arm still is extremely profitable and and successful. >> Sure. >> All the shutdowns.
1:46:50
>> But in China, there's less of the you're not somebody's not embarrassed to not be using an iPhone. >> Problem? Yes.
1:46:57
The problem is is that Apple hasn't done anything bespoke for the Chinese market.
1:47:03
M >> the competitiveness there is just unbelievably just it's amazing.
1:47:08
It's like no other part of the world.
1:47:10
And >> are there are there compelling AI hardware? >> Yes.
1:47:17
I mean I mentioned the robotic thing.
1:47:18
I mentioned um well like the foldables are taking over the universe there, right?
1:47:24
And so you know Apple's an American company launching American devices, European devices, and they're trying to shoehorn it into the Chinese market.
1:47:31
And they've never really done anything just for the Chinese market or built around the Chinese market.
1:47:35
Um, >> what about the sock? >> iPod.
1:47:39
Okay, maybe that was I mean I don't know. >> But why?
1:47:42
It's such a massive It's such a massive market. Why not?
1:47:47
>> Because they're a global company. >> Mhm.
1:47:49
>> But I think the foldable is going to do extremely well in China.
1:47:51
You know, it might do better in China than it does here. >> Interesting. >> I don't know. I'll be having one. Yeah.
1:47:57
>> I'll tell you that much right now.
1:47:58
>> What's the future of the iPhone Air?
1:48:02
It's like the price difference. >> Sam Alman's got one. >> Great. Yeah.
1:48:06
[laughter] The the >> But he doesn't care about money. >> The price difference.
1:48:11
>> He just breaks it and buys a new one.
1:48:13
>> He doesn't get paid by OpenAI.
1:48:14
>> He's probably got both. >> Yeah. >> Yeah. He doesn't get paid. Yeah.
1:48:18
>> Um the iPhone Air and the iPhone Pro, it's like negligible pricing wise.
1:48:21
It's the same price if you get the battery pack. >> Oh yeah. >> Right. It's like 99. 99 plus 100 plus. Yeah. Right.
1:48:28
Buy you're going to get the battery pack.
1:48:29
Um, so I don't know, like you look at the features comparison, you look at the cameras.
1:48:35
I mean, most people are going to always pick the Pro uh over the Air.
1:48:39
There needs to be more of a price gap. >> Yep. >> Between the two.
1:48:43
And eventually, you know, those two lines are going to merge.
1:48:47
Could take five years, but like eventually you're going to be able to get a Pro >> as thin as an Air or an Air with the same bells and whistles as a Pro.
1:48:54
I mean, you look at the MacBook Pro and the MacBook Air today from a performance >> these are different laptops and they're exactly the same. >> Yeah.
1:49:02
So, what you've got the Air Pro >> and they're identical.
1:49:06
>> It's the same thing because the chips, right?
1:49:08
The big differences between the two uh are it's a little lighter. >> Mhm. >> Okay.
1:49:13
The display is terrible on that thing compared to that thing.
1:49:15
Like, if you use like I tried out the 15-inch MacBook Air, okay, >> the thing is sleek and slick and awesome.
1:49:23
But like I've been ruined visually by how amazing the display is on the MacBook Pro.
1:49:28
>> Okay, maybe I got to upgrade.
1:49:29
>> And so, you know, you have to think about the different changes.
1:49:33
>> You want to talk about big things happening at Apple this year.
1:49:34
It's that new MacBook Pro. Yeah. >> Right.
1:49:37
You got the OLED, you got the thinner, you got the touch.
1:49:40
>> Cannot wait to drop $4,000 on that thing.
1:49:42
And >> it's going to be a touch screen. >> Yeah.
1:49:45
>> I cannot wait to get my fingerprints all over.
1:49:47
When somebody if somebody's looking at my computer and they touch the screen, I'm just like, >> "Isn't it the worst thing ever?"
1:49:55
>> Like I like >> you got to be carrying a polishing cloth.
1:49:58
You got to be carrying the Apple official polish. >> 20 bucks, you know?
1:50:02
>> Apple could make a Pixar lamp or something.
1:50:04
>> Well, Apple makes one.
1:50:04
Apple makes a Pixar lamp that just kind of polishes your touchcreen. >> That'd be great.
1:50:09
And AI for for screen cleaning. >> Yeah.
1:50:12
You know, the Apple polishing cloth.
1:50:13
They got so much, you know, flak for that thing.
1:50:15
Like you would think that it was $75, but it was $20.
1:50:20
>> Come to think of it, like it's really not that big of a deal.
1:50:23
>> Also, I know someone who is OCD and is obsessed with keeping their uh screen perfectly clean.
1:50:27
And I was like, "What's the secret? Like what?
1:50:28
You must have some secret formula like Windex or something that you're using."
1:50:32
And he was like, "No, just the Apple polishing cloth.
1:50:34
Like that's the one that works."
1:50:36
>> Well, I need to get one of those.
1:50:36
>> I was like, "That's a glow."
1:50:37
>> You know, one did come with my Vision Pro. So, >> yeah. There you go. Try that.
1:50:40
>> They threw one in as a bonus.
1:50:42
>> Isn't that nice of them?
1:50:43
>> That's so nice of you. >> $34.
1:50:44
99 you get a free polishing cloth.
1:50:45
that's 20 [laughter] bucks off.
1:50:45
Uh yeah, maybe maybe the the the Turnist narrative can center around like the as the models commoditize like the hardware becomes more important.
1:50:55
You want to be able to run different models locally and so pushing that like >> the hardware is great. The chips are great.
1:51:03
>> The software I'll even tell you is >> maybe not >> it's between good and great. Sure. >> Okay.
1:51:10
I'm not going to say it's only good.
1:51:11
I'm not going to say it's as great as the hardware, but it's good enough. How's that?
1:51:14
The AI is like the worst in the industry.
1:51:18
>> Yeah, but I mean right now you're seeing people go out and buy Mac minis to run AI.
1:51:23
>> Okay, let's think about it.
1:51:23
You see these people on Twitter doing that, right?
1:51:26
How many extra Mac Minis do you think were sold because of all this jazz?
1:51:29
I would guess I would put the overunder on 500 units. >> 500? I thought it was 10,000.
1:51:35
>> You seen it on Instagram.
1:51:37
>> There's 40,000 GitHub.
1:51:38
>> Okay, if it's on Instagram, maybe I'm wrong about 500.
1:51:40
>> So there's 40,000 GitHub stars. It's clearly a big meme. 60.
1:51:43
>> Maybe it's in the thousand.
1:51:43
But yes, I mean it's a quarter million a quarter or something.
1:51:46
>> And I think so much of it is just performative.
1:51:48
>> Go on the Apple store.
1:51:48
No, people >> go to the Mac Mini. >> They're in stock.
1:51:53
>> Are they all in stock? >> They're in stock.
1:51:54
>> Look at the ship dates. >> Funny thing.
1:51:55
Two weeks ago I go to the Pasadena uh I go to the Pasadena Mac store.
1:52:00
You know what's out of stock? Apple Vision Pros. >> What?
1:52:04
[laughter] >> Probably just because they're like not making them anymore.
1:52:06
>> A lot of a lot of the buying I think is purely statused and just >> No, no.
1:52:11
selling the Mac Mini and people just saying I want a signal that I'm AI native and I'm at I'm at the >> they're going to use it for two weeks and then forget about it. Yeah.
1:52:20
I I think that's going to be uh a real thing.
1:52:24
>> I mean, how many people actually need to live in these type of workflows? >> Not many.
1:52:28
Just the hackers, which is a niche community.
1:52:31
>> It's a niche community. It's a great community.
1:52:32
But it has made me think maybe I should start using Apple's native file system and like actually bring my data out of the cloud, out of drive >> and store it locally.
1:52:46
>> You don't need to because uh Cloudbot will go and access your >> So what's the deal with the Mac Mini? >> Uh it's shipping.
1:52:50
Order now, pick up in store today. It's available. Order by 3 p. m.
1:52:56
Delivers 2 hours from the store tomorrow.
1:52:59
It has Yeah, it's everywhere.
1:53:01
They're they're widely available.
1:53:01
They are widely available. >> So maybe I am right.
1:53:05
>> Available tomorrow at the Americana brand today and >> it's literally available at every minis a year. >> Yeah. Yeah. >> You think that's it? >> Yeah. >> Yeah.
1:53:16
>> And so and so it doesn't take that many if they're projecting out, hey, we're going to sell it's not like they I would I would imagine they don't have all of them that they're going to sell this year s sitting in stock already.
1:53:26
I'm curious how the Mac quarter is going to go tomorrow, right?
1:53:31
Like there might be a little bit of a drag on that. >> Yeah.
1:53:36
What should people pay attention to?
1:53:38
>> Well, the China number to your point, >> uh the iPhone number is basically everything tomorrow, right?
1:53:43
Either they >> they grow 10% as they say they will or they won't. >> Yeah.
1:53:50
>> Either Tim Cook gets to keep his job or he doesn't. No, I'm just joking.
1:53:53
[laughter] Um but you think about like the product, right?
1:53:56
Like they didn't do much iPad or Mac last year, >> right?
1:54:00
This year is going to be the biggest year for the Mac >> in a long time.
1:54:06
>> Got new MacBook Pro is about to launch.
1:54:09
>> Same design or as those ones with the faster chips.
1:54:14
>> You've got the iPhone chip powered lowcost MacBook which is going to destroy PCs and Chromebooks and be like just this utter game changer.
1:54:22
You've got the touchscreen MacBook Pro end of year.
1:54:26
You've got a refreshed Mac Mini.
1:54:26
You got a refreshed Mac Studio. You got the M6 chip.
1:54:30
You've got the [snorts] first new monitors from Apple in four years. >> Where will those sit? >> The monitors? >> Yeah.
1:54:37
>> In terms of They'll sit on my desk. I'll tell you. [laughter] >> Yeah.
1:54:39
But is it going to displace the studio or the uh Pro XDR?
1:54:41
[laughter] >> Uh the one I know about is going to replace the uh the Studio. There's a new XDR also. >> Oh, okay. >> Yeah.
1:54:52
And timing >> remarkably longterm.
1:54:55
Like I cannot believe >> why is there no camera on that thing?
1:55:00
>> Because it was it was created 15 years ago.
1:55:02
Like it was it's so old and you and you go and you look at like you could YouTube uh search for the pro prois XDR right now and there's guarantee a new video.
1:55:13
Uh why you should buy one in 2026. It's still good in 2026.
1:55:14
Like it's still the best option.
1:55:17
>> I mean the display is just remarkable.
1:55:19
>> It's just amazing that it didn't commoditize fast enough.
1:55:20
What are you guys using in here? I see.
1:55:22
Oh, the >> We have mostly studios.
1:55:23
We don't have a lot of XDRs and but we have been eyeing that new Dell monitor that Michael Dell has been rapaciously pumping on Twitter.
1:55:32
>> It's amazing that so enthusiastic about that thing.
1:55:35
>> I'm sure I'm sure you've answered this a hundred times and I'm sorry, but why why would they never do a TV?
1:55:39
Is it just commodity space?
1:55:44
>> Commodity margin differentiator.
1:55:44
You don't you don't think people would happily spend >> they're going to buy a Pixar lamp before they buy a TV. You walk in.
1:55:52
Oh, you got the Apple TV. I've heard that story.
1:55:54
As >> somebody who's like, you know, been an Apple I I can't describe myself as an Apple fanboy anymore, but as a kid, as a teenager and a kid, I was right.
1:56:03
Like, >> what the hell happened?
1:56:06
>> Uh the photos, >> the photos, [laughter] the photos ruined Jordy.
1:56:10
>> It just like, you know, a 20-year relationship just over.
1:56:12
[laughter] Um, but I do think there's enough people out in the world that if you made a 10 if you made the $10,000 TV >> Yeah.
1:56:23
>> that they would buy it.
1:56:25
>> You know, the >> because because when I'm buying a computer, a TV, sure it's different, but when I'm buying a computer, it's not upgrade.
1:56:34
>> Look at the premium on the Samsung frame TVs.
1:56:36
Those flying off the shelves, and they're not better than an LG.
1:56:40
>> So, my TV in my living room, I bought an 18. Yeah. What are we in 26 now? eight years.
1:56:44
>> Yeah, but Apple would figure out a way to deprecate the hardware [laughter] is what they do.
1:56:48
They're the best in the world at >> Well, if they did, they would do it, right?
1:56:51
They were they were >> they're like Tim's like, "Where's my TV?"
1:56:55
He's like, "Sir, we haven't found a way to deprecate the hardware."
1:56:59
>> They got pretty down the road on a TV about 10 years ago and then they they killed that thing.
1:57:02
They had teams working on it. It was a big deal.
1:57:06
>> Um, but then they went off and did a car and went off and did a Vision Pro.
1:57:08
If you think about like their two big moonshots over the last decade, they were both utter failures.
1:57:14
>> The car obviously is just like your quintessential failure. >> Yeah, pretty much.
1:57:18
Um, but you know, they did get some good stuff out of it, right?
1:57:22
Like >> I would say the saving grace for Apple's AI, and you've said this a few times now, has been the AI chip and the AI hardware.
1:57:29
The only reason they have an AI chip, the neural engine they launched in 2017 was because of the Apple car that was designed to power the AI needed for a self-driving car and they shrunk it down for the phone.
1:57:38
So if the Apple car project didn't get ignited, you know, back in 2014 15, they would be even further behind in AI than they are today.
1:57:47
And so give a shout out to the Apple Car team.
1:57:49
[applause] >> And you know, >> I'm still pulling for car with a naturally aspirated V12. Can you imagine?
1:57:59
>> It' be Graz real gated manual. >> Gated manual.
1:58:03
>> I think they would have just destroyed Tesla if they just didn't set their bar so high when we talk about overengineering.
1:58:08
Like can you imagine like a Model Y or a Model 3 or even an S, whatever.
1:58:12
Just like with that Apple interior and the Apple ecosystem and the Apple interface, like why did they have to go bananas, remove the steering wheel, remove the pedals, have everyone facing each other?
1:58:21
Like why' they have to go toll they overshot it.
1:58:23
Why' they have to go like all Apple on us, right?
1:58:27
Like why couldn't they just do just do a car just be a luxury brand carry would have been amazing. >> Yeah.
1:58:34
And we got a sock instead. >> Got a sock instead.
1:58:36
Those were the project shipped. Okay. So you got to give it. And you know what? >> That was a success. >> Yes. >> Sold out. Got people talking. >> Yeah.
1:58:46
>> You know, we did a review of the sock on on Bloomberg and people ate that thing up. >> Subscribed. That's great. >> Where's your sock?
1:58:53
Smashing down the >> I've got socks. I've got You know what?
1:58:56
I have I've got sushi socks on right now. >> Oh, no way. He's a sushi fan. >> No, not from Apple.
1:59:02
>> No, no, the your favorite restaurant in the valley.
1:59:04
>> No, no, these are from my mother-in-law.
1:59:06
Yeah, >> that's fantastic. >> Well, we kept you.
1:59:08
>> Yeah, we kept you much longer. But thank you so much. This is so much fun.
1:59:10
I'm so glad you're in LA.
1:59:12
>> Do I get to do the gong? >> Of course. >> Hit the gong.
1:59:14
How give us a number how long you've been writing?
1:59:17
How long you've been following Apple?
1:59:20
>> Uh, writing since ' 09. >> 09. >> Overnight success. >> There we go. Hit the gong. >> There you go. Boom. >> With authority. >> The Germinator.
1:59:29
>> Theminator live on TV.
1:59:32
>> Have a great rest of your day.
1:59:34
>> Let's Let's do this again.
1:59:35
>> Yeah, we got to do this again.
1:59:36
>> We'll be following your coverage tomorrow. >> Yes, tomorrow. Lock in everyone.
1:59:38
And let me tell you about Gemini 3 Pro, Google's most intelligent model yet.
1:59:45
State-of-the-art reasoning, next level vibe coding, and deep multimodal understanding.
1:59:48
I was I was so close to hitting an ad read during that I held myself [music] back.
1:59:54
So, I'm giving myself two.
1:59:55
I'm going to tell you about vibe.
1:59:57
co, where DTOC brands, B2B startups, and AI companies advertise on streaming TV, pick channels, target audiences, and measure sales just like on Meta.
2:00:05
And we have another guest in person live [music] in the TV and Ultra.
2:00:09
We got Welcome to the show. >> Please grab a seat.
2:00:18
>> I think Mark German took his Diet Coke, but you're welcome to one of mine if you get thirsty.
2:00:22
But um please introduce yourself for everyone who might be watching.
2:00:27
>> Yeah, so thanks so much for having me.
2:00:29
Uh my name is Miles Brundage.
2:00:29
Uh I lead a new organization called Avery.
2:00:31
It's called stands for AI verification and evaluation research institute and the basic idea is that AI is becoming critical infrastructure and you know everyone is depending on it but we don't really like make sure that it's safe and secure the way that we you know audit critical infrastructure.
2:00:48
Uh, and you know, we need something analogous to the cyber security industry which propped up to kind of make the internet secure.
2:00:54
We need that for AI systems.
2:00:55
And so, >> how many how many emails have you sent to Peter from uh Multibot in the last 48 hours?
2:01:02
>> I haven't actually >> cuz you on on the show yesterday, he was like his number one thing was like he's like, I need somebody to just handle the inbound from security researchers [laughter] and I just can't I not even multibot can handle uh processing all the inbounds. >> Yeah. Yeah.
2:01:16
Security researchers are quite interested in that.
2:01:17
But yeah, so the basic idea is, you know, we're kind of like a think tank that is trying to figure out how to build that new industry.
2:01:23
We think that, you know, it's good for improving safety and security outcomes.
2:01:27
If you have this rigorous auditing, specifically frontier AI systems, we're not so much focused on kind of the downstream.
2:01:31
So like upstream, how do we make sure that this is infrastructure the whole society can can rely on and not have to rely on companies doing their own kind of testing or kind of like, oh, do I trust the CEO's vibe?
2:01:42
Like that's not a good basis for uh for trust in a technology.
2:01:46
And so, uh, you know, we we just kind of launched recently and we're, you know, kind of excited to talk about our work. >> So, it's a nonprofit. >> Yep. >> For now.
2:01:54
>> Are we thinking for been through, you know, enough, you know, what one uh kind of controversial nonprofit to for-profit transition was enough for my life. >> Okay. Yeah, that makes sense.
2:02:04
Um, talk to us about uh how you're breaking down.
2:02:08
I mean the the issue of just AI security safety, you can go so many different directions from, you know, the the GPT40 psychosis to uh fake news to uh paper clips and gray goo and and you know, a thousand years in the future.
2:02:22
Um what is what's most interesting to talk about?
2:02:26
What's most important to talk about?
2:02:26
Uh there's also just like security issues as we see old bot >> thing. Yeah.
2:02:31
So we kind of break it up into four categories.
2:02:32
So there's uh kind of unintended system behaviors and so that includes things like hallucinations um you know on the uh you know on the kind of more extreme end like big misalignment deception type type things where the where the AI system is kind of taking uh actions that are unaligned with the user's intent.
2:02:49
Um and then there's misuse of the AI systems themselves.
2:02:53
So kind of you know someone trying to carry out a cyber attack with claude which you know has been confirmed to occur.
2:02:58
You know Anthropic was like hey like you know people connected to the Chinese government are doing this.
2:03:03
So yeah, >> very real issue.
2:03:04
Uh the third category is what we call emergent social phenomena.
2:03:07
phenomena. So that's kind of like kind of emergent interactions between you know the human and the AI that lead to these like psychosis, addiction, you know kind of like you know degraded kind of like learning those kinds of things and you know these are all kind of different categories but ultimately you
2:03:22
know you should look at all of them and then the fourth uh category is kind of you know normal security issues and that includes tampering with AI systems theft of AIP uh AI IP which you know there have been a couple confirmed cases of that like you know Kim let's talk about let's talk about Kimmy what's going on there. Are you up to date on the story?
2:03:39
Are you up to date on the story?
2:03:40
>> Um, so people are saying that that you know there was some kind of like theft of like the model.
2:03:46
That's it seems to me like the more likely explanation would be kind of one of two things.
2:03:49
be kind of one of two things. One is distillation like they're kind of you know they sampled from the uh you know the cloud API and the other is just that there's a lot of samples already on the internet and so they did kind of general scraping and then you know the way that
2:04:02
these systems often work is they they've read a bunch of stuff from the internet and they're kind of selecting a persona of like okay what kind of thing am I and like they see a bunch of stuff on the internet where like AI type things are saying I'm Claude I'm chat GPT and that it'll just like default into that persona. Yeah, the the feedback loop of
2:04:17
Yeah, the the feedback loop of the the pre-training now that we have an internet that is aware of AI and LLM is fascinating.
2:04:25
I I keep thinking about that New York Times interaction between uh who was it?
2:04:29
It was a New York Times reporter who wrote about interacting with uh GPT4 in Microsoft and he was like mean to it and then that that article obviously was very well shared and so it got baked in the pre-training and he says that now if I go to a new LLM and it finds out who I am it's like sort of adversarial. So, this is weird.
2:04:48
And I've noticed I've noticed, you know, it's really uh what is it?
2:04:50
The the Rockus Bassilisk a little bit.
2:04:53
Like, I've noticed that I I'll go to LLMs and because I write a lot on the internet and I'm obviously live stream and there's transcripts all over the place, uh that they pick up on who I am and what I do much quicker than I think most people.
2:05:06
No, it's hard to be anonymous these days. >> Yeah. You say thank you. >> I do not.
2:05:10
Um but I'm also not mean either. Sorry.
2:05:13
I'm kind of like a I'm like a neutral actor when it comes to like I don't say thank you, but I'm also not like braiding them.
2:05:18
So, when I did that thing the other day when everyone's like, "Oh, like ask Chad GBT like you know what, you know, like oh, make an image of what your experience is like being my like assistant or whatever."
2:05:26
It was like it was like a it was like a it was like cozy like sitting in a chair like reading a book or whatever.
2:05:31
So, it was not like oh I'm abused or whatever. >> Yeah. Yeah.
2:05:35
Talk about where you want to take the output of the work.
2:05:36
take the output of the work. obviously think tank I think Washington policy makers but also there's a a feedback loop of if you put out a really uh insightful statement or analysis the labs might absorb that directly uh who's the main uh audience >> yeah so we're trying to be kind of a hub for various stakeholders so not just policy makers uh you know some other key
2:06:00
ones are as you mentioned AI companies themselves both upstream the frontier AI developers also downstream like enterprise customers ers might want to know that, you know, they're they might be like, "Oh, well, I had meetings with, you know, Sam and Dario and so forth, but like I want, you know, I'm about to make a$10 billion kind of contract, like I want something more substantive, investors, insurers." Um, and so that's
2:06:18
Um, and so that's one of the things that, you know, we we just put out a big analysis with with various folks on, you know, how do you drive demand for this auditing?
2:06:26
You know, make sure that it's high quality.
2:06:30
And, you know, I think some of the mo the like most honest signals, so to speak, come from the private sector more so than regulation.
2:06:35
And like I'm pro some kinds of regulation, but like in some sense insurers are a great case where their incentives are very aligned to not kind of mispric the risks and they might, you know, be supporters of high quality audits.
2:06:46
And in fact, like one of our uh donors is the AI underwriting company, which is like one of the other players.
2:06:51
I know you had Tudo on recently.
2:06:52
They're like one of the other players in the >> Sure. Sure. Sure. Yeah.
2:06:55
Uh so uh the the the even though uh a big tech company might come to you to read an analysis uh they're not the ones that are actually funding the nonprofit.
2:07:07
It's from a more >> disperate. Yeah.
2:07:09
So we we're trying to avoid like depending on you know depending on industry too much although like you know very pro there being kind of like companies that are making money off of this and kind of selling their services to industry as long as they're good disclosures of conflicts of interest and so forth.
2:07:23
We we you know since we're a think tank, since we're doing a lot of policy analysis, we want to be kind of as as pure as possible.
2:07:28
So we we don't have any majority donors.
2:07:30
So far we haven't taken any cash from frontier companies.
2:07:34
We do take API credits so that we can kind of like you know audit their audit you know their systems also kind of like you know use you know open eyes models to assess anthropics and vice versa that kind of thing.
2:07:45
So we have like you know six frontier developers who've provided credits. >> Cool.
2:07:48
Um, what about uh I I I've been grappling with this fact that uh when the AI safety question came up, it was driven, and Dario touched on this in on his essay, how it was it was driven a lot by these like sci-fi doom scenarios.
2:08:04
And I think a lot of people in tech sort of looked at AI as a tool, sort of incremental, okay, yeah, it's autocomplete, it's knowledge retrieval, it's Google search, whatever.
2:08:12
Uh and then and then we wound up getting AI safety issues, but they manifested in very different ways than what was actually predicted.
2:08:21
No one was predicting the GPT psychosis necessarily or some of the other things.
2:08:26
People were predicting, oh, this will like swing the election and and and drive like everyone will be falling for fake news.
2:08:32
Videos that are fake, they do go out, but they get debunked pretty quickly.
2:08:36
Like I feel like we've responded pretty well to that, but then there's been a whole other host.
2:08:40
So how do you think about the timeline of risk and how you want to how far you want to look into the future? >> Yeah.
2:08:46
So you know the the way that we think about it is that you know on the one hand like you know people at at my organization Avery have various perspectives on these things like I personally am on the like AI is going very quickly and we could see some crazy stuff very soon end of the spectrum but I don't think you need to believe that in order to be pro you know AI auditing.
2:09:03
I think you know if you just kind of compare AI to you know other normal technologies and people sometimes say like well AI is not this this supernatural thing it's a normal technology well a lot of those normal technologies are audited for safety like if you buy a power bank it probably has been you know audited against like you
2:09:19
know underwriters laboratory standards for electrical safety so you know it doesn't catch on fire and stuff like that and so I think like even just getting to the normal technology you know level uh would be an improvement and then you know the case for auditing is like even more crazy if you're like, "Okay, someone could take over the world with this." >> Yeah, that makes sense. What were the
2:09:35
>> Yeah, that makes sense.
2:09:35
What were the biggest points uh that you agreed with in the Daario essay?
2:09:38
Was there anything you wanted to push back on?
2:09:42
>> Um, I mean, I would love to hear more about auditing as one of, you know, anthropics, [laughter] you know, policy platforms, but I, you know, but but, you know, more seriously, I directionally broadly agree.
2:09:50
I mean, I'm I'm, you know, I'm uh I'm maybe not the target audience as someone who, you know, has written and read a zillion kind of like these are the three five big issues in AI risks and here's what to do about them.
2:10:03
So, like broadly I agree with this is very serious stuff and you know we should take this seriously.
2:10:07
Jordy, >> uh, how are you, how do you, uh, yesterday we were talking about, uh, the risk of I think people talk about sort of agentic AI systems and sort of getting on some runaway path where it's taking actions out in the world.
2:10:25
And >> I don't know if there's enough discussion around maybe how an AI system like that could actually recruit everyday humans, potentially millions of people, to kind of like join their cause.
2:10:37
How are you looking at that >> risk in the in the context of >> overall AI psychosis, right?
2:10:43
These these uh >> point about like the just turn it off doesn't work if there's like 10,000 people that are like >> camped out on top of the data center. I love it. >> Yeah.
2:10:54
I mean, I haven't thought so much about the kind of like recruiting angle, but what I will say is that it is important to audit not just like the models themselves, but also like how are they used?
2:11:03
Do these platforms have good practices for like detecting if there's some crazy shenanigans like that going on?
2:11:09
And I mean just you know an example you know this this clawed code thing where it was being used by these Chinese hackers that's like not like the individual interactions were okay you know because they were basically like decomposing the prompt in like various subcomponents that are benign.
2:11:23
Um, and this was kind of like a known issue in the research was like, well, you know, if the if the model is just refusing stuff that looks like it's obviously malware or whatever, then you know, that's one thing, but if if the user kind of breaks it down into small chunks, um, then that's a big problem.
2:11:38
And so, >> and the risk is if you say like it's going to it's going to be easy to make it refuse to say like build me a bioweapon, but if I ask just like how do I learn about pipetting and then how do I how do I learn about this particular step and it doesn't know. >> Yeah.
2:11:51
and potentially like across multiple accounts.
2:11:52
And so like how do you kind of stop that in a way that is kind of like you know preserves privacy obviously >> not even just multiple accounts but multiple >> people >> models models. Oh yeah. >> Yeah.
2:12:04
No, it's like it's a super hard problem and and yeah, I mean I I would say like I'm interest and again like you know we I think it you know might be that we need to kind of just accept that like models you know more than like a year old or so like given how how fast kind of things get cheaper and get open source like we should just assume they will be like maximally misused in the worst possible way and just focus on like the very newest ones.
2:12:25
Um but you know for the very newest ones I think you know we need to get better at detecting those sorts of things because right now some of the companies will say we found this stuff other companies it's not so clear that they are actually like trying to stop >> so politicians have uh focused in on uh the the sort of risk or reality of rising energy costs.
2:12:46
>> Are you frustrated that so much of their attention is is on that versus some of these other issues?
2:12:50
Um, I mean, I think the I think the energy cost thing is more of a real thing than the water uh kind of issue.
2:12:57
And so like I so it's I don't think it's crazy to to worry about that, but um but yeah, >> but it's certainly taking up mental bandwidth and and just like dominating the narrative when maybe there's bigger bigger risk. >> Yeah.
2:13:10
>> Yeah. So I mean the you know my kind you know the things that I personally focus on are like you know there should be some kind of you know articulation what what what counts as safe enough or secure enough like and you know there's starting to be things like this and you know California New York um that say
2:13:24
okay you know you uh you know they may don't say exactly how safe but at least you should share you know whether you've measured catastrophic risks and share that you have a safety you know a security policy like eventually we should kind of ratchet up the standard there but you know some set of standards
2:13:39
then you need some evidence that people are actually following those standards and that's where auditing comes in that's where kind of um you know kind of transparency and and publishing system cards those kind of things come in and you need some set of incentives so uh to like make sure that people actually you know get face some penalty and so doing
2:13:56
all that in a way that doesn't crush you know small businesses and focus focuses on you know the very frontier of of the AI systems that's kind of what my focus is more so than the energy stuff >> what does an actual good audit or benchmark look like you can't just do like bad stuff per million tokens I assume. Uh how do you actually uh get to
2:14:11
Uh how do you actually uh get to some sort of quantitative metric that's tractable understandable but still valuable and not gameable? >> Yeah.
2:14:21
So we're kind of going through this transition right now from the earlier period period where like the strongest safety case you can make was you know what people call an inability argument.
2:14:30
It's like the model's too dumb to be worth worrying about in cyber security or bio.
2:14:34
And so like kind of just like show that it's too dumb.
2:14:37
That was kind of like the early wave of safety analysis and like we're starting to move into the world where actually like have you taught it to behave well and like let's just assume that it's capable because they're getting very capable.
2:14:50
Are the mitigations good?
2:14:50
And then it's mitigations at the model level like refusing stuff.
2:14:54
Um there's uh mitigations at the system level like you have a classifier outside the model that kind of like blocks certain you know outputs and so forth.
2:15:02
there's kind of at the platform level like detecting you know multiple fraudulent accounts that are you know coordinated and kind of like trying to steal the uh you know kind of like trying to distill the model and so forth.
2:15:12
So those are the kinds of things that uh you know there's starting to be work like for example there are organizations like meter and Apollo research secure bio transluce etc that are kind of looking at different aspects of that um but right now it's like largely voluntary and you know they're they often only look at a very small subset of those risks.
2:15:30
They're not really looking at the full the full gamut. >> Very cool.
2:15:33
Well congratulations on the launch.
2:15:36
I want to ring the gong for >> right here. Right here. Right here.
2:15:40
>> We love an AI nonprofit. Boom.
2:15:44
Thank you so much for coming on down to the TV pit culture.
2:15:49
>> Let me tell you about Plaid.
2:15:49
Plaid powers the apps you use to spend, save, borrow, and invest.
2:15:53
Securely connecting bank accounts to move money, fight fraud, and improve lending now with AI. >> That's right.
2:16:00
>> Uh we have some slight changes to the linear lineup.
2:16:03
Uh we're shifting some folks around.
2:16:05
There's some breaking news.
2:16:07
We're going to bring on uh different guests.
2:16:08
We're starting our lightning round at 1:30, but in the meantime, we can uh continue through the timeline.
2:16:15
Uh, does that sound good, Jordan? >> Sounds great.
2:16:18
>> Bubble Boy is talking about Claudebot in the race for Mac minis a bit over the last few days.
2:16:22
I think it's become a very scary phenomenon.
2:16:25
Uh, very scary realization that explains this crazy phenomenon.
2:16:29
Put simply, building a gaming PC will be nearly impossible in the next 5 years.
2:16:33
In fact, it already is for the vast majority of consumers.
2:16:35
But I will go one step further.
2:16:37
In the next 10 years, having any type of personal computing device will be unattainable.
2:16:42
Fab capacity will be allocated to its most productive and profitable use, which is cloud and AI data centers.
2:16:46
Even today, most of the software you run already won't work without an internet connection.
2:16:53
But now, with the opportunity cost being so high, consumers will be shafted and the only option uh will be moving to the cloud.
2:17:00
It's looking increasingly like the only hardware you will have is some terminal that connects to the cloud with no workloads running directly on your own hardware.
2:17:06
Your device will just have the most basic single-core processor and four gigs of RAM at most.
2:17:11
That probably sounds like the specs on the on the on the Sweet Pea ear pod AirPods or whatever the OpenAI AirPods are.
2:17:18
Um, what's uh what's interesting is I was digging into the the the amount of of leading edge capacity at fabs like TSMC that are currently dedicated to um to AI workloads or creating GPUs that can run AI workloads.
2:17:41
And it's about 50% right now. It's about 50/50.
2:17:43
So TSMC obviously fabs Nvidia chips and TPUs I believe but also they do Apple silicon which is not really used for you know Frontier LLMs in the cloud.
2:17:57
Um the question is like how much more can the Frontier AI labs eat into that?
2:18:02
If is it just a bidding process and if Nvidia goes to TSMC and says hey we want to buy 80% of your capacity and you got to tell Apple to take a hike would they actually do that?
2:18:13
Uh TSMC has been a little more conservative.
2:18:15
Ben Thompson's been writing about this a lot that TSMC has been a little bit slower to invest the capex to build the next AI fab or the the next frontier fab uh because it's something like $50 billion in uh in capex and if the AI uh trend sort of slows down they could be caught holding the bag since this is a multi-year buildout.
2:18:36
But uh it's interesting that the the energy narrative we we are constrained on energy but we're only using like 1% of energy in the west on AI and data centers that can power AI.
2:18:48
We're using like 50% of our of our AI of our fab capacity on AI.
2:18:53
Crank up both but uh one requires actual building of new capacity.
2:19:00
If you want to see an order of magnitude increase, the other requires uh if you want to see an order of magnitude increase in energy, going from 1% to 10% of energy used for AI, uh you can just slide chips around the board. >> Yeah.
2:19:12
I mean, I definitely agree with the bubble post like even stuff like gaming. >> Yeah.
2:19:17
>> Uh like on the MetaQuest Xbox edition, like you there's no actual like you're not playing the game on the device. It's like streaming. >> Yeah.
2:19:23
>> So I I think this is like very >> I think this is like definitely true. >> Yeah.
2:19:28
>> Everything goes into cloud. >> Yeah.
2:19:29
And so like it's that the the the interesting thing is that it's uh it's sort of like the you will own nothing and be happy thing.
2:19:35
Like uh the average consumer might just be happy with that because they're like yeah it works.
2:19:39
I don't care where the workload happens.
2:19:43
Of course like the George Hots of the world want the tiny box.
2:19:45
They want to be able to run it permissionlessly.
2:19:46
They want to be able to run their own model.
2:19:49
Uh and uh and there's always different Yeah. Yeah.
2:19:53
Not your keys, not your coins. That whole movement.
2:19:54
there's always been uh a lot of debate around uh you know the the sovereignty of your individuality and uh I think that will continue and I think that uh the I I I I don't think the Mac Mini is going to go out of stock.
2:20:09
I think that uh many of the trends in Apple hardware availability will continue and I think Apple has a lot of leverage at TSMC to continue manufacturing their devices.
2:20:17
Uh while we were talking with uh the Germinator and Miles, the Fed announced they voted to keep rates steady for the first meeting of 2026.
2:20:28
This was predicted by >> uh Kalshi uh uh so no real surprises here.
2:20:36
Uh >> yeah, we yeah, we were tracking the KHI market on this because uh there was that big uh back and forth with Jerome Powell.
2:20:44
We were playing the we are Jerome Powell song.
2:20:45
Uh and it seemed like there was an incredible amount of pressure to um to for to to pressure uh Jerome Pal into potentially cutting rates.
2:20:54
It seems like there the guidance is that there will not be a rate cut.
2:20:55
Um and there was another piece in the journal about some of the forerunners for the next chair of the Fed.
2:21:02
Uh there's four people that were in the running that I saw, but none of them were um completely >> Trump has basically said there's a bit some problems with each of them.
2:21:11
also [snorts] says, "I'm worried that when I'm interviewing them, they're going to say one thing and then be a little bit more independent once they take the job."
2:21:21
>> And also, it's a balancing act because if you install someone who is not uh who doesn't bring credibility to the market, uh you don't get the you don't get the desired outcome.
2:21:30
So, you can't just have someone that uh that cuts cuts rates to negative 50% as some people might like.
2:21:38
Uh 11 also >> build intelligent real time conversational agents [music] reimagine human technology interaction with 11 labs.
2:21:46
>> Meta also reported earnings eat on uh bottom line and topline.
2:21:50
Uh stock is up 4% after hours expanding 2026 capex.
2:21:59
What I'm seeing I'm trying to get up to speed here.
2:22:02
>> Uh Tesla had uh reported earnings as well. >> How's Tesla doing?
2:22:06
Uh Tesla saw negative revenue growth I believe for the first time but the stock is up 3.
2:22:12
7% after hours >> and Microsoft also beat consensus on top and bottom lines and uh the conference call starts at uh 2:30 in just uh about an hour.
2:22:26
Um shares fell 4% in extended trading Wednesday after the software maker posted slowing cloud growth.
2:22:32
Um they did Nice little markup on AI. >> Yeah, not bad. Oh, yeah. Yeah, exclude impacts. Okay.
2:22:43
>> Anyways, we'll we'll try to put together some more information here >> uh in the near future.
2:22:48
We did need to >> give a shout out to Colin and Samir.
2:22:51
We went on their podcast at the end of the year, sat down, talked about uh >> they posted the full two hours.
2:23:01
[laughter] >> They ripped the full two hours on X. >> I wasn't sure.
2:23:03
So Colin Samir have have a number of very interesting formats.
2:23:07
They have a podcast feed.
2:23:07
Uh they uh I watched a recent video that they put out.
2:23:12
It was like, you know, 10 lessons.
2:23:15
It was almost like a a YouTube video trailer for a full episode and uh and I was I was wondering like we talked a lot about a lot of things.
2:23:23
How uh [laughter] yeah, how what will this actually uh land? How will they edit it?
2:23:28
Uh what will they be thinking?
2:23:30
uh grid to see the full hour and 50 minutes up on the timeline. I love it.
2:23:34
And oh, they got Max Sparrow in there.
2:23:36
This is a nice little intro. I like this edit.
2:23:39
>> Yeah, it's quite nice. So, go check it out. Super fun conversation. We love those guys. >> Yeah.
2:23:43
Us talking about our business, the business of media.
2:23:44
Uh the next chapter, >> anti-scale strategy. >> Yeah.
2:23:50
>> Some stuff we mentioned here.
2:23:52
>> And uh we have a surprise guest today.
2:23:55
the the new Neolab Flappy Airplanes.
2:23:59
He's going to be joining the founder.
2:24:01
Aiden has been on the show before. >> Yeah. During the teal.
2:24:04
>> So excited to have him back on and get the update there. >> Yeah.
2:24:08
>> Uh Emily, before we move on to Emily Sunberg, I got to tell you about Octa.
2:24:12
Octa helps you assign every AI agent a trusted identity so you get the power of AI without the [applause] risk. Secure every agent. Secure any agent. Yes. >> Barry Weiss.
2:24:21
uh everything she says internally immediately becomes external messaging.
2:24:28
Uh tough tough job but uh she put in uh she was talking with the team and saying she wants to put a huge emphasis on scoops.
2:24:38
>> This is about CBS >> CBS. Yes.
2:24:39
Not >> so she's the editor-in chief at CBS but uh she was acquired in through the free press.
2:24:44
So the free press is operating the CBS is operating and there's a there's an effort to sort of merge the two.
2:24:50
free press obviously very fastmoving startup culture CBS has been around for uh decades.
2:24:54
So, um, the the the quote that has hit the timeline is, uh, Barry Weiss also said that the network CBS will now put a huge emphasis on scoops, but not scoops that expire minutes later, but investigative scoops and crucially and crucially scoops of ideas, scoops of explanation where this is where we can soar and where we'll be investing, she continued. Scoops ideas.
2:25:24
People are people are roasting this but I think >> I think one fires me up. >> Yeah.
2:25:31
>> Also you just look at the modern media landscape.
2:25:33
There are a lot of scoops of ideas that are they create these like everyone watched that podcast.
2:25:36
Everyone is repeating that idea from that person. They set the tone.
2:25:41
Uh they introduce this concept, this philosophy, this thesis.
2:25:45
They unpacked it for two hours exclusively in this one place.
2:25:48
You got to go there to get it. >> Yeah. Yeah.
2:25:52
And it starts a conversation that wasn't happening.
2:25:53
It's not just like here's some facts that then immediately get copy and pasted everywhere.
2:25:57
That's what she's talking about.
2:25:59
Scoops that are expiring.
2:26:00
>> It's not about you have to uh start a conversation that that wasn't. >> Yeah.
2:26:05
>> Yeah. I mean I I I I go to Darwash Patel the the uh the Leopold Dash and Brener you know it wasn't a scoop I guess but situational awareness was uh posted as PDF but also uh you know made available in a multi-hour sitdown podcast where he unpacks those ideas that flew around that started a conversation about where
2:26:25
AI is going its impact on the financial markets uh Andre Carpathy same thing on Doresh there's been a number of things where uh certain ideas certain scoops certain explanations have recontextualized things and I think that CN uh CBS has the the production horsepower to bring a really polished product together. Um but if there's
2:26:46
Um but if there's something that's been already revealed, the facts are out and then you just do you're reinstantiating on on CBS that's not like mustwatch necessarily for the modern media consumer. So I don't know.
2:27:00
I think I think it's a reasonable thing to say even if it's sort of a funny phrase.
2:27:03
Um, >> keep running through the timeline. I'll be right back.
2:27:06
>> Okay, so uh, Miles says that when Google hit a $500 billion valuation, they had 90 billion in revenue and 20 billion in profit.
2:27:14
OpenAI is raising at 800 billion with 30 billion in estimated revenue in 2026 and 30 billion in free cash flow.
2:27:23
Anthropic is raising at 500 billion with 20 billion uh, in 2026 revenue and negative 15 billion in free cash flow.
2:27:30
Was Google just free money?
2:27:30
Why did these price so differently? uh mainly growth rate. Let's see.
2:27:34
Uh Miles says, "Google is growing at 20% at 90 billion in revenue with significantly better quality of revenue and had profits."
2:27:42
Yeah, but I mean Anthropic and OpenAI are growing at a,000%, right? Or 500% or 300%.
2:27:47
Uh the the growth rate really does matter.
2:27:51
And I don't know, maybe there's there's some sort of uh you know, excitement just around uh the AI bubble.
2:27:58
But Tyler, what do you think?
2:28:01
Anthropic at 500 with 20 billion in 2026 revenue and negative 15 billion dollars in free cash.
2:28:07
>> They've 10xed it like three times.
2:28:07
I could I could see a few more 10xes. >> I agree. >> I could see OpenAI. Couple more 10Xs. >> I agree. >> Deep mind. >> Yeah.
2:28:16
>> I want to see Demis in the, you know, in the hot seat. >> Yes.
2:28:19
>> You know, like two or three more 10xes. He's CEO. >> Oh, yeah. >> I want him up there. >> Yeah. Get him up there.
2:28:23
I I still think it's so funny that he was like, "We're not doing ads."
2:28:27
It's like that's the easiest layout to be like we'll do it when we're going to do it. I don't know. >> Yeah.
2:28:31
Also like he is actively being funded by ads. >> Yeah.
2:28:34
And and they did put out that they did put out that note that uh that Google's uh AI powered search is now powered by Gemini 3 Pro, right?
2:28:41
Uh and then they've also monetized those.
2:28:44
So maybe there's not ads in Gemini, but there's a Gemini powered product that has ads in it.
2:28:51
And so I feel like that's >> I think it was just kind of a like free dunk. It was a free dunk.
2:28:55
No one's really going to fact check him.
2:28:57
>> But the more nuanced Duncan and we're going to have Eric Seford on the show to discuss the roll out of OpenAI's ads product is that uh I mean Ben Thompson earlier today said that uh it it h expectations were low and it did not meet them just in terms of the robustness of the AI product, how they're pricing it.
2:29:12
He compared it to Netflix's early roll out of of of ads.
2:29:16
The ad product specifically saying like you're going to pay uh CPMs for impressions based on like you know these categories.
2:29:22
Like it's not just going to be this robust Facebook level like black box that just gives you you put dollars in you get dollars out.
2:29:29
Like everyone wants like the ATM machine [laughter] when it comes to ad products. Truly.
2:29:33
I mean like it sounds silly but people like advertisers really want this like massive scaled self-s served platform that they can just go and and and and partner with and it feels a little bit more bespoke, a little bit more incremental and it and and truthfully a lot of people are saying like they should have started working on this earlier.
2:29:49
We should have they they should have uh you know been building along this line earlier but of course they had a lot a lot of other stuff to work on. Anyway, Shopify.
2:29:56
Shopify is the commerce platform that grows with your business and lets you sell in seconds online, in store, on mobile, on social, on marketplaces, and now with AI agents.
2:30:07
Uh, moving on back to DeepMind.
2:30:07
Google bought DeepMind 12 years ago yesterday.
2:30:14
Uh, they it was [laughter] such an insane steal.
2:30:16
Run said they bought it for a buck 50. Lol. Insane steal.
2:30:18
The numbers were so much smaller back then.
2:30:23
in the opening episode of HBO's Silicon Valley.
2:30:25
Have you seen HBO Silicon Valley?
2:30:26
We talked about this, right? Stressful.
2:30:28
So, you never watched the full thing?
2:30:30
>> Yeah, I've never I've never actually watched it just because I A show like that would be like before bed and it just reminds me of >> work.
2:30:36
Yeah, there were some hilarious hilarious moments in that show though.
2:30:40
Uh >> in the in the opening episode of HBO Silicon Valley, the protagonist is offered a $4 million valuation for an insane new technology which is so absurdly large to him that he throws up.
2:30:51
[laughter] It's like $4 million valuation really does feel low nowadays.
2:30:54
That is that is like pre preede.
2:30:57
Like there's nothing that YouTube, DeepMind, and Android were acquired for less than $2. 5 billion combined. What a time. That is wild. And >> insane pickups. >> Insane. Insane pickups.
2:31:11
Um let me >> uh I don't actually think we covered this, but the update on the SpaceX IPO, they're timing it.
2:31:19
uh uh they're looking at a June IPO, so uh not too far away.
2:31:26
And Elon is apparently trying to time it uh to planetary alignment and Elon Musk's birthday.
2:31:31
So, Celeste uh the Financial Times says celestial calendar meets high finance as billionaires personal impulses shape plans to raise 50 billion in record listing.
2:31:41
Uh >> yeah, I think he's he's allowed to to plan his IPO around his personal impulses.
2:31:49
Yeah, >> this is this is his baby.
2:31:51
>> Some numerology, some good uh some good vibes.
2:31:53
He's just managing the vibes.
2:31:56
>> Um >> before we move on, let me tell you about Phantom Cash.
2:31:58
Fund your wallet without exchanges or middlemen and spend with the Phantom card.
2:32:04
>> Uh yeah, I don't think there's many other uh Musk's birthday is June 28th. >> Okay.
2:32:10
And on June 8th and 9th, Juniper and Venus will be quote within a little more than one degree of each other in the sky.
2:32:19
>> Sorry, you said his birthday is June 28th. >> Yes.
2:32:21
>> With this Sunday, markets closed.
2:32:24
>> So, he's not going to be able to do it.
2:32:26
>> He's got to delay it a year.
2:32:26
He's got he's got to wait a full year until until [laughter] >> uh hopefully not.
2:32:30
Um [snorts] but uh yeah, a few days later, Mercury will also align diagonally with the two planets.
2:32:38
The unusual demand highlights Musk's personal imprint on SpaceX.
2:32:41
Of course, >> uh where major corporate decisions have often reflected the billionaire's beliefs and priorities. >> Yeah.
2:32:50
>> What what corporation has a CEO, a founder CEO, where the the the the founders beliefs and priorities don't shape the corporate decisions.
2:33:02
Yeah, that's >> they're trying [laughter] they're trying to like they're trying to like spin this.
2:33:08
Uh it's not it's not really hitting. >> That's odd.
2:33:12
[laughter] >> Uh but uh >> there's another there's a massive scoop in the Wall Street Journal from Kate Clark.
2:33:18
She says SoftBank is in talks to invest a lot more money in OpenAI.
2:33:21
Um this is surprising tons of people because everyone said Moss is tapped out.
2:33:28
He can't possibly marshall the capital.
2:33:30
He'll have to sell everything. David Gogggins.
2:33:34
>> He's David Gogggins of of buying OpenAI shares.
2:33:37
>> Well, he's not going to It doesn't seem like he's going to get it from Saudi. >> Oh, yeah. >> Where's he? >> I don't know.
2:33:42
I mean, he has he has other stuff he can move around.
2:33:44
He can sell other positions.
2:33:46
Like, Soft Bank's a massive organization and uh and going all in. >> Yeah.
2:33:51
So, Soft Bank's in talks to invest up to 30 billion more in Open AI, Japanese.
2:33:56
>> So, pull pull up this video uh from Buco Capital and because I think it there's a chance that There's a small chance that Masa saw this and said, "I need to own more."
2:34:08
>> Are these actually Sora videos?
2:34:08
Are these just like random cling like nonsense?
2:34:12
Like >> Buo says two trillion in capex for this by the way.
2:34:15
>> Okay, get this off the screen.
2:34:15
I'm not a fan of this at all. This is bad.
2:34:17
Uh, SoftBank Group shares closed 3.
2:34:20
7% higher in Tokyo trading Wednesday after the journal's report on the latest talks over OpenAI.
2:34:28
So the Tokyo markets are bullish on getting more OpenAI shares.
2:34:33
Uh this is of course the hundred billion dollar round that will have a post money valuation of 830 billion if it succeeds in raising the full amount. Deal talks are ongoing.
2:34:43
Soft Bank's already one of OpenAI's largest shareholders.
2:34:45
The stake that grew to 11% in December when it invested 22. 5 billion.
2:34:49
Uh in a statement at the time, SoftBank chairman Masayushi Son said the firm was deeply aligned with OpenAI's vision. He'd love to see it.
2:34:59
Um on the other side of things, Jim Kramer is saying he's now a Jimmy boy saying, "I still pay for Chad GPT and Gemini, but I have been favoring Gemini of late.
2:35:09
My two neighbors at the office are cancelling Chad GBT.
2:35:11
I like how much Gemini knows about me and is friendly, but not cloying, and it's not a suckup." Well, you know, 5.
2:35:18
3 is coming soon, and I think that the uh the the conversational element, the the textual flavor might be coming back, and they might be, who knows, maybe they fine-tune right on Jim Kramer's tweets to let him know that, hey, we're not we're not we're not afraid of pushing back on you, Jim, if we need to.
2:35:38
Uh, OpenAI, we'll do whatever it takes.
2:35:42
>> Uh, anyway, we we have our next guest available. >> Here we go. >> Who who do we have? We have Asher. >> There we go.
2:35:47
and Aiden from Flapping Airplane.
2:35:49
Amazing in the ream waiting room.
2:35:50
Let's bring them in to the TV pin Ultra. [music] >> Hello. >> Welcome. >> Welcome back, Aiden. Uh good to see you. How are you guys doing?
2:36:00
>> Hey guys, we are doing awesome. It's good to be here. >> Fantastic. Welcome back. Uh >> big big day. >> Big day.
2:36:05
Please take us through moment while you guys are do the >> tell us about the name first.
2:36:09
Let's all do the flap for flapping airplanes. What a name.
2:36:15
>> Well, so you know we are we're a new AI lab.
2:36:16
We're focused on the data efficiency problem.
2:36:18
We're trying to train models that can be roughly as intelligent as humans without ingesting half the internet. >> Okay?
2:36:24
>> You know, in order to do that, we want to think at least a little bit in a way that's a little bit biologically inspired.
2:36:29
We don't want to build a bird, right?
2:36:30
You know, obviously airplanes are fantastic, >> but uh to maybe help them flap their wings is is the right metaphor for what we're doing.
2:36:37
>> Well, you're exactly right. >> That's hilarious.
2:36:38
Where does the milk fit in?
2:36:42
>> Uh I'm not here to comment on the milk.
2:36:44
It's a big part of our third culture, but I I can't say more than that. I I love it.
2:36:47
It's the most probably the most underrated drink out there.
2:36:52
>> Not enough people are taking advantage of of all the benefits, >> you know.
2:36:55
We got so many milk ads throughout my whole life.
2:36:57
The got milk campaign was fantastic and then it sort of just disappeared.
2:37:01
They haven't been running ads.
2:37:03
>> I went Yeah, I went through a phase where it was like borderline a gallon a day. >> Go mad.
2:37:07
Yeah, there's a phrase for it.
2:37:09
Anyway, [laughter] >> any I will be buying dairy securities.
2:37:12
I got a lot of milk >> coming to the [laughter] >> unlimited milk new new perk. >> That's fantastic.
2:37:19
Okay, so yeah, take us through like the status of the company like uh how far along are you? How big's the team?
2:37:26
Uh what what what are you thinking in terms of like release time?
2:37:29
Are you are you planning on sort of going heads down age of research mode, SSI mode and and releasing something when you've hit some some major milestone or do you want to be more iterative with it? >> So, okay.
2:37:42
So, we're we're about two months old. The team is now 11. We're super excited.
2:37:45
We've got people we really love who are who are both brilliant, but also just wonderful people.
2:37:49
We're really excited about it. >> Great.
2:37:51
>> Um I think I think we are sort of in the middle.
2:37:53
We're definitely age of research mode.
2:37:55
Like I think the goal is not to commercialize.
2:37:56
Not because we're not commercial people.
2:37:58
Like my background, even Aiden's background, Ben's background, all all reasonably commercial in some sense as oo in addition to deep research. >> Sure.
2:38:05
>> Um it's just that when you like when you get revenue, you have to focus on it.
2:38:08
you have to focus on providing quality for customers and that makes it harder to build, you know, deep technology. >> Yeah.
2:38:14
>> So, you know, our our goal is to try to to find the biggest market we can to solve the most important problem we think we can solve, which is the data efficiency problem before doing anything of that.
2:38:22
>> At the same time, like our approach is probably to be a build building a little bit more in public.
2:38:25
We'll release some research artifacts at least that I think will be will be cool reasonably soon, but you know, who knows exactly when the runs will finish or or how how many times they'll crash before they work.
2:38:35
our biggest training run today.
2:38:35
So, you know, bad timing next launch for our uh maintenance. But >> here we go.
2:38:40
Uh [laughter] walk me through why data efficiency and why the data efficiency a problem is important. Uh it sounds all good.
2:38:46
Oh, you get a really smart model and you don't have to ingest the whole internet, but everyone can just ingest the whole internet.
2:38:55
Like you can download it, you can scrape it there.
2:38:57
You know, there's plenty of models that have been trained on it. So, break it down. >> So, exactly.
2:39:01
I I think the goal is not necessarily in the long term to not train on the entire internet.
2:39:05
I mean, I'm it's research. I don't exactly know.
2:39:09
I think the idea is that like this is not needed, right?
2:39:10
And and the fact that it's not needed suggests that we're actually missing something because currently for the for the existing technology that we have it, it is necessary. >> Yeah.
2:39:19
>> So, why do I think it's an important problem?
2:39:20
you know, to the extent that AI has been hard to integrate into the economy and you know, we always see, you know, these Bloomberg articles that are like, you know, oh, like chat and search are working and coding is working, but like what else is AI really doing for me?
2:39:33
To the extent that's true, I really think it's because models are much less data efficient than humans.
2:39:36
But like if you wanted to learn a new task or put in a new vertical, it like it takes thousands of times more effort than it does to just tell a human what to do.
2:39:44
>> So I think if you can make a model a million times more data efficient, it's like a million times easier to put into the economy.
2:39:49
I also think there's just like tons of cool stuff that you can do in really data constraint regimes if you if you can learn to learn with less data.
2:39:56
For example, whether it's robotics or scientific discovery or even something like trading which we have to acknowledge is like the most valuable next token prediction problem in the world from a pure economic perspective.
2:40:06
>> Um these problems have very limited data and and you know existing AI systems aren't quite as good as them as as they are at other things.
2:40:11
I I think that you know learning to learn with less data is just tremendously valuable in all these things.
2:40:16
Talk to me about like fragmentation and steerability.
2:40:18
If you achieve sample efficiency like you're planning to, do you envision a world where you're sort of creating some sort of base model that is so sample efficient that just with like a basic prompt or a few examples, it becomes incredibly effective at a specific task and and sort of replaces like the heavy duty training data runs and the RL environments and these massive uh data collection uh efforts to do some sort of fine grain task.
2:40:49
ask at a very very high level to sort of like 59s of efficiency.
2:40:54
Uh is that what it looks like or is it or is it more like you will wind up vending a sample in efficient model that's for specific use cases? >> Yeah.
2:41:04
So one way we think about this is that it seems like the reinforcement paradigms of today are actually just shockingly inefficient.
2:41:09
You don't really get much generalization across tasks.
2:41:12
You you teach a model to do one kind of of learning and then you teach to the next one.
2:41:15
It's kind of like whack-a-ole or something.
2:41:17
And we look at this and we think this is kind of crazy.
2:41:19
I mean the first time I really saw RL scale what it brought me back to was actually good oldfashioned AI back in the dawn you know this primordial age of AI when people were kind of handd designing these convolutional filters for eyes and noses and things. >> Yeah.
2:41:32
>> And then then we're like wait just throw data at it just throw scale at it. What are you doing?
2:41:35
And in this really weird way we're kind of looping back onto that where it's like oh just make another environment bro just make another environment.
2:41:41
>> Just one [laughter] more just one more.
2:41:43
>> The next AI will not be just you know environment slop.
2:41:45
environment slop. I mean I think I think I think it is a piece of it like I do think that there's a long tale of tasks in the world right and like there will always be a place for people to produce custom data it's just like a question of how much operational difficulty it takes and like one route is just to slog through the operational difficulty
2:41:59
constantly you incur variable cost another route is to do a bunch of fixed cost investment into trying to make that variable cost lower so the last thing I'll just say to answer your question I don't really think we know what the end goal looks like entirely like you know AI is a big space like you know human level intelligence is not the ceiling it merely the floor on what is possible. Like if you can train models with vastly
2:42:18
Like if you can train models with vastly less data and and possibly more compute in very different ways, like what is going to happen?
2:42:24
Like we actually don't know.
2:42:25
Like I think I think it's it's actually unlikely that they will uniformly dominate frontier models even in the very best case scenario for us.
2:42:31
They're like going to know less facts.
2:42:33
They're going to be different.
2:42:33
They're not going to have memorized all Harry Potter.
2:42:35
Like that's actually a useful skill in some ways.
2:42:38
>> But I do think they'll be different and weird and like they'll have have interesting capabilities that that we'll find, you know, really valuable ways to use.
2:42:44
So, so yeah, I think it's an experiment we're really excited to run. >> What? >> Yeah.
2:42:47
What do you expect the economics to look like in a more sample in efficient environment?
2:42:51
I it feels like obviously you're not paying for a bunch of reinforcement learning environments and a bunch of data, but it also feels like the the the training cost might be lower.
2:43:00
Is that the correct assumption?
2:43:02
And then what does inference look like?
2:43:04
Does that does it get cheaper? Is it a smaller model? Can it run locally?
2:43:06
Are there any other uh downstream economic impacts that you think might come from a new architecture?
2:43:15
>> I mean, there exists this whole smoresborg of companies right now that are basically doing reinforcement learning for you. Yeah.
2:43:20
>> Or are taking your big pile of corporate data and making it useful in your model. Yep.
2:43:23
>> And this is great, but it's actually a huge pain.
2:43:24
And both sides deal are upset.
2:43:27
You know, the companies that are doing it are like, "Please give us more data."
2:43:29
And the and the their clients go to them and like say, "This is all we have.
2:43:34
Like, how can we possibly give you more here and these results are not meeting what we want?"
2:43:36
And the demand for better data efficiency a year is just huge.
2:43:40
And we really believe that even winning here will have massive effects on economy.
2:43:44
>> I also I don't I don't claim to know exactly what the economics are going to look like.
2:43:46
You know, for example, is this going to be accomplished with models that are smaller or bigger? Like I don't know.
2:43:50
I actually I I slightly suspect bigger.
2:43:52
Like I'm not sure it's actually going to make inference costs easier.
2:43:55
>> Um I mean, if you look at the brain as a comparison, you know, the brain spends a lot of flops per token relative to what a current model does.
2:44:00
Um so, so you know, we don't have to use the brain as a model, but it's like an interesting data point.
2:44:05
So I I don't I think it's a little bit for for us to say anything definitive about economics, but I I'm certainly intellectually interested in these questions.
2:44:11
I haven't been thinking about them for a while.
2:44:14
>> How are you uh approach like are you guys going to be spending all your time locked into the office just doing your own research or is this or do you have opportunity to go out in the world, find companies and organizations that have been disappointed by existing solutions and say like how do we kind of reapproach this problem? Right?
2:44:31
because you're saying like maybe it's not just more data.
2:44:35
>> I think we're going to start with research.
2:44:36
I I think the main thing is like when we work with companies, we want to make sure that we are like fully devoted to actually providing genuine value for them like and and not just using them as like a stepstone to to do interesting research.
2:44:46
So I think I think for now until we feel like we have some technological edge, we'll be very very focused on research.
2:44:50
I do think over time to your point though like it is helpful to research to actually be able to go out into the world and see what the problems people are facing are and and we actually do plan to do that but just not like in January. >> Yeah.
2:45:02
Uh what talk about uh fundraising uses of funds uh GPU poor GPU rich like what you know the the AI talent wars like what what what are the constraints on on your progress here?
2:45:16
>> Well so I mean first I I just want to say we're really grateful to our partners.
2:45:18
I mean, they've shown a lot of trust and faith in us, and we're going to do everything in our power to reward that faith.
2:45:22
So, it's it's it's humbling. It's really exciting.
2:45:24
We're we're really grateful to work with them. >> Thank for the milk. >> Yeah. [laughter] It was a no.
2:45:32
>> Um uh so on what the computer is for, I mean, the good thing about doing foundational research is like the stuff we're doing is often so weird that like it doesn't need to be done at gigantic scale first because with very high probability, it's going to fail at smaller scale.
2:45:45
And like if it starts to work at smaller scale, like if it's really working, if it works so well that you know you should you should see pretty strong signs of success.
2:45:51
So you actually you need much less compute to like get a 10x win at least to get it off the ground than you need to get like a 10% win. >> Mhm.
2:46:00
>> So you know that that's good.
2:46:00
I mean the the raise is is primarily for compute is is to answer your question. >> Fantastic.
2:46:05
um on on the talent wars, you know, I think the way we think about it is and this is very much informed by my experience as a member of of this organization prod and my brother's experience as a co-founder.
2:46:15
We're really not completely caught up in the same talent war as everyone else.
2:46:19
We're trying to find the next generation.
2:46:22
We're also hiring experienced people and like I think that has resonated with experienced people, but like you know we're trying to reimagine new ways of doing things.
2:46:28
Like I like don't really think that having trained a trillion parameter model before is like the thing that's going to make someone successful at that.
2:46:35
Like obviously being a good engineer, being smart, being excited and curious about the problem.
2:46:38
These are things that are tremendously valuable.
2:46:40
And like lots of people, you know, have that, but I I don't I don't know if like, you know, five years of experience uh is is really the thing that's most important. >> I love it.
2:46:48
Uh well, congratulations.
2:46:48
100 million $180 million raise, $ 1. 5 billion valuation.
2:46:53
I want to ring the gongs for you. >> Do it.
2:47:00
Uh super excited for you guys and the whole team.
2:47:02
I'm sure you'll be back on uh very soon and enjoy the research.
2:47:07
>> Yeah, have a great rest of your day.
2:47:10
>> We'll talk to you soon. >> Great to see you. >> Goodbye.
2:47:13
>> Up next, we have Alex Dylan from outtake. ai.
2:47:18
He's in the reream waiting room and we'll bring him in. >> Let's do it. >> The TVPN Ultrame. >> Made it. Alex, what's going on? How we doing, guys? >> Great to see you.
2:47:28
Great to see you as well. Excited to be back.
2:47:31
>> It's been a minute since we caught up.
2:47:33
Uh, >> almost a year, guys. Yeah, >> I know. I know.
2:47:35
Uh, what's, uh, give us the latest. >> Oh, man.
2:47:38
Well, we're excited to be here because we're announcing our $40 million series B. So, >> let's go. >> Let's go. [laughter] Let's go. >> Yeah.
2:47:46
It's been a it's it's it's been a crazy 12 months for us.
2:47:47
I mean we frankly if you if you just look at what's happening in our space it the the the the number of impersonations scams fraud uh has frankly exploded.
2:47:56
Um the the number of people that are tackling that in an effective way has not exploded.
2:48:03
Uh and so Outtake is increasingly like the one-stop shop for uh I'm proud to say not just uh enterprise but government as well at this point. >> Amazing. Yeah.
2:48:12
re reintroduce the company like what are because there's there's uh there's so many different problems that you're tackling and and I'd love to uh help people understand like how you guys are approaching it in different verticals and for different customer types. >> Yeah, 100%.
2:48:28
I I think the oneline abstract version of outtake uh is our job is to make fake things on the internet very expensive and make the real things very obvious.
2:48:37
And so what that means for enterprises or governments that we work with uh all these institutions have public surface areas where they represent themselves. Right?
2:48:49
If if you're if you're a commercial bank, you have a website, an app, ads you run in ad libraries, employees in public directories, all of these are surface areas that you use to communicate with the world. Um bad actors know that.
2:49:02
They try to man in the middle and effectively try to take over those communications, attack you, attack your employees, attack your customers.
2:49:10
Our job is to go find that fake uh fake content, remove it as fast as possible and make sure that you remain sort of a high digital trust institution.
2:49:19
>> Yeah, we we had Peter on from uh Multibbot yesterday and he's had so many issues in the last part of it.
2:49:24
He was like doing a really haphazard rebrand with not a lot of planning and had issues with I think his GitHub uh account getting taken over as well as an X account.
2:49:35
>> Yeah, like as soon as he stepped off the account handle, someone else sniped it >> using a bot. Yeah, using a bot.
2:49:39
Yeah, >> it was like superhuman speed.
2:49:41
I I I think that's you guys are touching on exactly the right thing there where it's like, you know, cyber used to be this thing where you're like prepping for the attack and you're like, "Okay, I know like as an institution I'm gonna I'm going to get popped, you know, once a year, once a quarter.
2:49:54
It's my job to sort of reduce the blast radius."
2:49:56
And so it was really about building these like insanely powerful walls, right?
2:50:01
That's not that's that's no longer the attacker uh/defense paradigm anymore.
2:50:06
Like you're just under attack at all times.
2:50:09
Like you are constantly under siege.
2:50:10
like you should pattern match more with uh a virtual drone swarm where yes, the bot will snipe your uh handle the minute that you let it go.
2:50:19
Uh it's not that someone's going to think about it threequarters later.
2:50:22
Uh and so that's kind of the fundamental shift in cyber right now where the economics of attack have sort of gone to zero.
2:50:28
Uh obviously uh defenders need to adapt.
2:50:34
uh talk about how you guys are because you guys have customers and you're kind of you're basically monitoring the situation for them on the internet.
2:50:40
Uh it's it's monitoring as a service.
2:50:43
Uh but also the like takeown element.
2:50:46
Talk about the kind of partnerships that you're needing to do your work with individual platforms.
2:50:52
Let's say somebody sets up uh a fake account for for Joe Rogan and they're selling like supplements through that account trying to make it look real using AI.
2:51:02
Joe Rogan, >> uh you can you can report these things. Sometimes there's a lag.
2:51:09
talk about like speeding up that process because part of part of what will make this what you said like expensive to be fake on the internet is if things get taken down far far faster which means like machine on one side coordinating with machine from the platform to make it so that you know you're not uh running an an AI ad of Joe Rogan for like 10 days before it gets taken down.
2:51:32
>> Yeah, you hit the nail on the head.
2:51:32
like the KPI that I think about as a northstar for outtake is how do I reduce the ROI for a digital criminal, right?
2:51:41
And and and you you kind of hit the nail on the head of the way you reduce uh that ROI is how long can their attack be active?
2:51:48
If it's if it's fleeting, then they put in a lot of effort to put something together uh or maybe not too much effort with Genai.
2:51:53
Uh but but if it's taken down before they get any reward, then then Outtake is doing a job.
2:51:59
Um, and and you touched on another really important thing, especially in a in a world of like vibe coding.
2:52:04
Something that I like to say a lot is is is Outtake as a company that you cannot vibe code, right?
2:52:07
Like you you can sure you can start to put together the capabilities to try to search a little bit.
2:52:12
Um uh but to be able to search in depth with highquality classification across all these platforms, you know, whether it's social media, clear web, dark web, you know, forums, Telegram groups, areas that are really hard to get into uh and actually extract what you need at scale and detect what you need at scale. Incredibly difficult.
2:52:28
But even more important and certainly not even frankly a thing that that software fully solves on its own is the capability to be a high trust partner to all these sort of uh counterparties, right?
2:52:40
right? where you just mentioned when we go and report something it's really important that whether that's a social media platform or a domain registar they know that outtake as a high trust institution has vetted that report is is high confidence by the way that we the
2:52:55
way we provide that confidence is we use our agents to basically go do cyber security investigations on absolutely every single thing we report we did by the way 20 million last year um and >> 20 million investigations you mean >> yeah that's like try to try to comprehend you know how many how many
2:53:13
humans that would have taken by the way and also to show you a sense for sort of scale and speed um of that 20 million about 80 to 90% of them happened in just Q4 for us right which you know for us ends this month and so that's like 18 million 17 million of those investigations are happening sort of
2:53:31
just in the last three to four months so maybe that gives a sense for like how much we're investigating how much more we're investigating sort of uh as a company at the moment And somebody might ask why why is this not happening at the platform level? Like why wouldn't Meta
2:53:44
Like why wouldn't Meta do this or Google do this or X or Tik Tok?
2:53:49
And >> uh maybe break down why you need a third party because of kind of the I can imagine there's a really big like incentive misalignment where like if you're if you're Joe Rogan, you don't want somebody popping up in an AI account selling supplements using your name and likeness.
2:54:05
But Meta is sitting there being like, "Well, we want more account growth and we want more ad spend on the platform."
2:54:10
Like, yes, this is a problem we're trying to handle, but it's not even top of the list, I imagine, because it's like driving usage.
2:54:16
It's driving new content, it's driving revenue.
2:54:21
>> I I I Yes, I think you're pushing at a really important point here.
2:54:23
I the, you know, now working with a lot of these platforms, I I think I have a lot more sympathy for them than I might have at the beginning.
2:54:31
I might have I might have said, "Yeah, I I you know, like there's not enough people that care about this.
2:54:34
I would actually tell you having worked with a lot of these teams, they they care deeply.
2:54:37
I think their problem um is is is is the scale at which they work, right?
2:54:42
So these platforms are trying to ensure and by the way I think rightly so ensure that they don't infringe on free speech um and they they they they don't unfairly punish people.
2:54:52
They don't put too many guard rails on on the one hand.
2:54:53
On the other hand, of course, they want to prevent all these problems that we're discussing.
2:54:58
discussing. And so outake has this unfair advantage frankly where we're not trying to do this at the scale of 3 billion 6 billion people at least not today yet right uh where what we're able to do is say hey we're working with uh the best-in-class sort of fortune 2000 institutions uh the federal government
2:55:15
state and local agencies um which is a much much smaller set of institutions and so our models are actually quite fine-tuned to say okay for this specific institution like what is real what is not by the way risk is different for all these institutions, not only between the institutions, but within the institution. Let's say you're talking
2:55:32
Let's say you're talking about again a commercial bank.
2:55:33
The way that it assesses risk on social platforms is quite different than the way it assesses it on on websites.
2:55:39
And so you almost want like distinct models that are fine-tuned for different surface areas.
2:55:43
All that to say, you want to be hyper specific on how you classify risk.
2:55:48
Um, and that's a that's a very very hard thing to do at the scale of these these massive platforms.
2:55:52
I think that's actually the fundamental issue.
2:55:55
It's it's really just a tech scale question for them.
2:55:56
What do you think might change on the consumer side?
2:55:58
I I saw an Instagram reel of someone highlighting uh just an AI video and he was sort of pushing for a digital ID or some integration with his social security number.
2:56:11
There's obviously the world coin world labs project eyeball scanning.
2:56:17
There's different things that you can build trust.
2:56:18
Jordy likes to if he finds an Instagram profile just scroll back and make sure that they were posting in 2017 before AI existed so that they could uh so so you can see okay well it looks like the same person probably a real person.
2:56:31
Um what what do you think is going to change on the consumer side of like proving your identity online?
2:56:37
>> Uh you you touched on a few really interesting things there.
2:56:39
So so one uh Jordy that that's great.
2:56:40
That's actually one of the signals we use Sure.
2:56:44
>> for our agents to sort of check like you know what is the history of this profile?
2:56:47
Um um the the you mentioned world.
2:56:50
So you know we we actually have a direct partnership with world ID.
2:56:53
Uh we actually had uh launched uh I would say a a research experiment project with them called verify where we took uh what they sort of built where as you pointed out um scan the iris give you a a unique way to prove that you are a a singular and unique human.
2:57:10
Um and and and and what we were really curious about is like what does that mean for obviously consumer um which by the way I think it means a lot of things.
2:57:19
They're doing a lot of incredible work with uh platforms like Tinder for example where it's like hey I want to make sure I date real humans.
2:57:26
Um though frankly you know side note it seems like AI dating is also doing quite well.
2:57:31
So so maybe there's a gen general market for that as well.
2:57:34
>> Our chat Ryan in our chat just said TBPN wasn't posting before AI. >> Not true.
2:57:38
Look at my YouTube channel. I started in 2020.
2:57:40
[laughter] My my favorite. >> Yeah. Yeah. Yeah. >> Okay. >> Um >> Yeah. Sorry.
2:57:47
>> All that to say, um those kinds of consumer credentials actually can and will bubble into enterprise workflows.
2:57:54
Um something that that Outtake did was say, okay, um you know, wouldn't it be powerful if if if you could project your security?
2:58:01
And and I'll I'll define what that means.
2:58:03
uh right now when you when when we interact with other institutions, you actually don't know if that institution, you know, has a great security protocol.
2:58:11
Um, one way they could prove that to you is every email they send, they would cryptographically sign, right?
2:58:17
Um, and so there would literally be in the header a signature that says like, hey, at the moment of send, I literally swiped my pass key that we all have on our computers.
2:58:24
Um, and and and it was physically me and and I can show you with a badge that that this was therefore definitively me, the human, etc. , etc.
2:58:32
And I'm projecting to you that I'm a high trust person to email with.
2:58:36
>> Um, and so something that that Outtake is very actively exploring when I talk about uh, you know, making real things obvious as part of our mission is exactly that layer.
2:58:45
It's like, okay, how do you go beyond just finding and removing things, but how do you help authenticate and prove to the world what is real?
2:58:52
is real? I I think frankly that's like when when when we think about our $40 million series B and and like why I think we're on this trajectory to be a public market company is it's it's the the the real prize the the the real prize for us is like how do you become a
2:59:06
trust layer for the internet like how do you say hey we've spent so much time mapping the landscape removing what's fake that actually we are the best source of truth on what's real right um that that's really what we think about >> amazing well I will >> who who did the who did the round Um, yeah. On a more serious note, uh, who did the
2:59:25
On a more serious note, uh, who did the round? No, I'm kidding.
2:59:27
Uh, very, very iconic group to get in the mix.
2:59:31
And I'm sure you'll be back. I I I don't know.
2:59:36
If I was a bet, if I was a betting man, I'd say a couple more times this year with the momentum that uh that you guys are on.
2:59:42
So, uh, congrats on the on the milestone and uh and uh yeah, thanks thanks for fighting the bad guys. >> We'll see you soon. >> Thanks for having us. you. Goodbye. App loving.
2:59:52
Profitable advertising made easy with Axon. ai.
2:59:55
You heard the Axon Claxon.
2:59:59
>> Get access to over 1 billion daily active users and grow your business today. >> That's right.
3:00:03
Up next, >> we have Mitchell from Phoenix.
3:00:05
We're going back to milk.
3:00:09
>> Do you know what this company does? Let's bring in Mitchell.
3:00:10
I'll have him explain it.
3:00:11
How is this company tied to milk?
3:00:13
You're the second milk related company of the day.
3:00:15
The flapping airplanes guy. Loosely. Loosely.
3:00:16
But please, I'll let you explain.
3:00:19
introduce yourself in the company, please. >> Yeah, of course. Second, milk related.
3:00:22
I don't even know what that is, but yeah, >> there was just an AI lab guest who was drinking milk and so I I made the Anyway, sorry. Continue. >> One of our products. No, awesome.
3:00:31
Well, nice to be on here.
3:00:33
So, yeah, of course, Mitch, one of the founders of the Phoenix.
3:00:36
We're um kind of using II like you said to help milk, but mainly to help dairy farmers manage their cows.
3:00:42
So, at the end of the world, make milk that that other gentleman gets to drink. >> Yeah.
3:00:47
How is the is the dairy farmer market extremely concentrated at this point?
3:00:52
Like are are you like almost proumer or is it like pure enterprise?
3:00:55
You get in like the three biggest dairy farm aggregators and then you're done.
3:01:01
>> Maybe not three, but it's absolutely much more consolidated since you know back when it used to be guys with a couple cows in the back of the yard.
3:01:06
No, our average dairy farm is 3,000 cows and doing tens of millions revenue. >> Wow.
3:01:11
>> It's really good business.
3:01:14
How do you actually plug in get data drive outcomes?
3:01:16
Like what what what are you tracking? Do they have ERPs?
3:01:18
How what's the actual data?
3:01:21
>> Yeah know it's really interesting as we've scaled dairy farms over the past like 30 years data collection has become a big part.
3:01:26
Like you walk on to one of these dairies there's computer vision happening there.
3:01:29
There's caller like all of the cows wear Fitbits.
3:01:31
I mean they're not made by Fitbit but same thing [clears throat] >> you know all that sort of.
3:01:34
So there's a hell of a lot of data collection being there um kind of up to date.
3:01:37
Most of it just goes to nothing.
3:01:39
So that's where we really come in using a lot of that data there to actually help with insights. >> Yeah. What Yeah.
3:01:44
Break down like I would love to just like the specifics of like what what the actual product does, how how it's driving results for for the farmers. >> Yeah, absolutely.
3:01:55
So when we go onto a dairy farm, we collect genetic information on every single cow.
3:01:58
So we literally whole genome sequence all 3,000 something cows on a dairy or however many there are.
3:02:05
>> We then combine that with what I call phenotypic data.
3:02:07
But that's the physical data, the computer vision, the Fitbit, the weather, how much milk every cow is making, how much she's eating, all of that data.
3:02:15
So we feed that into a big AI model which then makes predictions on the life of a cow.
3:02:19
So when she's a day old, I can predict with about 90% accuracy what her life is going to look like.
3:02:25
You know, how many whether she's going to get sick, when, how, how much milk she's going to make, is she going to get pregnant, when, how, that sort of stuff.
3:02:31
And we use those to just make the farms a lot more efficient as far as their day-to-day management.
3:02:35
And then what's the pricing model? Are you doing per cow?
3:02:40
>> Yeah, it's like per cow per year. >> Per cow per year. Okay. >> Yeah.
3:02:44
Instead of a seat, we got a cow. >> Yeah.
3:02:46
And then and then what what what interventions are farmers taking?
3:02:48
Do cows like trade hands in the secondary market.
3:02:53
If a cow is particularly valuable, >> they do.
3:02:56
Cows sell for millions of dollars. >> Millions of dollars. Yeah.
3:02:59
I think the the highest is now over tens of millions is the best cows out there. Yeah. Pretty good.
3:03:05
And and is that for like breeding purposes?
3:03:07
Somebody like >> elite line >> is particularly >> they really just make semen their whole life is pretty much it. >> Yes.
3:03:15
I mean what we really look like is actually kind of how it works.
3:03:17
When we go onto a dairy there are farm workers show up and ask the computer what should I do today and our computer will tell them.
3:03:25
>> So we are really managing the dayto-day on the entire farm just so much more.
3:03:26
I mean, nothing against our existing farmers, but you know, much more efficiently when you're able to look at both predict everything out with 90% accuracy and you're looking at every single piece of information.
3:03:38
>> It's dairy dairy super intelligence. It >> really is. It's here.
3:03:41
What uh how how big is it?
3:03:44
Like how many farms are you guys active on today?
3:03:46
Is is is the primary like bottleneck now?
3:03:50
Just like having enough people to >> to sell the product.
3:03:56
>> Uh well, actually I'm the person who sells.
3:03:57
So I'm literally every single farm we close, I'm the person who spoke to them, but supporting them, you know, is quite a lot.
3:04:02
So, um, obviously that doesn't scale forever either.
3:04:04
But yeah, so we're on close to half a million cows now.
3:04:08
So I mean that's about 5%, right?
3:04:12
>> Yeah, >> that's pretty good.
3:04:13
>> And we only launched in August.
3:04:13
So we have scaled to like I mean last year we did about two and a half million.
3:04:18
I did two and a half million last week.
3:04:21
>> Um, so >> yeah, we're going pretty fast as far as scaling.
3:04:24
And really what I need now is just people to join.
3:04:26
I mean, at the end of the day, it does still, even though it's an AI doing it, it requires >> How many dairy cows are there on Earth? >> Um, hold on.
3:04:35
On Earth, there's like 100 million.
3:04:36
In the US, there's like 10 million. >> Mhm. >> Okay. Okay. >> Job's not finished. >> Job's not finished.
3:04:41
>> But you raised some new money. Tell us about the round. >> Yeah.
3:04:44
So, we did our, you know, our seed round raised, you know, over 5 million.
3:04:48
We got some >> got some great investors that came in.
3:04:54
So, you know, led by initialized overwater, these sort of people who are really helping us scale up this mission and um really just making the world of agriculture much more efficient. >> Amazing. >> That's amazing.
3:05:03
I I like Yeah, being able to walk up to the computer on a farm and just say, "How do I tell me what to do?
3:05:13
>> Increase this cow's mouth milk yield."
3:05:16
[laughter] >> Computer, milk this cow.
3:05:16
It's basically what's happening.
3:05:18
[gasps] >> Thank you so much.
3:05:19
kill this cow, move this cow, milk this cow, treat this cow, all that. That's great.
3:05:23
Yeah, I can I can imagine the farmer of the future walking around with a headset on just like seeing, you know, little little AR popups together. >> Very very cyber punk.
3:05:31
Cyber punk farming is here. >> It is. It is.
3:05:34
Well, congratulations on all the progress.
3:05:36
Thanks so much for stopping by.
3:05:38
>> Great to meet you >> and we will talk to you later. >> Goodbye.
3:05:41
>> Let me tell you about Railway.
3:05:41
Railway simplifies software deployment.
3:05:44
Web app servers and databases run in one place with scaling, monitoring, and security built in.
3:05:51
>> And we have Gabe from Rogo.
3:05:51
And up next, [music] >> first time on the show with some amazing news. How you doing, Gabe? Welcome to the show. >> What's up, guys? How's it going? >> What's up?
3:06:02
It's >> We've been looking forward to this one for I feel like a year at this point.
3:06:05
>> Yeah, I'm shocked it's your first time.
3:06:07
Thank you so much for for taking the time.
3:06:08
Uh, since it is your first time, uh, please introduce yourself, the company, and the news. ST's great. I'm Gabe Single.
3:06:15
I'm the CEO and co-founder of Rogo.
3:06:17
Uh Rogo is a Genaii tool for front office investment bankers, investors, private equity professionals, you name it.
3:06:24
Uh the kind of mental comp folks have is Harvey for finance.
3:06:29
>> Uh we just raised a $75 million series C led by Capital >> and it's incredible.
3:06:38
Um what uh I want to get right into it. Do you agree with Pat? Is AGI here? >> More or less.
3:06:47
You know, [laughter] AI is smarter than a lot of people I know.
3:06:50
So, it depends on your definition of AGI. >> Okay.
3:06:53
>> Okay. Well then in the investment banking context uh a lot of the value that an investment banker brings is interpersonal negotiation contacts a rolodex understanding subtext what the motivations of a certain CEO are uh stuff that might not just be be
3:07:12
scrapable off the internet uh what what where where does rogo actually plug in what's left for the human being what's the future of investment banking look like >> look we want the mo the best deal makers to be more productive. And there are
3:07:24
And there are guys right now who are slinging phone calls all day long, 5 minute, 5 minute, 5 minute, 5 minute.
3:07:30
Someone like a Blair Efron at Centerview.
3:07:34
>> You know, that guy is so productive because he has a team of junior bankers underneath him that know what they're doing.
3:07:40
>> Yeah, >> there should be a lot more dealmakers.
3:07:42
There should be folks who are able to go out, advise companies, uh help with M&A restructuring, etc.
3:07:46
uh and focus on the human part of the work, the connectivity, the reading folks, the negoti negotiation rather than, you know, spending time building powerpoints, spending time, you know, benchmarking comps, all that kind of stuff.
3:08:00
>> How uh how is how has the product evolved over the last year specifically with new model advancements, reasoning, etc.
3:08:08
Uh we uh I I have a lawyer buddy and and just like the difference in how he perceived Harvey 12 months ago versus today is like just totally night and day.
3:08:19
Like back then he was like, "Yeah, like it's it's it's cool."
3:08:21
And now he's like totally oneshot by it.
3:08:24
And I'm I'm imagining that uh something similar is probably transpired with you guys as well.
3:08:32
Yeah, I think especially in the enterprise there was a little bit of that first mover's disadvantage early on where you were kind of bottlenecked by just the model quality uh and everyone had ideas for how they wanted these tools to transform legal work, finance work, healthcare, etc.
3:08:46
But if the models were incapable of doing it or it was just it wasn't quite at the threshold where you could do serious work, I think you know people felt a little spurned by those products.
3:08:56
And I mean, if I look at Rogo today compared to two years ago, it's like two years ago, we were the most incompetent intern you would ever hire.
3:09:03
And if you had hired a human intern like that, you would have fired them.
3:09:07
>> The benefit is Rogo gets a lot better, a lot faster, and then it stays that way.
3:09:11
And then it's going to be a lot cheaper than that human intern.
3:09:13
But the products changed enormously, both in actual capabilities and ability to move across what a banker's doing, PowerPoint, Excel, file drives, sharepoints, data rooms, etc.
3:09:23
, but also integrations into the back office, right?
3:09:26
Like people don't realize, but to be an investment bank, you not just have to be, you know, creating Excel models, calling people, but who's thinking about the conflict checks, who's thinking about all of these kind of back and middle office tasks that help an investment bank run.
3:09:39
And over the past two years, we've really embedded across a lot of those workflows, too.
3:09:43
>> Can you share a little bit about how uh how your customers are actually using Rogo specifically?
3:09:48
like we saw this massive boom with Claudebot, now Moltbot.
3:09:52
And it felt like a lot of people were just excited about Claude code over WhatsApp or Telegram, right?
3:09:59
And it's like these models existed.
3:09:59
We didn't get a new frontier model.
3:10:01
We got a new way for a lot of people just to interact with these models.
3:10:05
And I'm wondering if you're seeing a shift from the, you know, the laptop crew maybe doing more of those compliance checks, more of those, hey, let's pull some comps or pull some data together over their phone specifically. Yeah.
3:10:20
So, we are about to roll out our mobile app, which we have been kind of long in the tooth doing.
3:10:23
Um, but I hope that that, you know, we start to see more of the phone crew.
3:10:26
The reality is the most important folks in in our ICP, uh, they're just on their iPads and phones all day anyway.
3:10:32
You know, when I was a banker, I didn't even realize my MDs had computers.
3:10:35
They were, you know, never had them, never in Excel, that kind of thing.
3:10:39
Um, I think you'll see the UX's of these agent tools start to evolve into like some sort of ability to manage, you know, your agents and tasks and the kind of UX around like your army of bankers or lawyers or whatever it is.
3:10:55
>> Um, yeah, I'm I'm I'm pretty excited about that.
3:10:58
>> What about other integrations?
3:10:58
Um, are you seeing folks like is there is there uh, you know, demand for like the the Slack type of integration?
3:11:06
we're in a we're in a war room on a particular deal and we're adding rogo to pull stuff in into that context >> except it's it's got to be teams if we're talking about investment banking.
3:11:19
Um but yeah, all of those things, right?
3:11:21
Like you should have a data room that is being populated by an agent and a data room that's being diligent by an agent.
3:11:26
Uh it should be completely integrated in all the places where your data already lives.
3:11:30
I think the complexity for our business is, you know, when you have a kind of a cloud bot or claude code running around on a desktop unconstrained, there's security risks.
3:11:40
You know, I was using cloud code to create a fake data room that we would then demo in the rogo product.
3:11:44
And as part of that, I gathered every prior fundraising deck we had, all of our financials, cap table, and I asked Claude to organize it, and it deleted all the files.
3:11:53
And I was like, can you [laughter] un can you please undo that?
3:11:56
And it was like, no, I'm sorry, that was that was irreoverable.
3:11:58
It's like if I did that with one of my clients, I would not, you know, we would not be raising this round.
3:12:03
And so there's a lot of of nuance at figuring out how to take these agents, which are most powerful when they are unconstrained and do have access to your files and your desktops and your systems and making sure they can be deployed in a way that's thoughtful and secure and so on. >> Yeah. Yeah, that makes sense.
3:12:18
How how are you thinking about integrating rogo with like legal like legal workflows and and just like like at what point does a lot of this stuff really converge?
3:12:28
Uh it feels like you know there's kind of an old an old par like kind of paradigm around these the way you know a company gets uh bought or sold that will evolve. >> Yeah.
3:12:45
I mean, especially in places like private credit, there there's there's a blurry overlap between what is, you know, legal work and due diligence versus investment work and due diligence.
3:12:55
I would say when I look at some of the like prior generation comps for me in this space, Thompson Reuters, Faxet, Bloomberg, etc.
3:13:00
Um, they actually really subdivided the the legal verticals and products from the finance.
3:13:07
I mean, Thompson Reuters actually sold off all their finance assets to Refinitiv and into Lseg.
3:13:12
And I think it was partly because it was actually, you know, there there's less synergies often than people expect.
3:13:18
Um I think if we're just talking about private equity, private credit, yes, a lot of synergies between those types of workflows, but like you know, a lot of what investment bankers are doing, um or you know, the the top private equity folks, top public markets investors, they're kind of outsourcing all the legal work to someone else.
3:13:35
Like that's not even under their umbrella.
3:13:39
>> Uh what about headcount planning?
3:13:39
How are how's the industry kind of thinking?
3:13:45
It's probably one of the more awkward conversations uh for for a lot of people involved.
3:13:49
But uh >> I mean it's not it's not taking teachers jobs, right?
3:13:55
Like we're talking about investment bankers and investment bankers are going to get more productive and [clears throat] so firms are excited, people are excited, bankers are excited.
3:14:03
If you're a junior banker, you will be generating fees yourself earlier in your career rather than later.
3:14:07
If you are a bank, you're going to expect higher fees out of your existing bankers with fewer resources, but it's going to change the trajectory of of what it means to work in finance.
3:14:17
Uh, and you'll go from being the automation piece of the bank where you're just turning out materials to having to think about, yeah, how do you go out and do deals on your own?
3:14:27
>> Yeah, makes a ton of sense.
3:14:29
>> Do you think uh like do you think we'll see an acceleration in overall deal making?
3:14:34
Like are there are there opport are there are there deals how many deals don't get done just because it's going to be such a huge headache?
3:14:40
Like >> there's a lot of deals that don't get done because it's such a headache.
3:14:45
I mean, especially in the private markets, like think about all those mom and shop businesses throughout the US or even, you know, businesses in emerging markets that can't pay Goldman Sachs $10 million to run a process, can't pay JPM $10 million to restructure.
3:14:59
Especially in the US where all of these assets have been getting picked up by private equity firms that just have sourcing arms that call, you know, your local roofing business in Minnesota.
3:15:09
You know, now those businesses should be able to get banking services themselves, actually run a fair auction and get a better price, and it should result in a more liquid, more transparent, more efficient private market. >> Go to deal sleds. What you got for us?
3:15:24
Recommendations for upand cominging bankers, the next generation.
3:15:26
What are they putting on their feet?
3:15:33
>> You know, I have one pair of deal sleds, and it's the ones my parents got me when I got my job as an investment banker at Lazard.
3:15:39
And they're Ferragamo loafers.
3:15:42
Those are the ones I wear to this day when I have to go into, you know, whatever investment bank we're >> Somebody should make some shoes that are just like Mac minis.
3:15:49
You're just [laughter] walking around, you know, walking around the office, >> Silicon Valley Bank.
3:15:54
>> Really signaling signaling like, you know, I'm I'm with it. >> That's a lot of fun. Yeah.
3:15:57
Mean, >> what do you what's the biggest bottleneck right now?
3:16:00
I'm assuming you're you're going to hire a bunch of people from this round. Where what's what?
3:16:05
Talk about the use of funds, all that.
3:16:08
We're 100 people today um primarily in New York.
3:16:11
We're opening a London office and then probably AP pack later in the year trying to get to around 200 250 people equally split across you know product and engineering uh and go to market and most of our go to market post sales what we call kind of like forward deployed banking is ex finance ex bankers ex investors. >> Mhm. Very cool.
3:16:30
Where's the name come from Rogo?
3:16:35
Our lawyers, our initial name was Athena and our lawyers were like, "You can't do that.
3:16:39
There's Gabe, there's probably a thousand products called Athena."
3:16:40
So, John, my co-founder, and I spent months trying to come up with a name.
3:16:44
Um, and I'll, you know, at some point we're just looking up old Latin words.
3:16:49
And Rogo means I ask in Latin.
3:16:51
And so, we thought that was a good one. >> I love that. That's great.
3:16:54
Well, thanks so much for coming on. Great. >> Yeah. Great.
3:16:57
Great to finally meet and uh, congrats to the whole team. >> Congratulations.
3:17:00
Have a great rest of your day. We'll talk soon. Goodbye.
3:17:05
And our last guest of the lightning round, Sierra Peterson from Voyager Ventures >> Energy Time Reream Waiting Room. How are you doing?
3:17:17
>> Hey guys, delighted to be here. How are you? >> We're great. >> Great to have you.
3:17:21
Uh first time on the show.
3:17:23
Would love uh introduction on yourself and the firm and then we can get into the news. >> You bet. So I'm Sierra Peterson.
3:17:28
I'm one of the two founders at Voyager Ventures.
3:17:33
We are an early stage venture capital firm investing in these foundational technologies that power civilization.
3:17:37
So this energy, materials production, advanced comput, physical AI.
3:17:44
These are the biggest markets in the world and the ones that are foundational to ongoing growth in the long term and competitive advantage in the near term.
3:17:51
So I have been working in energy for my entire career.
3:17:53
Got my start 21 years ago at the International Energy Agency. >> Overnight success. >> Exactly.
3:18:01
[laughter] long time coming.
3:18:03
Uh and I mean that's really the story of our backgrounds.
3:18:08
Myself and my co-founder at Oyer Sarah have been working in energy and industrial uh modernization and climate stabilization now for decades and between us we've built five climate tech and energy companies.
3:18:22
Uh we've been active in policy making at the International Energy Agency as I mentioned and also in the Obama office of energy and climate change back when we had one.
3:18:32
Uh and we built Yeah, exactly.
3:18:35
Uh we built companies that um were responsible for financing more than $3 billion worth of distributed energy assets. So, a lot of solar.
3:18:42
Um, we've been active in the research programs at Harvard and MIT, advancing everything from energy market research, which was my graduate focus, and Sarah's at MIT in industrial biotech.
3:18:55
Uh, and we've been investing now for 10 years plus.
3:18:59
Actually got started as an angel investor in 2015 in energy tech.
3:19:00
Those companies did well.
3:19:03
And so, we teamed up to launch Voyager in 2021 and raised a hund00 million debut fund.
3:19:08
U, we've recently John's going for the gong.
3:19:13
>> How much did you raise recently?
3:19:15
>> Hey, no, that's not the big the the big news announced at 275 on two.
3:19:16
So, >> we were waiting for that. Uh, very cool.
3:19:27
I want to talk about energy because I feel like your your timing here is great.
3:19:31
I feel like people in uh are spending way more, you know, uh it's gone mainstream uh both at the national level and politics.
3:19:43
>> Just in [clears throat] the last few weeks, we've had uh Scott Nolan on to talk about uh you know nuclear uh uh fuel production and then a separate company that turns that actually into fuel pellet, tricofuel and like there's a there's a company in every part of the stack.
3:19:57
uh are there any particular subcategories that you're tracking that you're particularly interested in?
3:20:01
And then I want to talk about the financing of all this as well, but uh what what what do you think is the most like breakout success in energy these days? >> Oh yeah.
3:20:10
I mean we're energy maximalists.
3:20:11
I have been working in energy for my entire career. Yeah.
3:20:13
And when I and when you think about what's truly foundational to powering civilization, it's energy.
3:20:18
That is the foundation of all of it.
3:20:21
And so, you know, we it's actually a really exciting time to be investing given advances particularly in electrification.
3:20:28
I mean, we're seeing the cost of solar drop so fast.
3:20:35
You have to look at it at a log scale. It's amazing.
3:20:37
And what that means in terms of, you know, the distribution of energy at a worldwide scale is incredible.
3:20:42
H and for creating, you know, a truly global and resilient energy system.
3:20:47
Um that's you know enables near-term stability for users of electricity and long-term prosperity at a global scale.
3:20:55
That's like that's this is a technology that will power the next era of progress and beyond.
3:21:00
Um so electricity massive advances on edification massive right now.
3:21:07
>> Are you optimistic about American solar?
3:21:09
It feels like the in terms of solar panel construction and manufacturing China is doing so well that it's been really hard to get a foothold.
3:21:16
really hard to get a foothold. We we we've talked to some teams that have you know taken Chinese-built assets and >> T1 energy >> T1 energy is interesting but uh I haven't seen as much as much excitement from the startup world around really
3:21:31
scaling solar but then you have you know huge voices like Elon is a solar maxi there's lots of people that are very uh very you know uh just uh beating the drum on like the fundamental physics of the sun is so much energy we're going captured at some point. So, it feels
3:21:46
So, it feels inevitable, but it doesn't feel like there's a boom or narrative being driven. But what are you seeing?
3:21:54
>> Oh, it's totally inevitable.
3:21:54
You can't argue with the numbers.
3:21:56
I mean, 90% of installed energy generation for the last year that we have record is 2024 is renewables.
3:22:04
>> That [clears throat] is at a global scale.
3:22:05
That is certainly true in the United States as well.
3:22:06
I think it's just under 90%.
3:22:08
And [clears throat] I mean, that's regardless of political whim.
3:22:12
That's regardless uh of um you know overall organizing policy approach.
3:22:16
Um that's just simply because solar is better that it's cheaper, it's better performing.
3:22:23
It you know it enables a true distribution of electricity like we've never seen before and coupled with advances in batteries. That's the unlock.
3:22:31
You can you really create a transactive synergy system which we've never seen before and really will be fundamental to power and growth for decades ahead.
3:22:38
decades ahead. So huge yeah huge believer in solar's potential now um and you know in combination with electrification both of [clears throat] energy generation uh also um in end use applications I mean electricity now is the you know the most frequent way in
3:22:55
which anybody around the world accesses power that's exciting it means that power is programmable um and it enables a new era of machines to be made which is extraordinary in terms of you know extending human capability abilities into ways that we've never had before. Uh when you couple that with automation
3:23:12
Uh when you couple that with automation um and artificial intelligence, like we're just seeing the foundations of so much innovation right now to really one boost systemic stability in a rising vulnerability and uh volatility market.
3:23:29
You have fraying trade relationships, geopolitical tensions, physical risks that are intensive given climate change.
3:23:36
And when you can control your inputs in terms of electricity, in terms of your means of production, and you can take advantage of advances in advanced manufacturing to create better products, you can really control your own destiny. That's exciting. >> Yeah.
3:23:50
How are you thinking about uh financing hard tech companies, energy companies, these like capital intensive companies at the earlier stage?
3:23:56
It feels like uh more and more founders are going to need to get familiar with venture debt.
3:24:03
It's been sort of a cautionary tale from the software startup era. I'll venture debt.
3:24:09
That's what will burn you.
3:24:11
That's what will really put you in danger uh if you're if you misplan and then you you get over your skis.
3:24:16
But how do you think about the role of venture debt or debt instruments or more uh more complicated financing schemes for some of these companies?
3:24:25
>> Yeah, it's a great question.
3:24:25
I mean, I think one of the things that we're quite excited about is the sophistication uh of the financing instruments that are available to early stage startups.
3:24:31
I mean, it used to be that you had to burn super high cost of capital venture or private equity on assets and that didn't make sense for anybody.
3:24:40
um you know now we're seeing really tailored um and interested um bespoke nondilutive financing um you know all types and I think it's a recognition of the scale of these markets and the durability that they have and you know this um longived um and reliable yield.
3:24:59
I think we're also seeing though and this is something that's been um you know a recent development ongoing sovereign interests and recognition in these technologies as being tools of the national interest.
3:25:13
You know 44% of our own portfolio um is either in partnership or in active conversations to sign a deal with the US government.
3:25:22
>> Uh that's extraordinary.
3:25:22
I mean, you know, you see Inqel last year, I think, became one of our our second most frequent co-investor.
3:25:29
Um, and >> it used to be completely unthinkable.
3:25:32
It was like you'd never go to the government.
3:25:35
They'll never buy anything.
3:25:36
Like Palunteer was this weird company that was like off in the dark just doing one deal and they were the only ones and then Anderol sort of did it again and then all of a sudden it was like every company has some sort of deal with the government.
3:25:46
really change everything >> because when you think about what are the inputs and tools of sovereign advantage of competitive advantage, it's controlling your destiny in terms of energy, controlling your destiny in terms of supply chains for inputs to manufacturing, industrial capacity and then increasingly compute capacity.
3:26:04
It's a national asset and all [clears throat] of that hyperscaler stack that all relies on electricity. >> Yeah.
3:26:12
What about uh like what what are the prospects for making uh bringing the solar supply chain uh to the United States?
3:26:22
We have like earlier we mentioned T1 energy which was kind of a weird >> setup uh in the way that it was created and and now operated by a new company but uh can we I think one of the reasons that people that that maybe Americans and and uh the tech industry isn't so excited about solar at this moment or at least broadly is that we're not making the panels here.
3:26:43
It doesn't feel like uh something that uh is that we really own. >> Yeah.
3:26:49
I I mean I I think I don't know that we're not excited about solar when you you look at the solar deployment.
3:26:55
Well, I'm just saying like the the attention is almost entirely on nuclear >> like nuclear >> like that now >> like I'm just saying from when you when you would like a word but >> yeah sure but think about [laughter] I I would just say the number of entrepreneurs that have come on the show this year that are talking about nuclear
3:27:12
the the nuclear deals are are uh and and I'm not saying it's like warranted but I'm just saying like nuclear is is certainly uh >> well it feels like at least to me and you can give your feedback back on this take, but um it feels like nuclear was uh was potentially economically valuable viable. The the technology worked, but
3:27:32
The the technology worked, but there was a regulatory blocker in place that made it very very difficult to actually just stand up the technology.
3:27:41
With solar, there's no that that regulatory blocker doesn't exist.
3:27:44
It's more of a market force of China manufactures them really really cheaply and like it's the DJI drone question versus GoPro like where like we just need a ton of people, a ton of manufacturing capacity, a ton of CNC machines.
3:27:59
We need like a massive facility and that's more of a capital formation and then are you willing to go up against the Chinese manufacturing engine? That's >> so critical.
3:28:09
My point is just that energy is critical solar energy >> it still feels blocked. pretty big part of it.
3:28:16
So I would imagine we want to make as many of these as we can in the US.
3:28:20
>> So I was curious like >> update on the on the effort there.
3:28:21
I know you've uh invested in in it and uh curious. >> Yeah. >> Yeah.
3:28:27
I mean I think we're seeing a real reexamination of global supply chains uh for critical uh aspects of ongoing sovereign advantage and industrial >> competitiveness.
3:28:37
uh solar being, you know, one, critical minerals another, uh access to novel battery technologies.
3:28:42
Um these are all aspects of an overall industrial policy that's, you know, increasingly favoring uh domestic production and control of energy producing assets and manufacturing inputs.
3:28:56
Um I think you know to your earlier point like any technology is going to need to clear the global clearing price uh and the the massive reduction in cost.
3:29:09
It's been extraordinary for solar production for panel production.
3:29:13
Um it's you can't get much cheaper.
3:29:18
[laughter] >> There's kind of nowhere to go. >> That's exciting.
3:29:22
I mean, think about that as an input to other industrial processes.
3:29:26
That's I mean, it's truly extraordinary opportunity to have abundant energy uh that doesn't rely on, you know, fossil fuel supply chains that's truly distributed and can be programmable.
3:29:38
I mean, you couple it with energy storage, I mean, you just have a whole new paradigm in abundance for electricity. >> Be exciting. >> Very exciting. >> I Yeah.
3:29:46
What uh you guys se uh >> stage agnostic preede?
3:29:53
>> Oh, we're early stage early. >> Yep. Yeah.
3:29:55
So, we raised 275 million.
3:29:55
H it's more seed series A series 275 million for uh seed and series A.
3:30:01
We went out to market with a 250 target.
3:30:04
We were overs subscribed on that. I had a 90% first close.
3:30:08
There was a lot of >> Yeah. Thanks, guys. >> Amazing. >> It's been great.
3:30:14
>> Well, congratulations. >> Well, excited.
3:30:15
I'm sure I'm sure a bunch of your companies will be on and congrats to the whole team.
3:30:20
>> Yeah, good to meet you. >> Likewise. Talk energy. >> Thanks. >> Talk soon. Cheers.
3:30:27
>> Um, did the Klein team just join Codeex Open AI?
3:30:31
Jason Lou asked the question. >> Yeah.
3:30:35
So, I think a lot of the team members did.
3:30:36
Asha, remember there was a whole controversy about >> you know the smell. Imagine the smell. >> Oh, yeah. Oh, he was declined. >> Yeah. Yeah. So he moved over.
3:30:44
I think a number of the of the like lead engineers there moved over. >> Yeah. Yeah. Yeah.
3:30:50
>> Will Brown says it's a very rare reverse wind surf >> and said well they certainly didn't decline >> decline the uh the ask. Yeah.
3:30:59
interesting to see these uh the you know the the the really the really headline grabbing um the uh the really headline and grabbing AI trade war narrative has sort of calmed down but everyone's hiring, everyone's poaching, everyone's trying to build their teams.
3:31:15
>> It's a war working through >> uh anyway uh there are >> uh there's a post here somebody is calling [laughter] out Augustus to Rico.
3:31:27
They said he has no couscous. >> Greed. >> Greed.
3:31:30
[laughter] >> He I to be fair, I've never seen the man eat couscous.
3:31:34
He might not have any couscous.
3:31:37
I think they were trying to say conscience.
3:31:40
Uh but yes, Augustus has has stumbled into such a funny life.
3:31:43
[laughter] I'm really He's uh Yeah, he's he's he's grinding through it.
3:31:51
>> Picked uh something more controversial than pretty much anything else.
3:31:54
Yeah, but he's he's >> seemingly more it's somehow more controversial than defense. >> Yeah. Yeah. It's just it's just odd.
3:32:02
It it just triggers all sorts of people.
3:32:05
I wonder um I feel like the the fact that he's able to go on podcasts and communicate so clearly.
3:32:10
He's one of the Joe Rogan CEOs in your parlance like uh that's allowed him to get through it.
3:32:16
It feels like if he were on his back foot and not able to just say, "Hey, I'll talk to anyone for an hour. I will just talk talk."
3:32:24
that uh if there was a negative news cycle about him, it would be a m much harder situation. But yeah, I don't know.
3:32:30
>> Uh >> last post for the day. >> Okay.
3:32:33
>> Emily Sunberg went mega viral. 62,000 likes.
3:32:37
>> I did not see the likes on this.
3:32:37
I didn't realize how big >> it says the logo for Sydney Sweeny's lingerie line looks like it's for a salad dressing company that launched in 2019.
3:32:48
>> Sir, >> this is for Siren. >> Oh, Siren. >> Siren. >> That's wild.
3:32:51
And it does it does give off a little bit of Sweet Green.
3:32:54
So >> Sweet Green launched way before 2019. >> Wait. So Oh. Oh, wait. Wait.
3:32:58
Is that the company that she's referring to?
3:33:02
What is she actually referring to?
3:33:03
Because I saw Dirt quote tweeted it.
3:33:04
Uh, and that's a media and technology company with a similar logo.
3:33:10
And Dirt says just says, "Hey."
3:33:13
And got 66,000 likes as well.
3:33:13
And Emily says, "Bro, but oh, I guess it's Sweet Green that they're that she's actually referring to.
3:33:23
>> Salad dressing company.
3:33:23
Isn't Sweet Green like a salad company?" >> Yeah.
3:33:26
I mean, she's just she's just she's just posted.
3:33:28
Alex Conrad says, "Someone blended Cava and Graza branding in Chad does have a little Graza in there."
3:33:35
>> But >> I think this I think this company will do very well.
3:33:38
Yeah, I think Sydney's like one of the most commercial people in Hollywood, right?
3:33:44
She's like just leaning in.
3:33:46
She's got >> uh a you know somewhat short moment in time to create a lot of value and and uh certainly a lot of the brand activations that she's done have broken through. So >> I'm bullish. I'm bullish.
3:34:00
>> To go back to the Clawbot thing like like this siren clearly exists in a very different part of the economy as Sweet Green.
3:34:08
Like sweet green I think of is like you walk down the street you see it.
3:34:11
Maybe if they start opening a ton of retail stores but like in terms of >> flashbang.
3:34:14
[laughter] >> Are you flashbanging me?
3:34:16
Why are you flashbanging me?
3:34:19
>> Because the chat earlier was going crazy.
3:34:20
They were trying to throw me off. It was insane.
3:34:23
>> I was barely hanging on. I was >> It was brutal.
3:34:27
>> Everyone loves uh everyone loves the flashbang.
3:34:29
Uh fun fact, Sydney Sweeney is launching her lingerie brand on, you guessed it, Shopify. Harley, that's right. That's the story. >> Fantastic news.
3:34:38
Uh, well, the last thing before you plant the bomb.
3:34:41
Uh, we have a new ad read for Railway. I already did it.
3:34:46
I'm going to do another one. Railway.
3:34:47
Railway is the all- inone intelligent cloud provider.
3:34:50
Use your favorite agent to deploy web apps, servers, databases, and more.
3:34:54
While Railway automatically takes care of scaling, monitoring, and security.
3:35:00
[applause] >> Well said, John. Uh, we'll close it out.
3:35:04
Nathan uh is highlighting Bloomberg article the the title of the article Nvidia lacks clear successor for superstar CEO who built company and Nathan of course says man is barely getting started >> and I agree who's who's asking who's
3:35:22
asking this question this is driven the stock 10x and it's the biggest company in the world and they haven't [laughter] like they have massive margins of cash flow like there's been no mistakes like literally like nothing 62 years He's got another 30. He's got another 30. He's He's got another 30.
3:35:34
He's going to be riding Nvidia straight into the singularity and we will be too here on TVPN. Last post. Last post. Plant the bomb.
3:35:45
>> Eric uh Sufort says >> Pinterest dropped nearly 10% yesterday on news that the company is cutting roughly 15% of its workforce to focus on AI.
3:35:56
Eric, I believe, is coming on the show tomorrow. >> Yeah. Yeah. Yeah.
3:35:59
really like talking with him about all things.
3:36:03
>> We're going to dive into this. We'll dive into GPT ads.
3:36:07
>> Um, pretty remarkable.
3:36:07
I wonder where Pinterest will land.
3:36:09
There's been rumors or or uh, you know, pitches for an acquisition.
3:36:13
Leave us five stars on Apple Podcast and Spotify. >> We love you.
3:36:17
>> We'll see you tomorrow tpn. com for the newsletter. >> Nice work, brothers.