0:00
Heat. Hey, Heat.
Search this video
0:09
[Music] Hey, [Music] hey, hey. [Music] Heat. Heat. [Music] Hey, hey, hey. [Music] [Music] Heat. Heat.
1:46
[Music] [Music] [Music] Heat. Heat. [Music] Heat. Heat. [Music] [Music] Heat. Heat.
2:50
[Music] [Music] We came to this world.
3:19
[Applause] [Music] to shape our future.
3:36
We came to this world to reach the stars to shape our future.
3:49
[Music] We came to feel the new Yeah.
4:04
[Music] [Music] Let me see. [Music] Team Deathmatch.
4:48
Delta team, you are clear to engage.
4:52
[Music] You're watching You're watching TVPN.
5:02
Today is Tuesday, July 1st, 2025.
5:02
We are live from the TVPN Ultra Dome, the temple of technology, the fortress of finance, the capital of capital.
5:10
We have the current thing hot off the presses. It's Siri V2. Apple is making moves. Apple's making moves. They are weighing. They are listening. They're listening.
5:23
They're weighing using anthropic or open AI to power Siri in a major reversal.
5:29
Everyone thought they were going to do it in house.
5:30
Can you imagine the actual like customer feedback that they get on the Siri product?
5:34
It's like people try to get Siri to do a simple thing.
5:36
It fails and they're just hearing somebody swearing or something while they turn it off.
5:42
I hadn't even thought about that.
5:45
Like, we've heard you loud and clear. We've heard you loud.
5:46
We've heard you loud and clear. Yeah, you guys are Matt. I don't even know.
5:50
I don't even know. Is that really I feel like people the expectations for Siri have also been low because it was introduced a decade ago and it was like the the thing that sets the timer and the thing that you ask to get the weather and that's it and so people haven't been asking for super
6:05
intelligence or chat the chat like experience and so or even a companion certainly not yeah c certainly not you don't hear about people falling in love with with Siri it was kind of pitched like that though back in the day it was very much like Siri we're giving it a human's name because it will be as integrated into your life as a human. And that was the original pitch. Now
6:22
And that was the original pitch.
6:22
Now it's possible, but not possible with the team they have on staff.
6:26
So they're partnering with Anthropic or OpenAI. Bidding war. Bit of a bidding war. Yeah.
6:33
And uh Apple allegedly has been bulking at some of the prices being thrown out. Okay.
6:41
Anthropic wants how much?
6:41
Billions a year and and ramping up over time. Yeah.
6:46
What's interesting is that like it could go the other way, right?
6:48
We're like we're like anthropic or open AI.
6:50
I could pay Apple like Google pays Apple is valuable.
6:54
Knowing knowing Tim Yeah.
6:56
He'll figure out a way to make it a profit center. I would think so. Yeah.
6:58
I mean the the the data but the issue is there's real cost to running these products for now. Yeah.
7:03
Like I mean it it seems like the cost will come way down.
7:08
And I mean I I believe that the current Cha GPT OpenAI Apple deal is revenue neutral for both.
7:15
I don't think there's any money changing hands, but it's also kind of a mess of a product because you have to say, "Hey Siri, go to chat GPT and look this up."
7:23
And then it pops up a prompt that says, "Would you like to use chat GPT even though you just said that and I don't understand that what you said?"
7:32
And then you have to click yes.
7:33
It's like come on Apple forward progress. The markets like it. It's up 2%.
7:38
Mark German has the story in Bloomberg.
7:42
We will be diving into that.
7:45
said it's sideling its own in-house models and potentially a blockbuster move aimed at turning around its AI effort.
7:51
We should go to the Mag Seven uh market caps to track the horse race of the biggest tech companies in the world.
7:59
Let's pull them up on the screen and see where people are tracking.
8:02
So, Apple's in third at 3. 1 trillion.
8:04
Good to see that they're over three trillion. Stocks up 2% today.
8:08
Nvidia and Microsoft Meta is climbing as well.
8:10
And one of these companies is not like the other.
8:13
Meta does not is measured in Tesla.
8:15
Oh, measured in in the poultry billions.
8:20
Congratulations to 2% up. You love to see it. They're up 2%.
8:24
Um, yeah, Meta being in sixth.
8:29
Doesn't feel like a sixth company to me based on the team and the and the moves they've made this week, which we will continue to dive into.
8:37
We have Swix on the show today to break down some of the AI talent moves.
8:42
We have a bunch of other folks coming on the show today. Yep. Uh, talk about that.
8:45
Anyway, other news, uh, the Senate passes Trump's mega bill, the big beautiful bill, after an all night session. They went all night. I like that.
8:54
And, uh, it's got to get through the House still, but uh, Elon is fuming.
9:00
Play the Ashton Hall sound effect for coming for the bill.
9:02
It's absolute war on the timeline.
9:05
the timelines in turmoil because Elon Musk said, "Anyone who campaigned on the promise of reducing spending but continues to vote on the biggest debt sealing increase in history will see their face on this poster in the primary next year." Pinocchio mode.
9:20
Liar voted to increase America's debt.
9:23
It's actually never been easier to make your enemies look like Pinocchio. Yeah.
9:27
Wait, but but I feel like if you put like I don't know I I don't even know who voted for like let's just say like Ted Cruz or something.
9:33
If you put like Ted Cruz's face on here, like are you going to lose the Pinocchioess?
9:38
Like I feel like we'll find a way.
9:41
I feel like it could be a little bit a little bit rough. Uh anyway, Elon's upset.
9:44
He's trying to start a new party. Uh he the America Party.
9:48
He said vox popularuli vox de he wants to start the American party and he is pissed because poly market yes uh is showing that the uh is expecting the reconciliation bill the big beautiful bill to pass uh by 70% by July 5th and 62% oh yeah 62% by July 4th and almost certainly by the end of the month.
10:11
So this will be the one to watch.
10:13
They're voting I guess the first vote is tomorrow. Yeah. We'll be following that. Yeah. Be interesting. Lots of pork.
10:19
But if you're AGI pilled, it doesn't matter. Yeah.
10:24
Uh we are going to be printing at 10% GDP, folks, in just just a few thousand days when super intelligence arrives.
10:31
These these debts, we're not days away.
10:34
Many pe many people are saying, if you're if you're against this bill, you're just telling yourself about your AGI timelines.
10:40
You're like, I don't really think super intelligence is coming.
10:43
coming. Yeah, it it is um uh I I think Teal was pushing uh generally pushing on Elon a bit in the interview he did last week around like okay so are are humanoids humanoids are going to be important obviously making them are they going to impact the economy are they going to accelerate
11:03
growth um you know this is the post 1970 stagnation you know we've been growing at 2% Sachin Nadella is asking the question of will AI accelerate GDP growth because if if GDP growth accelerates a lot It changes the dynamic around everything like not just this
11:20
debt which we've been kind of lightly joking about but also to Elon's credit he's also said when when he initially exited Doge and just stopped being a special employee he did say the only way out is growth so he knows this too but I think he can hold both of these ideas
11:37
which is we need to grow but he's also uh frustrated the Hegelian dialectic of course um yeah I mean if if GDP is growing really really fast you could even keep tax rates the same or lower them and still pay down the debt and do more redistribution. Like you could be
11:50
Like you could be paying people like you could have like something that feels like a massive social safety net, something that feels like UBI or or socialism but actually attack people less because the growth is so significant.
12:02
So it all changes once super intelligence arrives.
12:05
Maybe maybe Zoron is just so AGI pilled himself. That's true.
12:10
That he's like, "Yeah, we're gonna have plenty plenty to go around.
12:14
we're gonna have such high levels of production.
12:17
Sure, the state can just seize a little bit.
12:20
Alternatively, maybe he's a hardcore Bitcoiner and so if he wants if he wants there to be no billionaires, no, it is impossible mathematically to be a billionaire in Bitcoin because there's only 21 million of them.
12:31
So, the richest you could possibly be is a millionaire if you owned over 5% of the Bitcoin supply, right?
12:38
So, there will never be a Bitcoin billionaire because there aren't a billion of them.
12:42
Well, whereas there are billions of dollars.
12:44
Maybe he doesn't want people in the network to vote to expand the supply. Yes.
12:47
So, he's probably against the expansion of the Bitcoin supply, but maybe he's a hardcore Bitcoiner.
12:52
He's a hardcore Bitcoiner. He's against inflation. Yes. Yes. Exactly. Exactly.
12:57
Anyway, in other words, we got to hit that gong Jordy for Rahul Vora.
13:02
Where's the hammer founder?
13:04
You got the hammer of superhuman.
13:04
He has sold his company to Grammarly for an undisclosed sum.
13:13
Nothing like hitting a gong for an undisclosed sum.
13:15
You just don't you don't know. Could be a million. Could be a billion. Who knows?
13:19
But no, but I I've used uh Superhum for many many years now.
13:24
How many When did they really get out?
13:26
Was it like six years ago?
13:26
I think it was been older.
13:27
I I feel like it came on the scene like 20 years ago.
13:29
on the scene like 20 years ago. I remember I was I think yeah I was graduating college and there was this concept that the thread boys at the time were talking about luxury software because the idea of the idea of charging $30 for email which was just default
13:43
free they grew really quickly uh apparently at the time of this acquisition they have about 35 million of ARR very impressive um of ARR okay so that's like what 100,000 users paying roughly something like that something like that between 100 and 200,000 users paying $30 a month. Yeah. Got it. Pretty Yeah. Got it. Pretty cool.
14:03
I mean, yeah, serious business at 30 35 mil, you know, if it's run effectively.
14:08
Could have been, you know, kind of a lifestyle business throwing off a couple million dollars in free cash flow. Um, clearly go bigger. Yeah, potentially. I'm really interested.
14:15
We should have the CEO of Grammarly on because I want to understand. So, they acquired Kota. That was surprising.
14:21
Kota was another company that had a lot of uh, you know, meaningful traction, a lot of fans.
14:27
It was a great product, but I I don't think was breaking out, right?
14:30
It wasn't, you know, on a path to being a public company itself or or a billion dollar y outcome.
14:36
And then um now they're adding superhuman uh to the network.
14:41
I don't see yet exactly how these tools all interconnect, right?
14:47
Grammarly uh helps you with grammar in the browser.
14:49
Superhum uh does have some AI functionality.
14:52
I'm sure I'm sure uh and I know KOD does as well.
14:54
unclear to me yet how uh if it's like the same user that's going to be using all of these things, if it's truly a productivity suite. Yeah.
15:04
suite. Yeah. or it's a basket of I mean selling a point solution in the enterprise is hard with when you look at the the story of Slack versus Teams like Microsoft just has so much dominant distribution there actually having someone rip out their productivity suite
15:19
and say as a company we're using superhuman for email and Grammarly for writing and you know this other product for docs and uh and then not to mention the fact that notion's running somewhat of a similar playbook with notion mail and notion docs and and then Google has a similar productivity suite. Um, so I
15:34
Um, so I mean it's a tall it's a tall order, but uh will be cool to see how these different products interact because they're clearly um excited about uh tying them all together in some interesting way. Y very cool.
15:46
Anyway, um you mentioned humanoid robotics and uh you know there's another player. It's not just Elon. It's Jeff Bezos as well.
15:56
Amazon now has as many robots as employees in its warehouses.
15:58
So the number let's give it up for the hardworking robots. Yes. Hard work. They haven't strike yet.
16:07
The number of Amazon employees per facility uh over the last 10 years has kind of oscillated from 750 800 employees per facility went up to a thousand.
16:17
Now it's down around 700 down 30% over the last 5 years. That's the postcoid era.
16:23
And the packages handled by Amazon end to end per employee has skyrocketed over that time from from zero uh to over 4,000.
16:35
And so the robots are taking over.
16:37
They're about to cross their 1 millionth robot deployed in an Amazon warehouse.
16:41
So congrats to Jeff Bezos.
16:41
Hope you enjoy the wedding.
16:43
Yeah, we should hit the gunk for that. Maybe the gunk sound.
16:47
Yeah, I saw a post somebody maybe it ended up in the timeline, but but I'll say it now.
16:50
I was like, "A lot of people have opinions on Jeff Bezos's wedding.
16:52
I ordered a phone last night at 10 p. m.
16:58
and I got it this morning.
16:58
I think he should be able to celebrate however he wants." So, I'm happy for him.
17:01
And built a huge business. Very helpful.
17:04
This was kind of And even looking at like Amazon puts up over hundred billion dollars a year of profit. Yeah.
17:12
Let the man have a $50 million a year wedding. A year. Not a year. Hopefully not a year. Hopefully not a year. He's only done it twice.
17:23
Some guys get to that stage of life and they just enjoy weddings, right?
17:27
So they just have, you know, Yeah. frequent.
17:30
There was this narrative around the wedding actually that uh oh, he only invited celebrities cuz like Sydney Sweeney went and everyone was like, "Oh, is he really tight with like Sydney Sweeney?"
17:38
Like and but but but and I was saying like look, he invited 200 people.
17:43
200 people. you're seeing Sydney Sweeney because like if you're a if you're a celebrity like you know paparazzi photographer you're not taking a picture of like RJ Scaring from Rivian who Bezos invested in and might have been there I don't know or like Andy Jasse the
18:00
current CEO of Amazon probably got the invite but is doesn't get in the paparazzi so he's having a party he's having a party but you're going to invite your boys your business I'm sure I'm sure he did but yeah yeah the media is obvious on who are the biggest stars there because they are who drive click. But that's why we need to send our
18:15
But that's why we need to send our paparazzi people Andy Jasse. That's true.
18:20
Yeah, that that would Andy Jasse spotted me. Yeah. Yeah.
18:23
And then it's just like Sydney Sweeney off blurry in the left and it's like everyone's like uh you missed the most important person at the wedding. We did not. No, we didn't.
18:34
Andy Jasse is the most important person at that wedding potentially if he went.
18:37
Uh anyway, in other news, Commonwealth Fusion Systems, a fusion startup, uh has struck a deal with Google.
18:45
Historic news, we've signed a landmark agreement with Google, a multifaceted partnership that dramatically advances our commercial fusion energy mission.
18:51
Google signed an offtake agreement for 200 megawatts of power from our first ARC power plant.
18:57
Google is increasing its corporate investment in uh in Commonwealth fusion systems.
19:02
Google has the option to procure power from future arcs.
19:03
The deal is a very strong signal that the world world wants clean, secure power that fusion offers together.
19:09
We're catalyzing the fusion market.
19:11
So there's a couple players in this space, but all the interesting use of hashtags as well dating themselves a little bit.
19:17
They maybe have a veteran social media manager maybe more for the world energy power moves because people are searching that.
19:27
Didn't Elon ban hashtags recently? I mean ads.
19:30
Okay, so you can still rip them in in hard posts. Uh yeah.
19:32
Uh anyway, beautiful video. Good luck to them.
19:36
And if you want to learn more about Commonwealth Fusion Systems, go to hashtagfusion energy. Stop it. Stop it. No, this is great. Super exciting.
19:45
So, I mean, this is uh this is like the flip side of the Sam Alman Helon deal where OpenAI and Sam Alman, he's clearly thinking about power being very, very important for the long term.
19:56
Crusoe has been very important in finding stranded energy, building natural gas plants, and helping scale up Stargate in Abalene, Texas.
20:06
And Commonwealth Fusion Systems is kind of the other side of that, and they're partnering up with Google. So, good to see it.
20:12
In other news, Microsoft claims that their new AI framework diagnoses four times better than doctors.
20:19
And people are kind of putting this in the truth zone, says uh Dr. Dominic Ing.
20:25
I'm a medical doctor and I actually read the paper.
20:28
Here's my perspective on why this is both impressive and misleading.
20:30
There were some interesting things where in the control set, I believe the doctors couldn't use Google.
20:36
So, it wasn't necessarily like a current doctor versus the Microsoft AI system.
20:41
It was like a like a like a doctor just sort of acting like a human GPU. Yeah. Yeah.
20:47
just being like completely pen and paper, which isn't really the benchmark that you need to go against, but still impressive.
20:56
Maybe the 4X is a little overstated, but the average doctor in the United States when they're sitting with a patient is googling Google whatever's happening and then Reddit at the end and then going to Reddit and then OpenAI is scraping that. Yeah.
21:10
And you know, summarizing it. Yeah.
21:12
But it is cool because now we have a paro frontier.
21:13
We have a graph of uh the average total diagnostic care per case, how much it costs uh versus the diagnostic accuracy and Microsoft with this team.
21:24
This is led by Mustafa Sulleon who in an interesting trade deal was over at Google in DeepMind, co-founded DeepMind, then did Inflection AI and got like Aqua hired in one of those like zombie deals to go to Microsoft.
21:39
Microsoft. He's now the head of AI at Microsoft and he uh has been leading this this um this push that's much more it feels maybe more Microsoft coded than what Apple's or that what Google is doing on the deep mind team where Google is focused on like solving these very
21:57
fundamental questions in in science and health like how do you do protein folding that's not something that you go into like your primary care doctor say like my shoulder hurts and like they need to know how to fold proteins right but If you go into your doctor and you say, "My shoulder hurts." Like having an
22:11
Like having an AI tool, basically like a great fine-tuned chatbot that's HIPPA compliant and works within the Microsoft Teams ecosystem to pull the patient records in appropriately, fine-tune on the right data, and then and then suggest a bunch of things that the human doctor can then go investigate.
22:28
That's super valuable and that feels like something that Microsoft is equipped to roll out to uh the network of of doctors and clinicians that probably are already on Microsoft tool sets and and Microsoft stacks.
22:41
So, good stuff from Microsoft.
22:44
Um, we'll have to get somebody on the show to talk about that since AI in healthcare is all the rage.
22:47
Anyway, hit that sound effect, Jordy, as we move to our next important news item, which is that ramp.
22:56
com is available now for you to money. Save both.
23:01
Easy to use corporate cards, bill payments, accounting, and a whole lot more all in one place. Go to ramp. com. Um, RAMP.
23:06
Anyway, we have a fun post here from Vivid Void says, "I'm literally plotting on your rise, bro.
23:14
I'm literally up in the lab scheming on beautiful abundance for you.
23:16
I'm cooking up ways for a rising tide to lift all of our boats.
23:20
I'm praying on I'm praying for your gains.
23:22
I'm lighting a candle to your inevitable splendor, dude. GM.
23:25
And we mean that from the bottom of our hearts. Yes. To all of you.
23:31
Well, if you're looking to design a meme like this, head over to figma. com.
23:34
Think bigger, build faster.
23:35
Figma helps design and development teams build great products together, and you can get started for free at figma. com.
23:40
Uh, Signal had a post breaking down Mark German's analysis of Apple weighing anthropic or open AI in the major power shift in the Siri game.
23:50
Signal says, "Absolutely astonishing.
23:52
Apple used to own the full stack, silicon to software to services.
23:56
Now they're outsourcing the one layer that will define the ne the next decade of computing.
24:00
It's a metaphysical betrayal of their own DNA.
24:05
Anthropic and open AI don't need Apple, but Apple desperately needs one of them.
24:09
This puts Apple under the models it's integrating wild reversal.
24:11
Whoever they choose, they now owe existential dependency to.
24:16
And finally, if it consu if consumers realize Siri does not equal Apple anymore, that it's powered by OpenAI or anthropic, then what exactly is Apple's IP?
24:25
Thin shell over someone else's mind, that kills the aura of vertical magic. I do agree. It's exciting. It's certainly exciting. It's exciting.
24:36
Uh, I would I would push push back and say OpenAI and Anthropic don't need Apple.
24:41
I would say that Open AI, sure they don't need Apple, but integrating deeply at the hardware layer is going to be important.
24:49
Otherwise, they wouldn't buy Johnny Ives love from they wouldn't spend billions of dollars trying to develop their own hardware device.
24:58
Obviously, they're trying to figure out what's after phones or what what could complement phones. Yep.
25:01
what could complement phones. Yep. uh but ultimately I think it is a for both companies extremely important relationship in the context of consumer right is this going to impact anthropics codeg genen business no uh but will this it might it might in the sense that if I'm asking Siri to do tasks that require
25:23
writing code and the models are generating more code that's more training data right like that example of like I took a photo of Pat McAfee I took a screenshot from the Pat McAfee Show and asked how tall is his desk and chat GBT 303 wrote like a thousand lines of Python to go through and calculate how tall that desk was. And then I was able to say
25:46
And then I was able to say basically by thumbs up or thumbs down or asking another question.
25:52
I was able to give it human feedback on whether or not that code path was the accurate if it got me the answer I wanted.
25:58
And if I if I walk away from that, that's a signal.
26:02
If I ask another question, that's a signal.
26:04
If I give it a thumbs up or a thumbs down, that's another signal.
26:05
And so there actually is a world where if you need to test a ton of different realworld Yeah.
26:12
code generation tasks and and is the code that you're writing actually useful to the consumer, the average consumer, the person who's just asking Siri for random stuff, maybe that is valuable. And so I I I don't know.
26:25
I'm still I'm still curious as to like what is the flow of value at the same time?
26:29
Obviously, are they going to get better, more valuable data from just having tens of thousands or hundreds of thousands of developers building on true cloud code daily?
26:38
The other the other thing is is what what we're what's completely unclear so far is will consumers just be talking to Siri and it's still the relationship just like Siri might be running on some cloud somewhere.
26:54
Consumers aren't saying, "Well, Siri's not Apple because it's not their own servers."
26:58
Well, Apple does push on that point very precisely because they say around privacy.
27:03
They say, "Yes, when you use our models, it is on our cloud and it is super private and it is on device." Yeah.
27:09
But they can still partner with OpenAI or an anthropic figure out how to deliver that same end value prop. Yeah.
27:15
It seemed like a lot of the privacy narrative at Apple was about control of the ecosystem and the ability to imp to to really push on the Apple tax and justify it, right?
27:27
And and maintain that leverage.
27:27
Um I don't know it it's interesting.
27:30
Also remember Siri was an acquisition. Yeah.
27:34
Consumers didn't have an issue with that saying, oh this isn't Apple code. Yeah.
27:41
You this isn't homegrown.
27:41
This isn't farm to table. This isn't handmade.
27:44
Can you imagine if consumers cared about that?
27:48
It's like this was technically an aqua hire, not an acquisition.
27:53
Oh no, I don't want to wear beats.
27:53
Like the the earnout structure wasn't at to my liking.
27:59
Like only one board seat changed hands.
28:02
I wanted I wanted three board seats in the acquisition. No. Yeah.
28:04
It was telling though that uh uh Pesh here has a post Claude gave Tim Cook sticker shock.
28:12
I was referring to this earlier in the show. Yes.
28:13
Uh, Anthropic have asked Apple for a multi-billion dollar annual fee that increases sharply each year. Very acceleration pill. Yes. Enthropic. Yes. But, uh, yeah.
28:23
So, so ultimately I think that that feels like, you know, Apple having a cost center that can scale exponentially with usage is not going to be something they're super excited about because at the same point in time they're trying to show we are this ecosystem of software products.
28:44
Yes, hardware is important but really you know we're growing services revenue and so if their costs are growing sharply as defined here.
28:53
Uh but the surprising thing is is yeah you could imagine a world where a part you know open AI says we'll give it to you for free for the first five years or something like that right I mean it could even be that that is paying Apple because if they are able to drop the inference cost low enough that it's similar to serving a Google search and they're able to stuff ads in it that would be incredibly valuable.
29:22
Like imagine if you go to Siri and you say um I'm looking for a new backpack and chat GPT can dynamically insert out ads into that response within Siri.
29:31
That should be extremely valuable.
29:34
And so that I mean that's the that's the current structure that Google has.
29:39
That's the current arrangement that Google has.
29:41
I'm it's unclear to me how much how much different the market will be going forward. I don't know.
29:46
Um anyway, let's read through some of this uh Mark German report in Bloomberg.
29:53
Apple weighs using anthropic uh or open AAI to power Siri in major reversal.
29:58
I'll read through some of this and then I'll want your reaction, Jordy.
30:00
Uh Apple is considering using artificial intelligence technology from anthropic BPC or or OpenAI to power a new version of Siri, sidelining its own in-house models in a potentially blockbuster move aimed at turning around its failing AI effort.
30:15
The iPhone maker has talked with both companies about using their large language models for Siri.
30:20
According to people familiar with the mount, it asked them to retrain versions of their models that could run on Apple's cloud infrastructure for testing.
30:26
Uh who asked not to be identified uh discussing private deliberations. Interesting.
30:30
So I mean I don't know if Apple's cloud infrastructure has a lot of H200s and is ready.
30:38
I mean they certainly could not serve 4.
30:40
5 like a truly large model like that.
30:43
another hundred billion customer for Jensen for Jensen.
30:44
I mean, realistically like like Apple has enough of a relationship with TSMC that they could go and do kind of like uh you know that the the A3 the A or the uh the Apple silicon, you know, the M2 chip, the M3 chip, like the M4 chip.
31:00
They could do a different run with TSMC that's more optimized for inference of large models in the data center context.
31:09
Uh, but it does seem like it'd be a very different muscle for Apple since they have not been I mean when you think Apple's cloud infrastructure it's like storing photos.
31:15
It's like iCloud like that's probably their biggest service or like like I guess like Apple Music, Apple TV. Like they serve files.
31:23
They don't do a lot of machine learning inference versus say Meta that already has a you know hundred billion dollar data center that just does recommendation algorithms and machine learning and inference.
31:36
It's not fully like a transformer large transformer model, but it is, you know, multiparameter model for deciding what ads go where and what reels get served to you.
31:45
So, um, it would definitely be overtime to serve you the perfect reel right now. For sure. For sure.
31:49
And and and Apple just doesn't have that scale GPU cluster that I'm aware of.
31:54
So, uh, they'll have to call up Dylan Patel and some folks to get it done. Call up Jensen maybe. Who knows?
32:01
If Apple ultimately moves forward, it would represent a monumental reversal.
32:04
The company The company currently powers most of its AI features with homegrown technology it calls Apple Foundation models and had been planning a new version of its voice assistant that runs on that technology for 2026.
32:15
But it seems like they're backtracking.
32:18
Uh switch to claude or chatbt models for Siri would be an acknowledgement the company is struggling to compete in generative AI the most important new technology in decades.
32:27
Apple already allows chatpt to answer web-based search queries in Siri but the assistant itself is powered by Apple.
32:33
Apple's in investigation of third into thirdparty models is at an early stage and the company hasn't made a final decision on using them uh marking a change which is under discussion for next year could allow Certino based Apple to offer Siri features on par with AI assistance on Android phones which you don't hear about that much.
32:53
about that much. I see the Gemini demos on you don't your friends that use Android phones don't don't talk about how they're you know the Lone Ranger is on Android the Lone Ranger you know we got one in our group chat but other than
33:06
that it is rare but even then I feel like there would be more like viral X posts about cool Android features if they had really done well on the Gemini product like implementation and it's just like maybe not quite there but Gemini is great the models are are really top tier. The Gemini app on iOS
33:24
The Gemini app on iOS is not amazing.
33:27
Like the the the text to speech is really really rough.
33:30
Like you click you click dictate and then it will just stop dictating when you pause and then just submit their query.
33:38
And so you can be with Chat GPT.
33:40
I love it because I open it up and I hit that button and then I think and I say okay I'm trying to do a research on the Mag 7 and I want to look at Mag 7 their performance over the last decade.
33:50
So go back and pull what their market cap was 10 years ago and then what it was today and then create a table.
33:56
But then if I pause like Google will just submit that and I'll be like whoa no I wanted to add more and I wanted to give you more context but chatbt waits until I push the button to finish that.
34:06
So it's just like minor minor UI but that's more important to me than like does it have 150 on MMLU versus 149.
34:13
versus 149. like I care more about the UX of like am I did I just waste 10 minutes by sending like a query that isn't complete and I know I'm going to get a bad answer on um and so uh I I I I don't know but again Android is clearly of Siri here so it's
34:32
not even a competition um the competing project internally dubbed LLM Siri that uses in-house models to uh remains in active development uh so uh representatives for Apple declined to comment shares first closed up over 2% after Bloomberg reported on the deliberations. So, Wall Street clearly
34:48
So, Wall Street clearly likes the idea of one of these deals getting done, but very big question as to the structure of those deals.
34:55
And I keep triggering Siri because I'm talking about it on my phone. It keeps popping up. Fantastic. Incredible.
35:01
Uh the project to evaluate external models was started by Siri chief Mike Rockwell and software engineering lead Craig Federigi.
35:09
They were given oversight of Siri after the duties were removed from the command of John Gandrea, the company's AI chief.
35:17
This was a spicy little trade deal going on over there.
35:21
Andrea Federiki, a spicy meatball over at Apple and Certino.
35:25
Uh he was sidelined in the wake of a tepid response to Apple intelligence and series feature delays.
35:33
Rockwell, who previously launched the Avision Pro headset, assumed the Siri engineering role in March.
35:39
After taking over, he instructed his new group to assess whether Siri would do a better job handling queries using Apple's AI models or thirdparty technology, including Claude, Chat GPT, and Google's Gemini. Interesting.
35:52
After multiple rounds of testing, Rockwell and other executives concluded that anthropics technology is most promising for Siri's needs.
35:57
Um that led Adrien Perka, the company's vice president of corporate development to start discussions with Anthropic about using Claude.
36:05
What's interesting is that uh SSI not in the conversation here yet.
36:09
Uh also uh thinking machines not in the conversation yet.
36:12
And when you think about thinking machines, that recent uh information article about um Mera Marotti wanting to do fine-tuning RL for businesses.
36:22
It's like there's very few bigger businesses than Apple.
36:25
That's a pretty good client to go win. Yeah.
36:28
And so if you have a foundation model, you have to assume she's raised what? Two billion. So she can do it. Preed. Yeah. Yeah. The preede. It could be a thing.
36:38
Apple doesn't really work with a lot of preede companies. Yeah. Yeah. Yeah.
36:43
They're like, "Call us when you're like series C." I'm sorry.
36:47
Like, you know, I can't bet my career on a on a preede company, but call us if you can get a seed round done. Yeah. Yeah.
36:54
But I mean I I I think the I think the fact of the matter is like$2 billion dollars is enough to do a GPT4 level run to do a Deepseek style run.
37:02
And so you if you take all the best practices from DeepSeek V3 and R1 like the reasoning model and you spend the money to go do the training and you pull all the data together and and maybe you're not completely stealing the data from the GBT4 API but you're getting it somewhat legitimately through you know these these data brokers like Merkore and Scale and whatnot like you should be able to get a you know near the frontier level model done.
37:31
issue is Apple isn't set up to comp top AI researchers, right?
37:36
So, in some way in some in some ways it's not they're not willing to spend the money to put together to say, "Hey, we're we're getting kind of cooked in AI.
37:45
We should spend $5 billion like Zach just did to to put together, you know, an incredible team, but they'll have to pay for it by it sounds like they may end up paying annually to get access to the technology that those types of teams built." Yeah.
37:59
And so you you got to pay for it somewhere.
38:02
It's the capex opex switcheroo.
38:04
The CFO is like, "Okay, yeah, like we don't want we don't want to spend a bunch of money up front to reap this reward endlessly, but um we will pay a billion dollars a year, I guess." I don't know.
38:15
Um they they they seem not fully AGI pelled over there.
38:20
Uh the Siri assistant, originally released in 2011, has fallen behind popular AI chat bots, and Apple's attempt to upgrade the software has been stymied by engineering snags and delays.
38:30
A year ago, Apple unveiled new Siri capabilities, including ones that would let it tap into users personal data and analyze on-screen content to better fulfill queries.
38:38
I do wonder if Apple is thinking about thinking machines.
38:40
Um, you have to imagine that Zuck is thinking about buying the company, right? Sure, they talked.
38:47
It's only a couple billion dollars.
38:48
And by all accounts, I think I think to to be honest, it sounds like Zuck had offered somewhere around the last private market valuation for SSI. Yep.
39:00
And he probably even though all the investors had a 1x prep, right, he probably needed to do that to get at least the early investors comfortable with it.
39:12
So, but Ilia is a true missionary, true believer, wants to take the straight shot to SSI.
39:18
He doesn't care about Ilia doesn't care about the debt ceiling. Let's put it that way.
39:25
Meera, on the other hand, is doing reinforcement learning for businesses.
39:28
So, you throw you throw 15 billion dollar aqua hire her way with that team, maybe it gets done. I don't know.
39:35
If I'm Apple, what let's pull up the Mag 7 market cap again. What are they at?
39:40
Two trillion, three trillion. They're up there.
39:42
Um and uh and Apple is, you know, it feels like they could afford even 15 billion for thinking machines. Sure. Do you disagree? No.
39:55
I mean, they totally could, but they'd be looking at it as talent acquisition, I think. Yeah.
40:03
Because the team is I mean, it's such a young company. Yeah.
40:06
And but if you were if you were in in the big man but so you would do that.
40:12
You think Meera is going to be cool just clearing 74 million?
40:15
She's not going to make more than Tim.
40:18
You think she's going to be down to sell now and then just make have total comp less than the average AI researcher going to Meta? I don't know.
40:29
I mean that that the beauty of an acquisition is that like you can be structured to make more than the CEO at least in a short amount of time.
40:35
Like that certainly was the case with the Beats acquisition, right? Sure.
40:38
Like for those years when that earnout was active, they made more than the CEO. Yeah. And that's okay.
40:45
Anyway, um people with knowledge of Apple's AI team say it is operating with a high degree of uncertainty and a lack of clarity with executives pouring over a number of possible directions.
40:55
Apple has already improve approved a multi-billion dollar budget for two 2026 for running its own models via the cloud but plans for that beyond remain for beyond that remain murky.
41:06
Still, Federigi Rockwell and other executives have grown increasingly open to the idea that embracing outside technology is the key to a near-term turnaround.
41:15
They don't see the need for Apple to rely on its own models which they cons which they currently consider inferior.
41:20
And this makes a lot of sense because Apple is not great at uh at software that needs to be updated, you know, 25 times a year.
41:27
We're in this crazy AI model horse race.
41:29
You're you don't want to be waiting around for oh yeah, like at the next WWDC they're going to improve it when Sam Alman shipped like 25 updates.
41:38
Sam's going to have 40 buzzy announcements between now. Exa Exactly.
41:46
So Beats was Apple's largest ever acquisition.
41:48
paid $3 billion including four million of 400 million of Apple stock.
41:53
But Beats at the time was doing a billion dollars of annual revenue.
41:56
They only paid like a 3x revenue multiple. Wow.
42:01
That's Intel's modem business. Yeah. Yeah. Yeah.
42:03
Uh which was in 2019 for a billion. Okay.
42:05
And it was doing two billion in revenue. Yeah. Probably. Yeah. You look that up. I'll keep reading.
42:12
uh they don't see the need to rely on the own models.
42:17
Licensing thirdparty AI would mirror an approach taken by Samsung.
42:20
While the company brands its features under the Galaxy AI umbrella, many of its features are actually based on Gemini.
42:26
Anthropic for its part is already used by Amazon to help power the new Alexa Plus, which I haven't seen rolling out into the Amazon Alexexas.
42:35
Maybe the Alexexas aren't wired up properly and you have to get the next batch, the next revision.
42:40
So Intel's modem business was doing a over a billion.
42:42
So less than a 1x revenue multiple. Yeah. Basically a 1x.
42:46
And you remember in 2018 they acquired Shazam. Oh yeah. For 400 million.
42:48
How much were they doing? I got to ask.
42:51
I gota This is in the future if its own technology improves the executives believe Apple should have ownership of AI models given their increasing importance of how annual revenue was at the time of acquisition. I want to say 80. How much? Waiting.
43:11
The company is working on a series of projects including a tab a tabletop robot and glasses that will make heavy use of AI.
43:16
I'm really excited for the Apple tabletop robot that doesn't get enough coverage these days.
43:20
I think that's a really cool idea.
43:21
So Apple paid 400 million.
43:24
The app was pulling in roughly 38 million a year.
43:26
Oh, so over 10x revenue multiple. Okay.
43:29
So anyways, not a long history of making massive aqua hires.
43:36
Yeah, actually no history at all.
43:39
Well, whatever Apple's planning, they got to do it on linear.
43:41
Linear is a purpose-built tool for planning and building products, meet the system for modern software development, streamline issues, projects, and product road maps.
43:51
If Apple wants to be more like OpenAI, one way they could do that is by getting on linear because OpenAI uses linear. Oh, here we go.
44:00
We get to we get to theme machines.
44:01
Uh, Apple has also recently considered acquiring Perplexity in order to help bolster its AI work.
44:05
It also briefly held discussions with Thinking Machines Lab, the AI startup founded by former a OpenAI chief technology officer Mirror Morati.
44:13
Uh, but nothing really more than that in the story.
44:18
Uh, Apple's models are developed by a roughly 100 person team, maybe two two times big, two big, who knows? Maybe 50. It's a lot of pizzas. A lot of pizzas.
44:26
uh run by uh Ruming Pong, the an an Apple distinguished engineer who joined from Google in 2021 to lead this work.
44:35
He reports to Daphne Luong, a senior director in charge of AI research.
44:39
Luong is one of Gandrea's top lieutenants and the foundation models team is one of the few significant AI groups still reporting to Gandrea.
44:46
Even in that area, Feder and Rockwell have taken a larger role regardless of the path it takes.
44:51
The proposed shift has weighed on the team which has some of the industry AI industry's most in- demand talent.
44:57
Some members have signaled internally that they are unhappy the company is considering technology from a third party creating the perception that they are to blame at least partially for the company's AI shortcomings.
45:07
They have said that they could leave for multi-million dollar packages being floated by meta platforms and open AI.
45:13
Yeah, if you are a great AI researcher, the idea of being responsible for just integrating the work of OpenAI or anthropic is going to be much less exciting than trying to build frontier models. Totally.
45:27
And also like I don't know just in terms of missionary work like the idea of building like the the the the most advanced god that you can talk to in this crazy AI like that is exciting.
45:38
But there's also something that's like I think potentially like a really strong missionary pitch for you're you are going to create you're going to build models that are used at massive scale by every iPhone user and you can actually you know surprise and delight the customer to the tune of hundreds of millions of people.
45:58
Like I think that is I think that is actually like pretty pretty laudable work and very even going to work at super intelligence at Facebook knowing or at Meta knowing that knowing that your work will be integrated into the meta ray bands and you're going to have like a clue like experience just wearing glasses like that is a cool that is like a new dimension that you can build on.
46:22
I remember the initial Siri buildout, they they did a bunch of magical things like you could say like Siri, tell me a joke and it would come up with these like funny zingers that were like tuned to Apple's brand but still actually pretty funny and they hired comedy writers to work on it.
46:38
And so like I think that there still is a world where you you find you you develop the model, you own it, and you're doing the foundational like serious AI research, but you're still productizing it in the Apple way, creating something special.
46:52
You know, you already hear about this with the the clawed is different than chatbt like the vibe test.
46:58
like we haven't seen the vibes fork out really all that much, but there's definitely a world where the Apple LLM performs at a similar scale and speed and the reasoning and it's doing well on the benchmarks, but then you know you're talking to the Apple model and it has the Apple feel to it and like that could be cool if you if you can actually It could be so cool. It could be very cool. It could be so cool. Yeah.
47:22
Like if they brought if they could bring back some of the like flare from the brand campaigns that you could that you remember Apple by 1984 the like original iPod like these things that were just iconic and if you could think about how to like embody that that original Apple it's like Steve Jobs voice basically. Yeah.
47:41
Bicycle for the mind and and like encouraging that that that that wonder.
47:46
I I I keep going back to that idea of like what would Steve Jobs have thought about vibe coding and it's like he would have loved it. He loved Garage Band.
47:55
He loved giving everyone the tools to be an artist. Everyone's a musician. Everyone's a programmer.
48:02
Like these these things this like this giving someone the ability to do really really interesting things with Siri. Awesome.
48:09
I think I think it's very very on brand for for Apple.
48:13
Anyway, let me tell you about Vanta.
48:15
Automate compliance, manage risk, improve trust continuously.
48:19
Vantage trust management platform takes the manual work out of your security and compliance process and replaces it with continuous automation.
48:23
Whether you're pursuing your first framework or managing a complex program, head over to vanta.
48:29
com and tell them the technology brother sent you.
48:32
There are some more trade deals in this uh in this report in Bloomberg.
48:36
Uh one of Apple's most senior large language model researchers, Tom Gunter, left.
48:40
He had worked at Apple for about eight years and some colleagues see him as difficult to replace given his unique skill set and the willingness of Apple's competitors to pay exponentially more for talent. That's what I'm saying.
48:51
Apple this month also nearly lost the team behind MLX, its key open source system for developing machine learning models on the on on the latest Apple chips.
49:01
After the engineers threatened to leave, Apple made counter offers to retain them and they're staying for now. That's dramatic. Wow.
49:10
Um, in his discussions with both Anthropic and OpenAI, the iPhone maker requested a custom version that could run on Apple's private cloud.
49:16
Apple's such an interesting culture because it's counter to a lot of other cultures in tech that are like, "Oh, I might work at Meta for 5 years and then go to Google for a few years and then dabble at a startup and then go back to Meta."
49:30
and Apple, you see more of the the somebody who's been there for 15 years. Yeah.
49:36
But then at a certain point, even if you've been there a long time, you're looking around and you're saying people with my exact same skill set are making Yeah. 5 10 times.
49:47
This article highlights that that the Meta pay packages have been between 10 and 40 million. Yeah. Annually.
49:54
And I would trust more on that.
49:55
That feels more realistic than um what uh the New York Times had been saying.
50:01
about uh hundred million uh dollar uh signing bonuses and all that good stuff.
50:07
Well, you know, the source is not conflicted at all.
50:10
The guy that hates the fact that these pay packages exist Yeah. is quoting them. Yeah.
50:14
Uh yeah, it's so funny to source a a podcast between two brothers doing SciHops. It's like, come on. Come on, New York Times. Step it up.
50:28
uh the company already has already inter internally tested the idea of running L these LLM's other models other companies models on their own Apple controlled cloud services.
50:38
Um obviously that's very important for the data privacy narrative.
50:42
Other Apple intelligence features are powered by AI models that reside in consumers devices.
50:48
These models slower and less powerful than cloud-based versions are used for tasks like summarizing short emails and creating Gen Mojis. It's here for Jen Moji. Just bake down Gibli. Just bake down Gibli.
51:01
Have it be a filter in the iPhone app. Boom. You're done.
51:04
Like you have a banger product there. Something like that.
51:08
Just c like, you know, you can go in and make any photo in your camera roll black and white.
51:12
You can make it Hollywood like teal and orange.
51:14
You can make it, you know, super moody or super cool or super warm.
51:19
Like you can adjust the the contrast, the color temperature. Turn it into a cartoon. People will love that.
51:26
Everyone will be using that.
51:26
Uh Apple is opening up the ondevice models to thirdparty developers later this year, letting app makers create AI features based on this technology. Very excited for that.
51:35
Very bullish for Apple on that. Amazing. Yeah.
51:37
Which is why the fact that they almost lost the team working on that entirely.
51:41
They need that team badly is would have been crazy because if that product stagnates I saw a founder I saw a founder I forget their name posting uh hearing that you know the team that you're based he was building a product based on MLX and then he heard the MLX team almost left.
51:57
He was like freaking out.
52:00
He was like whoa no it's rough.
52:02
So last year OpenAI offered to train ondevice models for Apple but the iPhone maker was not interested.
52:08
They said we got it handled. Turns out they didn't.
52:11
Since December 2024, Apple has been using OpenAI to handle some features.
52:15
In addition to responding to world knowledge queries in Siri, chat GPT can write blocks of text in the writing tools feature.
52:21
Later this year in iOS 26, there will be a chat GPT option for image generation and on-screen image analysis.
52:27
Uh while discussing the potential arrangement, Apple and anthropic have disagreed over preliminary financial terms.
52:33
According to the people, uh the AI startup is seeking a multi-billion dollar annual fee that increases sharply each year.
52:39
or the struggle to reach a deal has left Apple contemplating working with OpenAI or others if it moves forward with the third party plan.
52:48
They said if Apple does so interesting you can imagine Open AI being more willing to strike some type of deal because they're they want to be effectively the next Google right in terms of this dominant consumer internet company. Yep.
53:07
Uh but at the same time, OpenAI is building devices to try to kill, you know, Apple.
53:12
You can imagine they they wouldn't say that. They wouldn't say that.
53:17
I I do believe that consumers there's like a finite amount of devices that they're willing to have.
53:23
And so if they want to compete in hardware, it's a little hard for Apple to to get excited. Yeah.
53:27
Although they partner with Google and Google has I have been surprised by the number of devices that people carry.
53:33
I mean, like I'd say like a decent chunk of San Francisco tech people carry AirPods, an iPhone, an Apple Watch, and an Apple laptop basically everywhere they go. That's four devices. That's like a lot. Yeah.
53:49
And then Yeah, maybe maybe a VR headset eventually.
53:51
Um well, if they do wind up buying a device, they got to pay sales tax.
53:55
They got to get a numeral HQ sales tax on autopilot.
53:57
Spend less than five minutes per month on sales tax compliance.
54:00
Uh anyway, uh in other news, John MW, the inventor of Muwing, has passed away at 96 in his castle. Rest in peace to John.
54:09
Uh professor John Mu, visionary orthodontist and founder of orthotropics passed away peacefully in June 2025 at the age of 96. in his castle.
54:20
A piring a pioneering figure in the facial growth and development, John MW challenged the status quo of traditional orthodontics, arguing that environment, posture, and function, not genetics alone, shaped the face.
54:30
His philosophy, orthotropics aimed not only to straighten the teeth, but helped promote healthy jaws, open airways, and confident faces.
54:39
through his work was though his work was often met with fierce resistance from the establishment.
54:44
Jon stood firm in his convictions driven by a lifelong commitment to prevention, integrity, and science and giving every child a fair chance at full facial development.
54:56
Today, his legacy works on through uh lives on through the work of his son.
54:58
Uh John Mu will be remembered not only as a clinician but as a courageous truth seeker who dared to ask better questions and never stopped looking for better answers.
55:10
So never stop chewing some mast gum.
55:10
Say uh pour one out for John.
55:14
I wonder how he processed the explosion of Muing's popularity on social media.
55:21
It must have come as a surprise because it was the last like few years of his life.
55:25
been working on it for probably I mean he's 96 and somebody probably printed out a tweet for him and said look uh some some trad influencer has gotten 10,000 likes on your on your work and a million views.
55:42
Anyway, we have some trade deal news.
55:43
Ring that gong Jordy Mary Denophrio has joined Cross Link Capital as a partner and a leader of the BMS crossover strategy.
55:51
She will be focusing series B+ venture investments across software categories.
55:57
She uh Fortune magazine has the story.
56:00
She was a longtime Bessemer investor and now she's heading over to Cross Link Capital to cross over into private growth equity deals.
56:07
So congratulations to Maryo.
56:13
In other news, uh Mirror Moradi's Thinking Machines is offering top dollar to technical talent much higher base salaries than Anthropic and OpenAI.
56:21
Thinking machines will give you 500k anthropic.
56:23
How about a safe 400k open AI? Best we can do is 300k.
56:26
No wonder everyone's jumping ship.
56:28
Um I think we're looking at two different things here.
56:32
Like this is this is like not necessarily for the top researchers who are these like star athletes.
56:38
This is maybe more for just like the general folks who are joining as like software engineers and developers and kind of like members of technical staff.
56:47
But it is interesting to see that uh Meera is uh offering aggressive. Yeah.
56:51
Getting aggressive and offering generous pay packages because there's obviously a lot of work and it's a very small team and even though thinking machines has some great uh great AI researchers on board, they got to have some folks to chop wood.
57:04
Uh and now that Meera is in the is in the business of selling to businesses B2B fine-tuning reinforcement learning for businesses to drive profit, she's got to get thinking machines on Adio customer relationship magic.
57:16
Adio is the AI native CRM that builds, scales, and grows your company to the next level.
57:20
You can get started for free. Adio. comtvPN.
57:24
You can try it out for free on us. So the 462. 5K average salary.
57:31
Thinking Machines Lab offered for technical hires is considerably higher than more established language model competitors Marotti is aiming to compete with.
57:38
So the uh the competition in AI talent rages on.
57:42
We will be joined by Swix in just a few minutes to discuss the AI talent wars in more detail.
57:47
Uh Aiden McGlaughlin, who's been on the show, says uh posts the the survivorship bias photo of the plane with the red dots and Benjamin D.
57:59
This is in response to Benjamin Dracker who says the open AI brain drain poaching by other labs kind of confirms that the smartest minds in AI don't actually believe open AI is on the cusp of AGI but rather that it's going to be a drawn out slugfest for several years at least and the economy still matters.
58:18
In other words, if you really believed that open AI was right around the corner from AGI and once there AI disrupts the entire economy to a post scarcity economy, why would you leave that organization right now in exchange for money?
58:29
It's a good question, but Aiden is of course saying that the true believers are still there. They haven't left.
58:36
It's only the mercenaries that left.
58:38
And he's got we got to have Aiden back on the show.
58:40
That was that was a really fun guy. It's fantastic.
58:44
Um, also in other news, they're saying Meta is now preparing a huge pile of cash to acquire Sweet Baby Rays. Sweet Baby Rays.
58:51
I would love to identified as a potential acquisition target. Yeah.
58:57
I mean, I wonder what their AI talent pool really looks like.
59:01
I mean, uh, Elon sells the Tesla tequila. Chamat has tequila. Maybe Zuck.
59:06
He, you know, he the meta.
59:06
He says he doesn't use caffeine, you know. Maybe Sweet Baby Rays. Sweet baby rays. Bands. Ray Bands. Sweet Baby Ray Band. Metas. There we go.
59:18
They come with sauce in the glasses. No, no, no, no, no.
59:22
When you're barbecuing, that's the ultimate augmented reality use case because it can take a picture and say, "Okay, I've set a timer automatically. Check the temperature. How long?
59:32
Oh, how long should this particular steak be cooked?" Sets the timer.
59:36
Lets you know, starts buzzing.
59:39
Little reminder when it's when it's ready.
59:41
This is actually a killer application. Perfect. Perfect ribs.
59:44
You can only make perfect ribs with the sweet baby Ray-B bands.
59:49
Well, if you're interested in in uh getting in on the action of the Mag 7, the public companies, head over to public. com.
59:54
Investing for those who take it seriously.
59:56
They got multiasset investing, industryleading yields, and they're trusted by millions.
59:59
Um so Elon Musk is starting a new party.
1:00:01
says vox popula vox de 40 80% of people voted for a new political party that actually represents the 80% in the middle. Uh interesting.
1:00:12
So we'll see if he actually does it.
1:00:14
There is a poly market tracking whether or not he will do it. It has been surging.
1:00:18
Um we will be tracking that.
1:00:21
Uh third parties not very successful in American history but you know not yet. We're in a new era.
1:00:25
Maybe Elon will will chair it.
1:00:27
It's unclear if he wants to be the head because he can't be the president because he's not a native born citizen.
1:00:34
Um, so who would he put in charge or who would he back?
1:00:39
Andrew Yang was running a third party for a while, the forward party.
1:00:40
Do you remember this whole thing?
1:00:42
That whole that whole thing?
1:00:45
I I do remember it at a high level. Yeah.
1:00:47
I mean, he had the backing of like Sam Alman, some really big tech people.
1:00:50
big tech people. um and he was running as a Democrat and then he uh and then he dropped out and kind of like was not seen as like an outsider and then he started the the forward party so not left not right but forward uh and didn't
1:01:04
really get that off the ground fully but um yeah I wonder what the Elon Musk uh I wonder what the Elon Musk party will be called I think the I think it's called the America party Oh the America party okay that won't be he said uh he called it Uh uh that was his words. What Elon
1:01:20
What Elon Musk's America party could look like. Yeah.
1:01:25
He seems he he he clearly hated politics. Yes.
1:01:28
Got out, but he's angry enough about the state of things that he wants to potentially get back in the game.
1:01:35
Uh but one, uh we're going to have uh Zack Kukov on the show in two minutes. Great.
1:01:42
Um he's going to be breaking down the big beautiful bill for us. Great.
1:01:45
What's happening on the ground in Washington?
1:01:47
One thing that was included was the Invest America Act uh which got approval from the Senate as part of the bill.
1:01:54
Uh it's on the verge of becoming law.
1:01:57
The first country on the planet to make sure that every child has a stake in the game.
1:02:00
Real private account ownership compounding in our best companies from birth. Very cool. So very very cool.
1:02:07
Great work from Brad Gersonner.
1:02:09
And otherwise, uh, we have some breaking news from TVPN, which is that we have partnered with Graphite. Uh, graphite. dev is the website. Let's pull it up team.
1:02:20
Uh, I've been very excited about this.
1:02:23
Uh, I was a happy graphite customer at my last company.
1:02:26
My entire engineering team was running on it.
1:02:30
Graphite helps teams on GitHub ship higher quality software faster.
1:02:33
It is the platform trusted by 45,000 developers at top engineering orgs like Harvey Ramplet. Who else we got?
1:02:46
We got a complexity Harvey Aana Data Dog for Da Dog the browser company Duck Duck Go. Let's go. A lot of them.
1:02:57
But um if you're building on GitHub like I'm sure you are, go and get on graphite. Uh graphite.
1:03:02
dev dev is the website and we'll have to have Merryill the CEO on the show very soon but very excited to partner with them and um you know Tyler will be uh running our TVPN or on graphite as well so it more to come.
1:03:20
Well we have our first guest of the show Zack Cukov for having me live from Mr.
1:03:24
Washington Washington correspondent Mr. Washington.
1:03:30
I even set up I don't know if you can see that's great.
1:03:32
Oh, that's the uh monument right there. Wow. No way. There we go. Beautiful. That's great.
1:03:37
We'll just have our production team zoom in.
1:03:42
Zoom in just on the monument so you won't be on camera. Perfect.
1:03:44
Just I'll get out of the way, but you can see the good stuff behind me.
1:03:47
That's the important thing.
1:03:48
So, yeah, break it down for us.
1:03:51
The uh the Wall Street Journal has a story.
1:03:52
Senate passes Trump's mega bill bill after all night session.
1:03:54
tax cuts and Medicaid spending reductions head to the House with a July 4th deadline reachable.
1:04:01
What's interesting in the bill?
1:04:03
How likely do you think it is to pass?
1:04:06
What are the major reads on this?
1:04:08
And then we want to get your your takeaway from like where the dividing lines are around the bill. Yeah.
1:04:11
So, there's a couple trip lines.
1:04:14
Uh the first trip line, I would say you guys have been tracking the moratorum, the 10-year moratorum on AI development.
1:04:22
This was the on AI regulation.
1:04:22
This was the first it was in.
1:04:24
Ted Cruz put it in, then it was out, then it was back in. There was a compromise.
1:04:29
There was a five-year agreement instead of a 10-year agreement.
1:04:32
And Blackburn has been sort of a key driver of getting it back out again. It's out.
1:04:37
Now it was done 99 to1 in the Senate, almost unanimous in the Senate to get this out.
1:04:42
This was a big win for Daariothropic, probably the number one beneficiary.
1:04:47
And by the way, somebody who certainly used a fair bit of political capital pushing for this.
1:04:54
That's fault line number one.
1:04:56
I would say fault line number two. Stay on that.
1:05:00
It's it's a it's a potential ban on AI regulation that was removed.
1:05:06
And so the current bill does not have a has a does not have a moratorum on regulation.
1:05:11
So it's kind of a double moratorum on federal regulation. That's right. California.
1:05:19
Yeah, states, whatever, who want to Marcia Blackburn, her home state, Tennessee, passed last year something called the Elvis Act, which uh was cutely named, but basically said, look, if you are building like a, you know, an AI audio business, you cannot use any likeness of a real performer's sound.
1:05:40
So, if you want to have uh John singing the blues or Jory singing the blues, I cannot do that without a prior deal.
1:05:46
That's because in Tennessee, Nashville is a dominant industry.
1:05:48
The music industry is a dominant industry.
1:05:50
And so part of Blackburn's opposition to the moratorium, which would have prevented bills like that from taking effect, is that it ended up superseding the work that her home state has done to protect its dominant industries.
1:06:02
By the way, net net this moratorium, the time that they got to a compromise was so watered down really wouldn't have impacted Tennessee or California anyway.
1:06:12
But it was a very dramatic swing from in, out, in, and now out once again for good.
1:06:21
And and what do you what do you expect uh at at the state level around AI safety going forward?
1:06:27
Is this going to is the fact that this was removed going to inspire certain governors to say, you know, m make this a part of um their platform?
1:06:37
And I guess like I I want to know more about like if I'm 11 Labs and I'm and I'm training a voice agent and I can't be sure that some clip from, you know, Elvis or, you know, some Justin Bieber worked its way in there.
1:06:50
Uh, does that mean that I just turn off Tennessee at like the DNS level or do some geo fencing or or is it more risky than that?
1:07:00
um where I I have to I I have to comply kind of at the national level and they're kind of forcing me to comply everywhere or can I just pull back out of that state because we've seen this with with uh Apple intelligence not being available in China, not being available in Europe for example.
1:07:15
But yeah, what would the impact on something like that be?
1:07:18
Well, think about states in two categories, right?
1:07:20
There are some states that are so large and by the way the analog that I would think about here is what happens in the textbook market.
1:07:27
So if you're a textbook publisher and you want to write some textbook about biology or whatever, if the largest states in the country, this category one, pass a law mandating something needs to be in a textbook, suddenly you're going to comply nationally because they are so large they warp the market. Exactly.
1:07:42
You have to That's right.
1:07:43
And so if California tomorrow says, "Look, we're passing a European style bill to severely curtail," which I think would be a disaster, to severely curtail AI development, that would in effect be a national bill.
1:07:55
And part of the, in my view, uh, mistake that's happening tactically right now is this self-imposed July 4th deadline is so restrictive that it forces all of these more nuanced negotiating points to fall by the wayside, right? It becomes blanket.
1:08:12
Either it's in or it's out.
1:08:15
There's just not time to get to a more nuanced middle ground.
1:08:17
So 11 Labs if Tennessee has the Elvis Act probably says look we just squirt around Tennessee.
1:08:23
I don't I haven't spoken to 11 Labs but if California passes then suddenly you have the whole country warped by this and Gavin Newsome and the California legislature are certainly ambitious enough to try as they did last session uh to try to pass something that would impact the whole country. Yeah.
1:08:38
Wasn't Wasn't there news out of California this week or or last week around some of this stuff? Yeah. Yeah.
1:08:45
There's there was a huge fight which had an unlikely set of bed fellows where you saw Andre and others leading the anti- legislative anti-regulatory brigade against some other folks who are much more pro- cretailing.
1:08:56
By the way, I I will just say the other interesting takeaway for tech, not to totally dovetail away from your question, but the interest other interesting takeaway for us, I think, in this is the deciding vote in the Senate on the overall package, the reconciliation bill was JD Vance.
1:09:13
And so if you're a Vans looking at 2028, right, you already see that Accaziocortez and others, AOC and others are going to use his deciding vote on this bill, which includes huge amounts of Medicaid cuts that Tom Tillis from North Carolina in particular thought were politically toxic as a wedge against him when he comes and runs for president in four years.
1:09:33
So sorry that was you didn't you asked a question about California and I dovetailed it back to the federal level, but that's the Yeah. Yeah. Yeah.
1:09:40
That's the big uh the big shock.
1:09:43
So, you asked about trip wires.
1:09:45
The AI trip wire was one.
1:09:47
Medicaid cuts, which are politically really toxic, right?
1:09:48
Because who wants the idea that you voted for the government to take away your healthcare?
1:09:53
That's another huge trip wire.
1:09:53
And then there's a package of things that maybe we care about that are nuanced, like for example, IRA tax incentives for clean energy, right?
1:10:02
Or new taxes on wind and solar from China.
1:10:05
solar from China. These are all things that might enrage the Republican the the very conservative Republican faction in Vig in the House which now has to still pass the bill and now complicates the negotiating posture because the Senate
1:10:17
watered down the Senate preserved the IRA tax incentives and watered down some of the tax cuts, got rid of some of the tax cuts on Chinese imports, which means basically you have a bill that's actually financially harder to pass because it's less balanced. And at the
1:10:29
And at the same time, you've now alienated a huge faction of the GOP.
1:10:33
the super conservative GOP side of the party who still have to vote for the bill to happen.
1:10:39
So, it's up in the air if it passes or not.
1:10:41
Uh, can you tell me about the EV tax credits?
1:10:44
I know Americans are struggling more than ever to buy naturally aspirated V12s.
1:10:49
I'm hoping that there's some move there to subsidize things like the Ferrari 12 cylindry, the Dol Trindi. It's a multi00,000 car.
1:10:58
Even if we can't subsidize that, maybe we can just take away the subsidies from electric cars to tilt the to tilt the playing field in favor of the of the larger engines, the supercharged V8s, the twin turbos, the G63s, these types of cars that Americans, you know, deserve to be driving.
1:11:15
So, what's the news on the EV credit front?
1:11:18
I agree with you that the backbone of the American economy is the Ferrari, right?
1:11:23
It's impossible to me to imagine a world.
1:11:25
Thank you for the laugh track. I appreciate that.
1:11:27
Where a bluecollar factory worker cannot own a sports car is beyond me.
1:11:31
This bill doesn't address that. I wish it did. Be great if it did.
1:11:36
It's not going to solve that huge unsolved problem.
1:11:38
What it will do, by the way, is continue to drive a wedge between Elon and Trump. Right?
1:11:45
Then you saw Elon take the Twitter.
1:11:45
I imagine you guys probably mentioned it earlier.
1:11:48
You saw Elon take the Twitter and start to really crusade against people voting for this.
1:11:51
That is in part because at least the House version of the bill was much worse on a lot of the subsidies for electric vehicles than the Senate version.
1:11:59
The big challenge is going to be going into now the conference committee.
1:12:02
It's going to be how do they reconcile these two challenges so that they can hopefully tilt the market back in favor of course of Ferraris and supercars nonetheless while also not alienating one of the two power centers of the GOP at least today in Elon who could easily primary somebody who votes against.
1:12:19
Yeah, that makes a lot of sense.
1:12:20
What what what about uh how how are the Democrats feeling about EVs these days?
1:12:24
Because it used to I used to feel like it was a Democratic cause to be pro-E incentive.
1:12:32
Let's save the environment.
1:12:32
Let's get people out of the gas cars.
1:12:34
I'm joking about the V12s obviously, but um but now it's become maybe let's punish Elon, but is this a cut your nose to spite your face situation for the Democrats?
1:12:43
I think it's a rough spot.
1:12:46
I mean, you guys remember when Elon wasn't invited to the big EV summit under the Biden administration, right?
1:12:51
That's part of what kicked off his political awakening is he he hasn't used unionized labor and so he was excluded from this EV summit.
1:12:59
There's been a huge polarization, a negative polarization against EVs among not only Dems in office, but the Democratic base, too, which is to all the points you brought up quite ironic.
1:13:08
You're not going to find many Dems willing to cross the aisle in support of electric vehicles to save this bill.
1:13:14
This is a bill that will live or die by Republican whip count and Republican unity.
1:13:18
If the whip count isn't there, it's going to be really hard to get it done before July 4th.
1:13:24
What about uh Elon's America party?
1:13:24
You think that's real or or just uh exciting?
1:13:32
Uh are those my only two options? Real or exciting? There's no door.
1:13:34
There's no door number three.
1:13:36
Yeah, I mean it's obvious like you know we were talking earlier uh third parties haven't haven't historically done well well he has the America pack so when you say Elon could primary someone he could potentially primary a Republican on the Republican ticket with the American pack and he could almost like be kind of like a head fake like oh this is part of the America party but it's not really a third party.
1:14:01
Is that more what we're thinking here how this plays out?
1:14:02
Yeah maybe maybe one historical analog.
1:14:04
It's not perfect, but think about like a Theodore Roosevelt, like a Teddy Roosevelt with the Bull Moose Party, right?
1:14:09
You think about it's a faction within a larger established party to your point that can serve as an engine to primary people who are in disagreement over Exactly.
1:14:17
Like the Tea Party or frankly what's happening now. Look at Zorona.
1:14:22
MAGA is also like a wing of the Republican party. That's right. That's right.
1:14:27
MAGA, the Tea Party, and now on the left, the polarization from the Zoron and Democratic Socialist world.
1:14:33
All of these folks are new factions in a way that by the way we haven't seen before.
1:14:36
The same way that 50 years ago you basically watched MASH on television that was your only option and now I tune into TBPN while I play Subway Surfers and have 30 other things hitting me at the same time.
1:14:45
This is the political version of that, right?
1:14:48
Where you say, "Okay, it used to be my options were a platform that I accepted wholesale on the Republican side or the Dem side. Now I can be a MAGA guy.
1:14:54
I can be a techright guy.
1:14:57
I can be a populist guy and I'm a new right guy.
1:14:59
And on the left I can be a democratic socialist.
1:15:01
I can be an abundance bro.
1:15:02
I have an infinite menu of things.
1:15:04
The America party is more likely if it sticks.
1:15:07
One more menu on the side of the right. Sure. Sure. That makes sense.
1:15:11
That makes a lot of sense.
1:15:12
What uh what time what do you expect out of the next 24 hours?
1:15:18
So, tomorrow morning at 9, House Rules Committee votes.
1:15:21
That's going to be a key vote to see how this thing does.
1:15:24
What I would say is if this gets done by July 4th, the thing that you're going to have to look to is that 9:00 am Eastern vote to see how much unity is there in the Republican party.
1:15:32
They can barely lose any votes that this thing goes through the House in its current form.
1:15:37
But I think it's unlikely that it's going to happen immediately. It'll happen. This will get passed.
1:15:40
But on their timeline, I don't know.
1:15:42
In the next 24 hours, if you see more senators like Marowski from Alaska who are saying things like, "The bill is flawed.
1:15:48
Please send it back to the Senate so we can work on it again."
1:15:51
That's a key indicator this thing may not have the legs it needs the next couple days.
1:15:56
Yeah, we heard about I mean everyone's saying like it's the biggest bill, there's all this pork.
1:15:59
Like are there any like green shoots where there's some random startup out there that's just super excited because maybe it's not a headline grabbing, you know, $10 billion thing, but it's like that's a hundred million that I can go after in energy or in defense or in ship building or anything.
1:16:18
Have you seen any of these?
1:16:20
Like they're called riders, right? Yeah. Yeah. Yeah. It's a good question.
1:16:22
Um Biden was famous for stacking a lot of these.
1:16:26
Like the inflation reduction act had a million ways you could go participate. This one's got fewer.
1:16:29
I'll say one funny piece of pork and I'll say one actual opportunity.
1:16:33
The real opportunity I would say is for fintech folks who are looking at the Trump accounts, right, which is giving huge amounts of new Americans who are underbanked or not banked before access to capital for the first time.
1:16:45
If I'm Robin Hood, I am loving that's a big change for me.
1:16:49
Funniest piece of pork in the bill.
1:16:51
Marowski, who was the key hold out in the Senate from Alaska, got a tax exemption for uh Alaskan native Alaskan uh whalers.
1:17:00
She said basically native Alaskan whalers need to be Yeah, that was honestly big delivery for uh for her constituency of people who kill whales in Alaska.
1:17:09
Let's give it up for the whalers.
1:17:11
They don't get enough attention in Washington.
1:17:12
I mean, we've we've talked about this before, but you know, energy is really important for the AI data center race.
1:17:17
Whales, they produce oil that you can burn. Sure.
1:17:20
Wasn't it Chase who said he would he would burn whale oil?
1:17:25
He didn't say he would do it.
1:17:26
He said that the hyperscalers are so desperate for energy right now that they would burn whale oil.
1:17:31
It's more fun to think about Chase launching Crusoe for whales. VCs are the new whalers. Yes. Yes. They really are. And the old Whalers. And the old Whalers. That's true.
1:17:43
Zach, this was so much fun.
1:17:45
We'll have to have you call in tomorrow or Thursday.
1:17:47
Yeah, when there's more news. Please. I'll be around.
1:17:50
It's my first wedding anniversary, so I'll be around.
1:17:51
Plenty of time to call in. Thank you very much. Incredible. Incredible. We'll talk to you soon.
1:17:56
Thanks for having me, boys. See you soon. Yeah. Bye.
1:17:58
Up next, we're going back to the AI talent wars, talking trade deals, talking$und00 million offers with Sean Swix. How are you doing? There he is.
1:18:11
with the suit looking fantastic.
1:18:14
We You wore a suit last time and we didn't give you enough credit because you were at your own conference uh the AI World Fair and I was just This is a sign of great respect. This is a great respect. So, thank you. Thank you.
1:18:26
And it's always great to great to catch up with you.
1:18:29
Um how how did you process the news over the weekend? What jumped out at you?
1:18:34
I saw a couple of your posts.
1:18:36
You were breaking it down.
1:18:36
Um but but uh h how would you tell the story?
1:18:39
How how how big is this for Meta?
1:18:41
How bad is this for OpenAI?
1:18:44
Um how are you feeling about the trade deals that we've seen break loose over the last few days?
1:18:49
I feel like people who have been in discussing like sports trades uh might be way more equipped to to discuss this.
1:18:59
this is like relatively new in in AI land and like honestly just like absurd uh you know like no like get that back guys like you know like I'm I'm happy you're getting paid but like also uh it definitely feels like for uh the a lot
1:19:14
of the AI researchers I talked to like like this is a little bit um sort of like a panic move and and in some extent warranted right like I think that Meta has um felt like it needs to do something different than what it has been doing especially leading up to Llama 4. Uh and this is different. This
1:19:31
Uh and this is different.
1:19:31
This is very very different.
1:19:34
Uh this is something that only Zuck can do I think.
1:19:37
So there was just just for some extra context there was new reporting from uh Kylie over at Wire just now saying that uh you know the there's been some back and forth.
1:19:48
What are the actual offers?
1:19:51
Obviously a lot of them are different, right?
1:19:52
It's not just like a set offer that everybody is getting.
1:19:54
If you join Meta, you get $100 million immediately.
1:19:59
In a briefcase, hand delivered.
1:19:59
I don't think that's what's so apparently some of them are up to 300 million over four years with more than a hundred million in total compensation for the first year.
1:20:08
This is Wired's reporting, right?
1:20:10
Mark Kerman was saying more like 10.
1:20:10
She said she needed to see an offer letter.
1:20:14
But the funny thing here is there's an OpenAI staffer who was responding to this anonymously and they said that's about how much it would take for me to go work at Meta. shots shots fired. Wow.
1:20:24
But I mean they have set right so now they have unfortunately they have set their own price. It is public now.
1:20:31
Um I think there was some back and forth between like whether or not Sam appearing on his brother's podcast was like a very calculated move and like was he dishonest? Uh this is not my words.
1:20:41
Is this what it seems like he might have been honest based on this reporting.
1:20:46
Like it seems like it's real.
1:20:46
I think it look in that ballpark, right?
1:20:48
like whether or not it's exactly 100 plus minus like it is at least eight figures and that is like far and far and beyond uh what any anyone else had been getting prior basically and it it is in there at least in the relative range of what Tim Cook and Satia made all of 2024. Yeah. Which is which is crazy.
1:21:10
Also with the 100 million number, it's so round that it just jumps out to you.
1:21:14
And I think it's much more viral than 112 or 89.
1:21:19
Like 100 is just such a round number.
1:21:21
It's such a big amount of money.
1:21:21
It's just instant generational wealth. Uh for sure.
1:21:25
Uh I think like that's exactly um it's weird that the calculus of like does this make sense for meta?
1:21:31
Yes, it does because uh I think uh you guys covered my uh analysis before that like basically they just cut VR spend and reallocate to AI spend and they have the funding.
1:21:43
They have the equivalent of anthropics entire annual spend. Yeah.
1:21:47
Meta's at all-time highs right now, right?
1:21:48
The the market and so and so it doesn't really m if if you could move a $1.
1:21:54
8 trillion company market cap by a couple points, you're doing pretty well. Yeah. Yeah.
1:22:01
Yeah, Zuck is no stranger to betting like 1% of his company, right?
1:22:04
To totally against the future.
1:22:06
Um, and he like I think of uniquely of of a lot of the sort of fang or what is it?
1:22:10
I think a lot of people are trying to come up with like new acronyms for like the the Magnificent 7.
1:22:15
So, I like the one that's uh among us. Among Us.
1:22:17
Yeah, Among Us is the best acronym.
1:22:20
Uh, there have been some wild acronyms flying around you're talking about.
1:22:24
like Zuck is uniquely empowered as as like a you know founder mode of the of the trillion dollar companies uh to actually do this and like so like uh it's really amazing that the leaders he's hired and the people that he's managed to hire.
1:22:35
Um I I would say that uh you know to to one part of your opening question like how effective is OpenAI? I would have said no.
1:22:41
Uh not not that affected.
1:22:44
They have like super deep bench.
1:22:45
Um even like the people that you don't really hear about like are actually like fantastic leaders in in their own right.
1:22:51
Um and so like in some in some in some extent opening eyes is like very robust to to already sort of leave like departures already.
1:22:58
Uh but the email from uh Marchen over the weekend I think was also like oh like I wouldn't I wouldn't say that if I were him.
1:23:08
I don't know if you you guys have the the saw of you. Oh yeah.
1:23:10
that knowing that that memo would immediately leak, it felt emotional and defensive versus, "Hey, John John's phrasing was was or or at least the way that you you said if you phrase this as, hey guys, we have billions of people coming to our website monthly.
1:23:33
We have hundreds we have billions and billions of revenue.
1:23:39
We are the category leader in the space.
1:23:41
We are doing something different than Meta is doing and we will continue to dominate and now is the time to rally.
1:23:47
This is the fourth time this has happened.
1:23:49
Maybe fifth time this has happened.
1:23:51
So we survived Anthropic and Daario leaving.
1:23:53
Don't forget surviving Elon Elon leaving.
1:23:55
We sur we survived Ilia leaving and we survived Meera leaving.
1:24:00
And there's four spinout direct competitor co-founder companies.
1:24:03
This is a couple people leaving.
1:24:05
We're going to get through it.
1:24:07
We've been Look at the growth chart.
1:24:09
We're growing, growing, growing.
1:24:11
I would have Yeah, definitely gone from a source of strength there. I don't know.
1:24:14
But yeah, what is your thought on like the narrative that like um OpenAI seems Lindy at this point incredibly resilient, able to lose key people and just keep plotting forward with, you know, iterative releases, models get a little bit better, UI on the app gets a little bit better, deeper research comes out.
1:24:34
They're just always staying on the frontier, just one step ahead of everyone else in that core chat product.
1:24:40
And yeah, maybe they maybe they might not be able to dominate in, you know, the the clawed code ecosystem immediately.
1:24:48
They got to make some plays to make that happen.
1:24:49
They definitely want to, by the way, they want to. Yeah, for sure.
1:24:53
for sure. But in terms of the core thing that's clearly dominant and huge and like gamechanging and profitable like they're doing great and they should and they should kind of uh they should lead from a source of strength there u
1:25:06
because they've been through so much and continue to deliver on that front and and and this doesn't seem like a cause for like the narrative around this this leaving is that okay Meta is going to catch up to a frontier LLM a reasoning model. They're going to have great image
1:25:21
They're going to have great image generation, but does anyone think that, you know, in three years time people's f people's home screens are going to switch from chat GPT to Meta Chat? No, I don't think so. But what about you? Yeah. Yeah. I I don't think so.
1:25:37
Uh keep in mind also Meta has its own institutional issues with AI products.
1:25:43
Uh I think they launched this like um AI chat feature that where every chat was public.
1:25:47
Uh I don't think I still think it's still there.
1:25:49
Well, I think it was accidentally sharing a box you had to check.
1:25:54
Yeah, people were accidentally sharing it.
1:25:55
But it kind of speaks to like yeah, when you're dealing with trillion DAUs, like you're going to have weird stuff happened.
1:26:00
But yeah, uh I would say that Sam would be the first uh of many people to acknowledge that there's such a thing as a 10,000x uh engineer and 10,000x AI researcher.
1:26:08
Uh but they they also have a very deep bench and it's not just about the bench, it's also about the data.
1:26:14
it's also about the data. It's also about the infrastructure that they've accumulated um that you know the TM um what's their what's their acronym um MSL um yeah meta super intelligence is going to have to build up effectively from scratch because if effectively Yanukun
1:26:30
has uh failed in his in his leadership of of uh meta AI there um and I think like that that's fine like that's that's a pivot like that's that's something that only Zuck can do but like yeah open is a very deep bench and like I think they are um like they're they're they're good there. I would say though that um
1:26:44
I would say though that um it's a really interesting like how well diversified this this uh landscape is.
1:26:52
Most people when I talk to uh you know there's there's often uh you know in AI circles there this discussion of who has the mandate of heaven.
1:26:58
Mandate of heaven meaning they're just like most likely to achieve a obviously uh of the big labs myself excluded uh of the big labs a lot of people are mostly saying deep mind uh is actually not open.
1:27:12
So I think like uh OBI is like in intensely aware that they're being um you know poached.
1:27:18
They're being out competed on on on pricing.
1:27:19
Um they're being out competed on some metrics like uh elements for deep mind.
1:27:24
Uh and they're and open is like extremely competitive.
1:27:28
Like whenever I talk to them, I'm so impressed that like they're not just arrogance.
1:27:31
They're not they're not like resting on their laurels.
1:27:34
They're actually just like we will win on every single front.
1:27:36
Um and I think it's just a very impressive company that way. Then let's flip over.
1:27:40
I mean, MSL is a very it's a cooler acronym than than Llama.
1:27:46
I always love Llama because it's a fun name, but it it didn't have the same like seriousness. Fair.
1:27:49
And fair felt also just like it's it's a little nice. It's not hardcore team.
1:27:55
No, we don't we don't care about fairness.
1:27:58
We We dropped safety from MSL is like going to the majors. Exactly.
1:28:01
Sounds like major league LLM, major league AI training.
1:28:06
Um but but I want to talk about the historical energy analogy because Zuck has been in this position before uh with the shift to mobile.
1:28:13
So when when Meta or when Facebook back in the day was a web browser, they initially moved over to uh to an iPhone app and the iPhone was growing and they kind of made a mistake going a little bit too far into the HTML 5 route, wrapping it in an in an iOS shell, but they didn't build the app natively and they were kind of falling behind.
1:28:37
But then, you know, from 2012 to 2015, Mark was extremely aggressive about acquiring Instagram and WhatsApp, rebuilding the team, hiring some great people.
1:28:48
This is when Adam Masseri and andrew Bosworth came in and really really built out monetization in the feed, so they had more money flowing through the system to actually pay for all this stuff.
1:28:58
And then yeah, we were covering yesterday that the the Zuck did a talent raid in the 2010 era of Google because he just wanted to get better at ads basically and then he built the greatest ad engine of all time. Yeah.
1:29:11
Uh so yeah, I mean yeah, how are you how are you interpreting like the comeback story and the fact that like Mark has also been here before?
1:29:19
I think that it's interesting how you have like Zuck has almost has to fail first and then he overcorrects and and goes really really hard and wins.
1:29:30
Um in contrast I think like VR never really had a failure because it didn't really uh you know have that had that uh uh that much of a of a success to begin with.
1:29:39
But like um yeah like um MSL could only have happened after the comparative disappointment of Llama 4.
1:29:48
Um if Llama 4 had had proportionately the same traction that Llama 3 had, I don't think we'd be sitting here today. Totally.
1:29:56
So actually this is like net positive that we went through that because they were like oh like this like we proved out that this is we've tapped out in this direction and we just need a whole new team uh and we need to scale things in a very different way and probably we just need to poach from the other labs. what they've done.
1:30:13
What do you think the Oh, sorry. Yeah, continue.
1:30:16
It's interest It's also interesting though that like what their their current strategy was their third choice.
1:30:20
Um and somewhere in the middle either first or second choice was perplexity and that would have led um that would have led um Facebook or or Meta in a very very different direction more consumery than right now which is building a lab.
1:30:38
What do you think the postmortem on Llama 4 is?
1:30:40
Llama 4 is? Is it just that they over rotated towards pre-training massive model behemoth these like you know not focused on reasoning not focused on all the little tricks and tool use and stuff that actually creates a delightful LLM
1:30:58
experience or chat experience these days or is there something else going on there like how do you tell the story of of what they got wrong with Llama 4 if anything I don't want to bag Llama for too much I by the way I put in a a thing in the chat for for people to compare. Um, Scout and Maverick were great and
1:31:12
Um, Scout and Maverick were great and then the the shenanigans around Ellen Marina were regrettable but like understandable honor at mistake whatever.
1:31:22
Um, but uh it seems that Behemoth will just never be released. Interesting.
1:31:26
Uh the assertion again behemoths was never released so this is all rumors uh was that they tried to basically adopt whatever that Deep Seek did for their large uh fine grain uh mixture of experts model and it just didn't work.
1:31:41
They didn't they didn't portrain it enough whatever.
1:31:42
Um we will never really know the answer.
1:31:44
Uh I think like there's probably also organizational issues around what data they can use.
1:31:52
Uh I I know this for a fact just because like I I've heard issues around like honestly like at a big corp when your lawyers get into the mix of training data issues that can get very inconvenient.
1:32:06
Um when uh you know I think like comparatively anthropic when they won their recent case uh they were more like you know seek uh you know bend the laws and seek founder mode founder mode ask for forgiveness.
1:32:21
forgiveness. I think now that we actually have a little bit of ruling around like okay like pre-trading on on copyrighted data is actually is transformative use therefore it's fair use therefore it's fair game go ahead uh actually we'll see but you the case is
1:32:34
still trying whether that you need to buy the underlying thing which could still be egregiously you know it could be incredibly expensive to to acquire the copyrighted material not at all you don't think you don't think that'll go through no no whether or not you underlying thing is fine. It's actually
1:32:51
It's actually just the the issue of copyright whether you have to pay per use.
1:32:54
Uh it's actually pay per inference.
1:32:56
Um now that that issue is out of Yeah.
1:32:59
But I I was saying the anthropic case is still like they had 7,000 7 million books that they obtained through that they don't that they didn't purchase and the the pirated ones the pirated ones.
1:33:11
And so it will be interesting if labs have to effectively strike deals with publish you know I don't know it's it's still in the course they they are doing that uh they probably just just should do that.
1:33:24
Uh it's a onetime cost comparatively comp compared to the value that they can extract out of the books.
1:33:30
The books are just guaranteed to be underpriced because the books are priced as though they're sold to one person whereas you know a model is going to be used by millions.
1:33:38
So like there is there's just there's just no like it's fine just pay it.
1:33:41
Um the the piracy thing is a is a legacy of books three which is a data set that was created in a in by let's say someone who was very pro piracy.
1:33:48
Uh uh it was it was a copy of books one and books two which is actual uh licensed uh public domain books.
1:33:56
Anyway, so so long long story short, like I think I think those uh cases will be resolved and like um I don't I I think that Llama and Meta's strategy there were probably affected by the quality of the data that they had um and the fact that they just couldn't they they like sort of tight they were playing one hand tied behind their back whereas all the other labs were not. Yeah. Yeah. Interesting.
1:34:15
Well, this has been fantastic.
1:34:17
Last last one one more question.
1:34:19
Any reactions to the news that Apple is working with uh you know potentially working on a deal with Anthropic or OpenAI?
1:34:25
philanthropic wants potentially billions out of it.
1:34:27
Uh very very convenient timing for the open model release uh which is rumored to be soon.
1:34:33
Um uh but like I no reactions uh yet.
1:34:37
I think uh Apple there's a lot of news around Apple that never materializes.
1:34:42
I think Apple's hasn't shown that it has gotten the memo like Zuck has shown.
1:34:46
Um so until until it wakes up like I don't know.
1:34:50
I think like around the time of the Apple intelligence launch last year.
1:34:54
Um I think a lot of people I I was I was commenting that like basically Apple wanted to be the first layer of intelligence and and it wanted to route uh so it always wanted to keep wanted to be Switzerland wanted to play nice between open topic and all the and uh whatever other labs probably not Google but um I think you kind of have to make a bet.
1:35:14
I think you I think you kind of have to go all in.
1:35:15
Like I'm I'm not that interested in in Apple making deals with OpenIthropic.
1:35:19
I still want open eye to do an open eye iPhone.
1:35:21
Uh I don't feel like I don't like the in your device.
1:35:23
Um uh and I and you know I just I just wish they just went all in and and took Apple on head on.
1:35:30
Uh because like all of these are half measures, man.
1:35:32
Like the like we're going to be sitting here like five years from now and Siri is still going to like kind of suck.
1:35:37
Um like it's the same deal as like Amazon making a deal with Enthropic for like smarter Alexa. Where is that today? Yep. I was just asking.
1:35:46
I haven't I haven't actually seen it in the wild. Yeah, great point.
1:35:49
This is really one more breaking one more breaking news for you guys.
1:35:50
I don't know you probably guys probably missed it but um cursor has hired the cloud code people from the topic. What do you mean?
1:35:57
What do you mean they hired like how many of them?
1:35:59
I heard they hired the notion mail team.
1:36:00
I didn't realize that they hired the the cloud code team.
1:36:04
That's an that's an interesting commentary on like the like what is is notion able to keep its people because the notion mail team was uh you know they they just launched it.
1:36:10
Yeah, they just launched it and now they've left trade deal.
1:36:14
But like cursor so cloud code developers have been talking about it non-stop.
1:36:19
I think uh people who are like adjacent to developers really pay attention to this one.
1:36:23
This is as close as anything is to the new being the new cursor. It is cloud code.
1:36:27
Uh people are going absolutely nuts and spending a huge amount on it.
1:36:31
Enthropic revenue is up 4x since the start of this year.
1:36:33
Um and and like cloud code only launched in February.
1:36:38
Uh and now cursor has poached the top two people from cloud code like the creator and the and the PM.
1:36:43
Um I like it's it's breaking news but like in terms in terms of in terms of like talent trades this is a really interesting one because it's at the engineering level not the model level and it's starting to spread out.
1:36:56
It's starting to be like extremely competitive here.
1:36:59
Uh and I think it's very exciting. Wow.
1:37:01
Well yeah we'll have to have you back soon since uh this is this is the top story every day now.
1:37:05
Uh this is this is the real life.
1:37:08
Uh thank you so much for joining. Great to see you.
1:37:11
Have a great rest of your day.
1:37:11
And next up we have Matthew Prince from Cloudflare coming in to TBPN. Welcome to the stream.
1:37:21
Some huge news from Cloudflare.
1:37:24
Uh we'll let him tell us the story. How are you doing? Good to meet you. That's me Juke.
1:37:29
Uh welcome to the stream.
1:37:32
Um I would love to start with um the news today and then I want to go into some of the the history of the company and and uh when you faced similar uh junction points and kind of uh turning points in the past.
1:37:45
Uh so why don't you uh kick us off with uh the big announcement the big news today from Cloudflare? Sure.
1:37:51
So about 18 months ago we started getting approached by a number of content creators saying that they faced a new new threat uh online.
1:38:01
Um, we spend most of our days, most of our business trying to fight cyber hackers, Iranians, Russians, North Koreans.
1:38:08
And so I was surprised when the new threat was AI companies.
1:38:10
And frankly, I kind of rolled my eyes uh at publishers.
1:38:14
I was like, publishers always complaining about, you know, the next new technology.
1:38:16
Uh, and and and and they said, just pull the data, look at what is is going on.
1:38:21
what is is going on. And it was really striking um that uh over time uh you could watch as the sort of deal that publishers had made with Google you know starting 30 years ago which was you can copy all of my content and in exchange
1:38:36
you send us send us traffic that that deal just didn't hold up anymore uh where they were copying the content but because they were providing answer boxes AI overviews uh they just weren't sending nearly as much traffic and and that was the good news. uh as you looked
1:38:47
uh as you looked at OpenAI, it was about uh uh 750 times harder than with the kind of Google of old to get traffic.
1:38:55
With something like Anthropic, it's 30,000 times harder.
1:38:57
And the reason why is because people are reading the derivatives, they're not reading the content.
1:39:02
It'd be effectively like if I asked an AI system, you know, summarize what happened on the technology brothers show today.
1:39:08
I have less incentive to actually go and watch your actual show.
1:39:11
And that means that if you're making money through advertising, if you're making money through subscriptions, if you're just, you know, getting value out of the ego of knowing that people are watching your show, uh that's going away in an increasingly AIdriven web.
1:39:23
And and we worried about that um quite a bit.
1:39:26
And more and more as we talked to publishers, we realized there was something that we could do about it because Cloudflare runs one of the world's largest networks.
1:39:33
A huge amount of the internet already sits behind us.
1:39:36
uh we could partner with literally publishers, everyone from Ade to Ze Davis and everything in between uh to say that let's change the the way that that uh the internet works and let's change the sort of model where if you're going to take content, you should actually compensate the content creators for it.
1:39:54
And so starting on July 1st, uh we flipped the script.
1:39:57
We made it as a default that if uh you're using Cloudflare um that you uh can by default for free uh block any crawlers from taking your content without compensating you.
1:40:08
And I think now the hard work be begins where we have to figure out now the business model around how we get compensation back to these content creators and make sure that folks like you who are who are creating real original content can continue to be compensated for the hard work that you're doing. Yeah.
1:40:22
Well, we used to joke saying that uh the the labs are welcome to scrape our content because we want to influence uh the the outputs of the of the model.
1:40:31
But no, it's a it's a serious it's a serious issue.
1:40:33
We we were covering one of Ben Thompson's pieces about this whole debacle a few weeks ago and it was funny because at the time we were like this is there's so many different stakeholders here.
1:40:43
Who could possibly try to figure out a solution to this problem?
1:40:48
And I obviously the answer you guys are uh a key player in this and you have the scale in order to kind of make changes happen. Can you Yeah.
1:40:57
Can you I guess like break down more kind of like the the kind of the market structure between and and how you see this actually working out and who deals have been done in over the past year because I know uh didn't News Corp do a bespoke deal with OpenAI for that exact thing?
1:41:17
it sounds like you are more equipped to kind of handle like the longer tale but also some of the bigger publishers.
1:41:23
So talk to me about kind of the state-of-the-art of what those deals look like when they're onetoone and then what things might look like going forward.
1:41:31
Yeah, so I think that um first of all uh I I think for some time to come uh large publishers will do deals with large AI companies uh where there might be places that people need help are either small publishers or small AI companies.
1:41:46
um you know and figuring out what has to happen uh there.
1:41:48
But I think even in the case of places where deals have been struck, what the challenge is is, you know, someone like OpenAI is paying a lot of publishers for their content and doesn't actually object to the the basic idea that they should pay for the content that they're using.
1:42:07
Um but but they do object to the fact that they're paying where all of their competitors are getting content for free.
1:42:13
any market that exists has to have scarcity.
1:42:15
And the problem, the reason why there hasn't been a market in this place space is that there isn't scarcity.
1:42:23
And so what changed yesterday is scarcity was in introduced really for the first time.
1:42:29
And I I think the closest analogy is if you think back to, you know, before iTunes launched, um we were all downloading music on Napster and, you know, all the the free um pirate services.
1:42:39
That's effectively how the AI companies are taking data from any content creators today.
1:42:44
And what you need is an iTunes.
1:42:46
You need someone to come in and say, "Okay, let's provide a way where people can pay for that content and they can go back."
1:42:51
Now, now that's evolved over time.
1:42:53
I think our first version in the marketplace will probably be relatively naive and we'll figure it out.
1:42:58
We no one's paying 99 cents for songs anymore.
1:42:59
Uh we've moved to something that's closer to a subscription model like a Spotify.
1:43:03
Um and and and I think that that's what what we will evolve to have over time.
1:43:10
And and if we do it right, you know, I think that we can have a large number of sellers of content obviously, but also a large number of buyers of content.
1:43:16
Both the big sort of general purpose models, but then also very specific models.
1:43:19
And we've got to make sure that new entrance uh can come into this market and have have a vibrant way of of participating in competing.
1:43:27
And so I think that for now, most of the deals will be one-on-one bespoke deals between publishers and AI companies.
1:43:33
You'll need to have some way of enforcing those deals and making sure that if you're charging someone, they can get access, but their competitors can't.
1:43:41
And then over time, I think we'll develop something that looks much more like a robust marketplace where um uh content providers can set sort of what their price is.
1:43:50
is there can be a bid for maybe getting that that content exclusively for some period of time or or or whatever else and that's going to develop uh over over the next uh over the coming months.
1:44:00
Is that the correct I is that the correct model to think that the the content producer would set the price?
1:44:07
I mean, that's certainly what happens on Substack and, you know, the the New York Times and the Wall Street Journal set their price, but I've always thought that this might evolve a little bit more like YouTube or like Spotify where there is a pool of dollars that are going into Chat GPT subscriptions.
1:44:20
I personally pay $200 a month and occasionally I fire off queries that probably land on the Wall Street Journal or the Financial Times and I would be happy for them to send 10 cents or a dollar or split off 50% of what I'm paying.
1:44:38
So you have $100 to kind of, you know, parcel out in the pool and then you see how much inference pulled what data from what what sources and and kind of have it all be driven by uh the like kind of the same volume uh uh like the volume algorithm that you see on on uh on YouTube or Spotify.
1:44:57
Does that seem like kind of what the long-term um outcome might look like or do we need to also get to some sort of like ad driven marketplace because that's ultimately where YouTube how YouTube prices these things?
1:45:11
Yeah, you know, I think it's going to be um I I think that we there's there's a lot that we still have to figure out.
1:45:17
I I I last week um traveled up to Stockholm to meet with Daniel A, the uh the founder of of Spotify, and and it was, you know, it was really um it was it was interesting.
1:45:28
He said, you know, the the ultimate model that Spotify settled on uh I think it's I don't want to put words in his mouth, but but it was basically like it was it was sort of the most naive of all of the different options that he had come up with and and it is works as it
1:45:42
is a pool of funds and then they divide it up based on on basically how many minutes of of content you you consume and there are different rates for music versus podcasts versus audio books and um and and it was fascinating to to see. So you could imagine a version where
1:45:56
So you could imagine a version where Cloudflare effectively negotiates on behalf of all of the content that sits behind us with the big AI companies.
1:46:04
People can opt in or opt out of being a part of that pool.
1:46:07
The more content that's part of that, the more valuable it is to the AI companies.
1:46:10
And then we would look at how we could split that up across how much crawling is actually um done across that.
1:46:16
I think one of the real keys will be how can we give hints to the AI companies on what is the content which is the most valuable to them.
1:46:26
So for instance, if Taylor Swift releases a brand new song and maybe does an interview about it, that's incredibly valuable content.
1:46:32
If you're a I don't know chatbot for teen lonely teen girls, right?
1:46:37
On the other hand, um it's probably not so valuable if you're sort of a chatbot trying to be the world's best doctor.
1:46:44
you know may maybe there's some value in in sort of bedside manner but but largely you know that's not going to be a big deal in your in your care.
1:46:50
On the other hand, if a research study comes out on acetaminophen's effects on on uh you know different types of chemotherapy, very little value for the chatbot for teen girls, but incredibly valuable for the doctor.
1:47:01
And so what we have to do is actually I think on a on an LLM or a AI model by AI model basis be able to give them signals back that say this is the sort of thing that you should be paying attention to, that you should be indexing, that you should be actually um scraping and paying for.
1:47:18
Um and and the same way that in you know the music industry we've got reviews, we've got different types of music that that people are you know you might be into country or you might be into classical and and and it kind of allows you to compartmentalize that.
1:47:29
I think that whole ecosystem has to develop but once that does um something very much like Spotify could be the right answer uh long term and again I think we're in a unique position to pull it together.
1:47:42
Talk briefly about stable coins.
1:47:42
I remember in this Ben Thompson piece he was people have been suggesting this idea that stable coins could be some part of the solution to to or backbone for a new uh business model for the internet.
1:47:55
At the same time while I'm bullish on stable coins someone like a cloudflare can set up the infrastructure to just use traditional financial rails to process payments.
1:48:05
So, I'm guessing that it's not part of of of what you're This was Ben's uh reaction to the the the original sin of the internet was advertising and this idea that Mark Andre when he built the browser didn't build in payments because it was this multi-party discussion to figure out how to actually integrate payments into HTTP didn't happen.
1:48:27
didn't happen. the first MCP spec that's the model context protocol spec for AI agents to interact with each other doesn't include payments but he was kind of saying that it might be on the horizon but would love your take yeah you know I think um the answer doesn't have to be 100% one or or the other uh and so I would guess that there are there are going to be um very
1:48:46
traditional kind of traditional um financial rails that will be ways to pay pay people out but there will also likely be something that is uh blockchain based or or stable coinbased Uh and and again, I think we're looking at um all of the different providers uh that are in this space, including potentially creating our own uh stable coin that would be part of how these these transactions um you know, take place. And I think you're exactly right
1:49:09
And I think you're exactly right that um there's a natural extension from what we're doing here to how we figure out around what the payment rails look like for MCPs because as your agent is going around interacting with things um that's also going to be something that is is either um you know charged for or that you that you actually um pay get paid to do.
1:49:31
Uh and and there's got to be a way for those those transactions to take place.
1:49:36
And to the extent that you can reduce the transaction costs for those transactions as far as possible, uh then that's that's a pretty interesting uh way of of uh of of being able to um process, you know, very high volume, low dollar amount uh uh transactions. So I think that's right.
1:49:54
My only one quibble is that I think that the original sin of the internet is not that it didn't include payments.
1:49:57
It's that it uh was that there was no privacy uh built into it and your IP address reveals far too much about you as and who you are.
1:50:04
We don't even have to get to cookies, but that's a No, we totally can.
1:50:09
What about we we we've talked about content creators, publishers.
1:50:11
What about uh the the apps of the world, the Ubers, the Instacarts that actually have really meaningful advertiser businesses, you know, businesses built on top of them?
1:50:24
Uh do you expect there to be any type of of this functionality rolled out to them over time or or is that something you think they're thinking about quite a bit?
1:50:32
Because if I can go into chat GPT and that's my default daily assistant, let's say in a few years from now and I'm like, "Hey, order me groceries and order me this Uber, uh, you know, I'm not in the app seeing ads potentially, you know, being upsold and and that could be a problem for those types of companies." Yeah, totally.
1:50:51
And I think that I mean um, you know, robots don't click on ads.
1:50:55
uh and and so and so the the sort of model that that part of the um you know revenue stream to support these services uh has to come from from somewhere uh and if today you know Uber is able to um monetize the time where I'm I'm waiting for a car to come by showing me an advertisement um that that means that they can effectively charge me less for the ride.
1:51:22
Um, and so it may be that if that ride is booked through an agent instead and they don't get that that advertising experience, and again, I don't know enough about um, you know, Uber's business model to know how how important that is to them, but if they don't get that, that maybe it is that there's some sort of premium that an agent booked ride receives versus a human booked ride.
1:51:43
And again, I think that that's um that that that's just we've got to figure out what that what that uh what that looks like.
1:51:51
I think that's going to be less up to Cloudflare.
1:51:54
Cloudflare. uh I think it's going to be you know our job will be how do we provide those payment rails how do we provide the systems how do we make it so so that we can have guard rails on as agents interact uh uh with applications that are there that you as the application provider can set those those rules and controls and then and then agents can interact within within that
1:52:14
system and MCP is is sort of the leading candidate uh for providing that but you're you're right that that's a sort of draft version of that that specification and there are a lot of things that have to you know, mature in order for that to be something that is the the real um kind of agentto agent uh system that runs that is likely to run what will be an increasingly AIdriven web. Yeah. Can you talk about any Yeah.
1:52:35
Can you talk about any previous crucible moments for either Cloudflare or just like the internet history broadly?
1:52:42
This feels like a moment where you're potentially trying to I mean deal with a shift in the in the overall internet structure or the incentive structure.
1:52:53
Uh but uh it's a reaction but it's also an action that should reshape the the uh the the relationship between different parties on the internet.
1:53:03
uh are there different moments in the history of Cloudflare or the history of the internet that you think uh we could draw on as historical analogies um to learn from to see how this might play out?
1:53:15
You know, I think that um you know that Google really uh 30 years ago defined what was the business model for the web, which was uh you you create content, search drives traffic, and then you monetize that content through advertising or or subscriptions or or ego.
1:53:36
Um and ego and monetization, not quite the right thing, but you drive value from it in one of those those three ways.
1:53:43
Um, and that that's held up for the last 30 years.
1:53:45
I I I I think that this is um the first time that I've I've seen something where I thought, wow, there there are really going to be um very very very fundamental changes to the structure of how the internet works if we move from a searchdriven model to an AIdriven uh model.
1:54:07
And and I think that um what I hope is that we can actually make this something where um what gets rewarded is not what gets rewarded today, which which largely is, you know, who can write the headline that produces the most um you know, uh sort of inflammatory response uh in in all of us.
1:54:28
You know, who who basically drives the most cortisol uh you know, gets the most clicks and and therefore the most ad revenue or the most subscriptions.
1:54:36
That's that's not a that's not a healthy way of looking at the world.
1:54:38
I think you know the alternative is if you imagine that kind of all of the AI models together are a good kind of assemblage of of an approximation on human knowledge.
1:54:49
If we can identify where the holes are and then create incentives for content creators to actually fill in those holes that actually is going to move humanity forward.
1:54:58
It's going to make the AIS better.
1:54:59
It's going to it's going to be a much more interesting thing.
1:55:02
And that's I think how we're going to we're going to be um creating it.
1:55:05
I I can't think in the last 30 years of of a a as seinal moment.
1:55:07
I think there are some things that are along the way.
1:55:11
I think 2016 was a really a big turning point year um for for a lot of reasons.
1:55:16
It was sort of the year that it went, you know, technology went from being able to do no wrong to being able to do no right.
1:55:21
Uh and and sort of rise in populism um around the world.
1:55:24
You have you have Brexit in the EU.
1:55:26
You have the first Donald Trump election.
1:55:29
You've got Xi and Modi consolidating power in Asia, the Filipino election.
1:55:33
Um, and I think that that was a point in time where, you know, also driven by um Snowden uh shortly before that in 2015, uh, that that a lot of the world lost faith in kind of the US model of internet governance and it wasn't quite ready to adopt the Chinese model, but that's that's definitely been a shift uh, which is substantial and I think we're still watching that uh, play out.
1:55:58
that's that I think is probably still remains the biggest threat to the overall internet.
1:56:01
But this shift to AI as well um you know again I I think is is is a real challenge but it's also an incredible opportunity if if again we can make the rewards incentives for content creators not be who can produce the most most cortisol but who can produce the biggest sort of incremental improvement in human knowledge.
1:56:20
um that that would be an incredible win and uh and something that that um you know I I don't quite know how to get there yet, but I I think yesterday or today and uh you know we took in July 1st we took a a real step in uh in in moving in in in that direction and making sure they compensate content creators can be compensated for for really producing uh information which is valuable to advance human knowledge.
1:56:46
What do you think about the future of investigative journalism?
1:56:51
Uh there's been a big shift in just content creation into independent creators, but a lot of that is actually downstream or separate from the type of work that investigative journalists do.
1:57:03
I'm thinking of like Seymour Hirs hanging out at a local army bar for a year before he gets one scoop that he can write a book about.
1:57:10
and and he and these investigative journalists, they tend to, you know, John Keroo was on the payroll of the Wall Street Journal for years.
1:57:17
He was writing stuff, but then he blows the Theronos case wide open, eventually a book and then a movie.
1:57:22
Um, and and I and I imagine that with artificial intelligence, the value of a scoop could actually go up because you could use an LLM to read through the clickbait repurposed version of the article to find, wait, their sourcing actually, this person posted it on the internet earlier.
1:57:40
They're the one that actually did the hard work to get the scoop.
1:57:43
they should get 90% of the value and 10% should go to the repurposed ragebait version as opposed to the economic model now which is maybe you get 90% more clicks on the on the viral you know repurposed content and the scooper actually gets 10%.
1:57:58
Do you think that's possible or or how do you think that investigative journalism will evolve?
1:58:04
Yeah, this I I don't know this is an answer to your question but but it but it's um I thought it was I thought it was interesting.
1:58:10
As I said, I met with Daniel A um up in up in in Stockholm and um and and he told me he said, you know, we were talking about this and he said, you know, one of the things we do at Spotify is we actually surface all of the different queries that are sent to Spotify that we don't feel like we have a really good response for.
1:58:29
Um, so if somebody searches for, you know, I want a a a song to a reggae kind of beat with it's all about how much it sucks when your sister steals your car and your dog.
1:58:42
Um, like not to get too specific, I'm not again I don't know what what what these things are, but there are people there musicians who take this list and they actually write songs based on what people are looking for. No way.
1:58:58
and that that they make tens of millions of dollars a year doing just exactly that.
1:59:02
And there's something about that that I think is really beautiful.
1:59:04
Um as opposed to the alternative of just you know just trying to do a little of more meto of what everybody else is doing like taking a different twist using signals from the market in order to go find that.
1:59:18
So you know again that's not investigative journalism by any means but I think it is actually advancing kind of the sort of good of humanity.
1:59:27
It's taking what a place where there's a a missing need and it's filling that missing need.
1:59:33
And so I am hopeful um you know and and relative I mean I'm a relatively optimistic person and and I'm hopeful that if we get this right that we might be on a golden age of of content creation where people actually are rewarded for doing things that advance human knowledge as opposed to just doing things that produce as much cortisol as possible. Yeah.
1:59:52
The the idea of a knowledge bounty, right?
1:59:54
It's like, hey, there there's there's groups of people or a single entity online that will just pay you to produce really high quality information around this specific topic.
2:00:05
And this the internet always did this sort of organically, but it would just be somebody spinning up a blog or maybe an Instagram account or something like that and they would just start blogging about something that was niche and then an entire network would form around it and other people would contribute to it.
2:00:20
So, I think there's quite a lot of precedent.
2:00:22
Uh but it would be exciting to like put that reward out before and not saying hey you have to do this and then maybe you'll get some ads associated with it later but like basically like the the pool of capital is here and you just have to deliver that knowledge that that people already want and we know it because they're searching it on LLM and things like that.
2:00:42
Can I I think that the um I think I think I'll tell you the sort of black mirror kind of dystopian outcome uh if we don't get this right which is I don't think that I don't think journalism goes away.
2:00:53
I don't think that research goes away.
2:00:55
Um, but you could imagine a world in which we go back to something kind of like the Medachche where there are five big AI companies and they basically employ each of them all of the journalists, all the researchers, all the academics to be able to do the research to feed their systems because you're still going to have to have original content.
2:01:13
People are still going to have to be able to produce that in order for the AI systems to, you know, fill in the holes in their kind of effective block of Swiss cheese that that is that is their model.
2:01:22
Um, but but I I think that that's that's a bad that's a pretty bad world where where again because you'd then probably have like the conservative AI and the liberal AI and I mean all these things that we've kind of that the internet kind of broke apart.
2:01:35
You could imagine AI as a system that is just this massive consolidator uh in the future.
2:01:40
And I I think that's a bad outcome.
2:01:42
think that's a bad outcome. And I think a better outcome is if we can figure out um how how can we reward people for again driving knowledge um but but but not get to a place where all of them have to be employed by by one of the five you know big yeah I actually talked
2:01:59
to somebody very high up at one of the labs who said that was was saying that like if you take the total uh the total salaries of all the journalists in America it would be a fraction of what the AI labs are spending right now just acquiring ing data and so it it's not a financial question. It is a political
2:02:15
It is a political cultural question. It's interesting.
2:02:20
That's also the argument why I think that and and I've been really positively um surprised as we've talked with because we've talked with all the big tech companies, all the AI companies about this and they and they all like with with maybe one or two very small exceptions um they all say yes we understand over the long term we have to pay for content. And so they get that.
2:02:41
What what what what's key though is it has to be a fair playing field.
2:02:43
Um and that fair playing field everyone sees that slightly differently.
2:02:47
So if you're a new incumbent you want to say well Google can't be advantaged.
2:02:50
If you're Google you have to say you know well we have it has we have to be respected for the ecosystem that we've helped build and flourish and we should be able to you know this.
2:02:59
So everyone's got a slightly different take on this.
2:03:01
But I I'm actually encouraged that that the the dollars are there.
2:03:04
Uh that there is an acceptance that this is is uh is something that needs to happen and that if you if you if you ask them, everyone says yes, we need to do this as long as you can create a level playing field.
2:03:19
And again, that's where I think our role comes in is how can we create that level playing field?
2:03:22
How can we create scarcity?
2:03:24
And then how can we make sure that content creators are being compensated for the the hard work that they're doing? Yeah.
2:03:30
Zooming out just for a second, what is the state uh how's the internet been for the last few weeks?
2:03:36
It's been a busy uh month in or last month in geopolitics, I'm and I know that makes your job a bit harder or at least I would imagine uh things get a little bit busier. Keeps it interesting.
2:03:48
I mean, it's whatever whatever you see sort of in in a in the sort of kinetic side, the physical side of a conflict is is usually replicated in the in the cyber um side of the conflict.
2:03:59
And so um you know we as the hostilities um ramped up between Israel and Iran.
2:04:05
First of all we saw very clear signals that that something was coming ahead of time which is it's interesting how there's sort of a digital residue which shows when uh there's going to be uh kinetic action uh uh taken and so and so we were we were watching uh that that pretty carefully.
2:04:24
uh big attacks um targeting both um Iran uh one of Iran's largest banks was hacked by what looked like a group of um is Israeli um activists um the largest cryptocurrency exchange in Iran was hacked and all its funds were drained u again looking like it was either by the Israelis or or Israeli um sympathetic uh attackers.
2:04:45
Iran has has tried to do the same thing against Israel.
2:04:47
Israel's um cyber uh defenses are much stronger.
2:04:50
Um but we're starting to see that Iran is targeting other um uh other largely US uh financial services firms uh critical infrastructure firms uh to try and see to sort of try and strike back from what what is um you know h how do they strike back against the US in a way that um that that is harmful but doesn't start World War II?
2:05:15
And I think that increasingly um unfortunately cyber will be a big piece of that.
2:05:18
And so we're um you know we're we're doing everything we can to to stay in front of that, make sure that anyone who needs it has has protection and that we can keep keep the lights on everywhere in the world. Last question for me.
2:05:30
Um there's a number of larger overarching tech narratives.
2:05:33
I'm interested in if you're losing sleep over any or tracking any in in particular focus.
2:05:40
Um the AI talent wars and kind of pay inflation for top performers.
2:05:46
Is that an issue or something you've been tracking or uh the difficulty to build larger data centers and infrastructure the GPU short of the chip wars?
2:05:56
Any any of those topics kind of particularly relevant to you or your business or or kind of keeping you up at night?
2:06:03
Yeah, you know, I think we're we're not trying to build a foundational model LLM.
2:06:08
So So we're not we're not as much competing um for the the bleeding edge AI researchers.
2:06:13
I will say that we've I've been um really amazed at at both from an academic side and then also just you know pure talent side how many people are interested in working on that problem of something like how do you score content in order to be able to create a marketplace and so and so that that's something that's been um you know pretty exciting.
2:06:32
We're we're very excited about the uh the the people who are coming to uh apply to work for Cloudflare to to work on problems like that and like today has been a record uh applicant day for us uh on the on the Congratulations. Record.
2:06:44
Sorry, I hit I hit the air air horn for you on the record applicant day, but it's good. It's good news.
2:06:50
Uh and then, you know, I think um you know, we we've uh we tend to put a little bit of equipment in a lot of places and so that we don't tend not to have kind of massive multi- Yeah.
2:07:02
You're not running into like power constraints or rare earth element constraints or water, right, constraints?
2:07:06
Yeah, we we we do in but in in in kind of a micro way where you know we put GPUs at the edge of our network.
2:07:13
we have to live with a power envelope that we're given by an ISP.
2:07:15
And so we're doing things to try and figure out how to just get the best power efficiency out of the the GPUs that are there.
2:07:21
The answer to, you know, more and more need of GPUs can't be everyone has to turn up their own nuclear power station.
2:07:27
Um so so I do think the place where we're pushing suppliers and AMD is doing some really interesting things in the space.
2:07:35
Qualcomm is doing some interesting things in the space.
2:07:36
things in the space. fruit company down in Certino is doing some interesting things in the space or how do you focus on yes performance but performance per watt and and we've seen this game before with with Intel saying oh that doesn't matter like just just get the most performance possible you can water cool
2:07:52
things all kinds of things that that tends to not be the winning strategy over the long term and so we do think that there will be a renewed focus on just energy efficiency around GPUs and and we're really pushing uh that in a way that because because again we can't build our own data centers in the business that we're in. Um, and then and
2:08:08
Um, and then and then chip shortage.
2:08:10
The chip shortages are funny.
2:08:12
Like it's there's never been a chip shortage in any time in the history of of silicon that doesn't turn into a glut.
2:08:20
Um because it's Hey, but this time it's different.
2:08:23
Yeah, this time this time is totally different.
2:08:24
So I think that I think you're going to see both um you're going to just see shifting demand uh that's there and and uh and a lot more.
2:08:34
It's Nvidia is not going to be the only chip supplier of GPUs.
2:08:37
Uh there are some really again great great other providers that are out there. Yeah.
2:08:40
Well, thank you so much for stopping by.
2:08:42
Enjoyed the conversation.
2:08:44
Congrats on the launch and come back on again soon. Yeah. Yeah.
2:08:45
Thanks for congratulations.
2:08:47
Good luck filtering through all those job applicants. We'll talk to you soon. See you. Bye.
2:08:51
Really quickly, I got some breaking news.
2:08:53
The best performing AI agent now comes with a $1 million guarantee.
2:09:00
AI, the number one AI agent for customer service, number one in the world, number one in performance benchmarks, number one in competitive bakeoffs, number one ranking on G2, and you'll start a free trial.
2:09:12
You don't want to be baking off with Finn. ai from Intercom. From Intercom. Head over to Finn. ai to get started. I'll be right back. We'll kick it off. Yeah.
2:09:20
We'll bring in Luca from Pudgy Penguins.
2:09:22
Um, we're going to talk about what's going on with the market structure bill, what's going on with the Genius Act, what's going on with his business.
2:09:29
So, we're very excited to catch up with him.
2:09:31
It's his second time on the show.
2:09:32
He came on during crypto day. Fantastic conversation.
2:09:34
And now we're excited to have him back on to talk about the state of crypto. How you doing, Luca? Good to see you. I'm doing good. How are you doing? I'm great. What is that behind you? The the Igloo.
2:09:45
Is this uh part of the Pudgyp Penguins expanded universe?
2:09:50
Yeah, the holding company that holds Pudgy Penguins, Abstract, and a couple other companies we've yet to announce that we've acquired is called Igloo.
2:09:58
So, everything lives within the igloo. That makes sense. There you go.
2:10:02
Has there is there a uh is anyone riffing on the meme of like selling ice to an Eskimo yet?
2:10:08
I feel like that's uh you know, I know that you're good at at doing deals with old line retailers, but can you sell ice to an Eskimo?
2:10:17
I can I can sell ice to an Eskimo.
2:10:20
I can sell fish to a penguin.
2:10:20
I can sell land to a mammal. Yeah.
2:10:24
What is what what is the latest?
2:10:28
I mean, we caught up on crypto day a couple months ago.
2:10:31
What's the latest in your world? How are you feeling?
2:10:33
I we've been in this kind of kangaroo market for the highle L1s.
2:10:36
Uh prices are up, prices are down.
2:10:39
Everyone's kind of waiting for the next big crypto catalyst, but uh maybe that's the Genius Act, the Market Structure Act.
2:10:46
Maybe that's legislation, but what else are you tracking?
2:10:48
Stable coins have been really hot with the circle IPO.
2:10:52
What's been top of mind for you?
2:10:54
Yeah, our big thing at the Igloo is actually like we feel like we are one of the most well positioned organizations for consumer and culture in crypto.
2:11:01
And it was actually a little fascinating cuz I actually felt like I missed a piece in terms of what I felt like was needed for this consumer crypto revolution. Mhm.
2:11:15
And I think it's really prevalent around this genius act, which is I think you're going to see crypto break through mass off the backs of stable coins because if you look at all of the consumer crypto applications and really their barrier
2:11:29
and bottleneck, you know, I I I thought it was product builds and I do think that is true to a certain extent, but I think the product builds are a byproduct of the current environment that rewards builders for building in. And I think
2:11:43
And I think that gets completely flipped on its face when average consumers are now transacting in dollars that are on chain.
2:11:52
It opens up a whole new vertical of product builds that I think are more oriented around maybe providing a a service or an experience and not providing a scheme to return your dollars two or threefold as like a financial mechanism for you to kind of speculate on. Yeah.
2:12:09
Uh, and so I think like the Genius Act is quite literally going to kick off the consumer crypto revolution.
2:12:17
revolution. And that is really apps that are built around I think the best properties of blockchains which is like on its face I think a lot of people listening specifically around your audience like maybe look at blockchains as like maybe speculative vehicles or like programming languages but I think
2:12:32
the best frame to look at it as is like a blockchain is a censorship resistant version of a Stripe or a PayPal that taps into global composability and global liquidity which basically means that like anybody with an internet connection can transact, you know, using these payment rails without some sort of intermediary. And that at scale, if you
2:12:52
And that at scale, if you actually understand like consumer product builds, you know, every every guy that I know that's building an AI or building consumer product goods at one point has had their account locked up or their funds withheld or are paying a huge premium on these transactions, whether it's 3% in some cases 5 to 8% if it's a higher risk product.
2:13:13
Uh, and I think this, you know, blockchain is very useful when you think about removing that middleman for those applications.
2:13:20
I think we've just been pigeonhold as an industry around like these builds that are solely around speculation and making the end user more money, right?
2:13:29
Like I go buy this, I should return a 2x or a 3x or a 5x.
2:13:33
And I think we're slowly transitioning into more traditional builds where it's like, hey, consumer crypto is really around this idea of removing these these layers that I think today make it so hard to build, you know, fun and robust applications that are, you know, segmented either g geo uh from a geography perspective, whether it's certain regions, etc.
2:13:57
Uh and and I think it's kind of transitioning to like more real product builds that I think more serious entrepreneurs are interested in.
2:14:06
So the Genius Act, I think, is going to change everything because the the floodgates for stable coins are here.
2:14:10
Uh and I think that was the missing piece to the consumer crypto revolution that I think crypto very much desperately needed. Okay.
2:14:17
I want your reaction to this older analysis by Ben Thompson during the first NFT boom uh where he was talking about the value of NFTTS as a as a as a disintermediated membership uh program essentially.
2:14:35
So basically the the the the link I see to uh stable coins is if you have a community uh and you're a creator, you might not want to have someone have a super fluctuating financial interest in your community.
2:14:52
And so when I think about the folks who are big on Patreon and it's $5 a month or $10 a month, but when I pay $30 a month for the Wall Street Journal, I get access to wsj. com. I can comment there.
2:15:05
There's a bunch of other things that you could imagine, but I can't really go around the internet and carry my Wall Street Journal login to their YouTube account and get custom YouTube videos.
2:15:15
Like the videos have to be embedded on the Wall Street Journal website, so it can't really follow me around on the web.
2:15:20
And Ben Thompson was talking to Adam Miseri at Instagram about this idea that Instagram might be open to a bring your own membership type of integration with web 3.
2:15:30
And so I I I think that we were almost there, but there was no one that really broke out into mainstream adoption where both you you got to get YouTube and you got to get Instagram to say yes, we are we are you know integrating this membership card as a native as a primitive on our platform to allow you know creators to create uh you know subscription level content.
2:15:57
Instead we wound up with every social media app has their own subscriptions.
2:16:00
So, I can go get YouTube YouTube premium subscribers who subscribe to me, pay me $5.
2:16:04
Then on X, I can have subscriptions $5.
2:16:07
They subscribe to me there.
2:16:08
Uh, but you got to follow follow me around and you can't just subscribe to me as the creator.
2:16:13
That felt like something that Web 3 could do, but we weren't quite there with stable coins.
2:16:17
Is that something you think is a reasonable uh dynamic that could play out postgenius act?
2:16:22
Is there still some sort of counterveailing force where the big tech companies might just want not want to give it away because hey, they're earning a cut on those membership revenues directly?
2:16:31
How do you see that playing out?
2:16:34
They're they're definitely not going to want to give it away, right?
2:16:36
Like those guys are pretty good at capturing and trying to keep revenue.
2:16:40
I think like one of the big, you know, bull cases around web 3 is like these universal loginins and identities, right?
2:16:46
That like follow you all over the internet.
2:16:48
Like that that's going to take a lot of pieces to put together.
2:16:52
But like things like the Genius Act, right, and everyone transacting in stable coins makes that so much more possible and feasible that I don't know of anyone who's really mastering that build just yet.
2:17:02
And to the point of NFTs, I think there's two bulk cases for NFTs.
2:17:07
One, which is the one that you referenced, which is access, right?
2:17:10
And that is a more mundane, probably less exciting, but probably bigger TAM because you could replace the entire ticketing industry and only fans and Patreon like you'd mentioned and info, right?
2:17:20
the guys can go and plug that in.
2:17:22
So I, you know, I I I I think there's that and then there's also the TAM of digital collectibles, which is, you know, physical collectibles today have a total addressable market of about 600 billion.
2:17:35
1% of that 600 billion is digital.
2:17:35
Yet digital collectibles have a slew of pros that physical do not, right?
2:17:39
They are they have no friction when buying or selling.
2:17:43
They have no insurance issues.
2:17:45
They have no storage issues.
2:17:45
They have no inauthenticity issues, right?
2:17:47
And and so I think there's a there's a gap in value to capture in terms of you know in a $600 billion market.
2:17:53
It's kind of like gold to you know physical gold to digital gold.
2:17:58
You know Bitcoin's captured 10% of it.
2:17:59
I believe that digital collectibles are you know whether you argue what side of the spectrum they're on they're not 1% they shouldn't have 1% of that market share.
2:18:10
And so I think there's going to be value to capture between that spread.
2:18:11
The other the other the other thing that stands out to me as a as just an observer is that I think when people are frustrated with the NFT market broadly, it's frustration with 90 it's basically saying like yes 99% of attempts to create new IP are going to fall flat because it's really hard to create IP that is actually like it's easy to create IP, it's hard to create valuable IP, right?
2:18:39
IP, right? And so there was a period in 2021 around Hollywood like 90% of scripts in Hollywood don't go anywhere but 1% it's like they they just hit the chain and they're they're they're you know people assume that there should be some value tied to them and so I think it's just very possible that that NFTts
2:18:59
become bigger and bigger and bigger but it's just a handful of really select projects that that you know become these sort of centers for gravity and get developed in all these different ways but you're not going to have a hundred 000 projects that each have movie offshoots, video game offshoots, you know, physical toys attached to them. And that's exactly the way it happens in
2:19:19
And that's exactly the way it happens in the real world where people people in Hollywood, like you said, will spend their entire life trying to develop a cartoon that actually turns into a franchise.
2:19:29
And many people will try and fail. Yeah.
2:19:31
And the difference with crypto is like the expectation is immediate value creation.
2:19:35
And um I just don't I don't see that happening. Yeah.
2:19:42
The the tokens, whether it's fungeible or nonfgeable, there's a superpower behind it.
2:19:47
It's a kryptonite and it's a superpower depending on how you harness the power.
2:19:50
But financial alignments can be your greatest weapon or your greatest weakness. Yeah. Right.
2:19:55
And how you harness that, right?
2:19:57
Like us being able at Pudgy to galvanize this group of 10,000 people to go and hit algorithms.
2:20:05
Our big breakthrough at Pudgy was I told all of them that Walmart had turned us down, that we were going to go live on Amazon and that I needed everybody to buy and if they bought it right all at the same time, we would trigger Amazon algorithms.
2:20:19
We'd be number one across all of these different categories and that they'd give me a lot of leverage to go do a deal with a major retailer.
2:20:27
Well, what happened was exactly that.
2:20:27
We launched within, you know, 30 minutes of launching on Amazon.
2:20:31
I'm AC I'm number one across basically every consumer product category, every toy category.
2:20:37
And then and then Laura Rush, who was the head of merchandising at the time at Walmart, pinged us and said, "Uh, we want to stock you. You got 3 months.
2:20:44
We're going to put you in a pallet in the center of the store in 2,000 locations in quarter 4."
2:20:47
And that was like the big paradigm shift for Pudgy where people were like, "Oh you know, these kids, you know, came out of nowhere and did it."
2:20:54
And and the rest is history.
2:20:55
We've basically, you know, carried that momentum up until where we are today.
2:20:58
Uh but that was our big breakthrough and that was because I think of that alignment and that incentive.
2:21:04
Everyone was incentivized to go and create this headline and this moment for Pudgy, at least the core community members who had quote unquote bags or collectibles or tokens within the ecosystem.
2:21:14
Um and if you can harness that and if you can respect that, uh it it can be a superpower.
2:21:19
It can be your greatest kryptonite as you've seen a hundred times over and there's, you know, plenty of horror stories which we won't get into.
2:21:26
Yeah, bit of a wildcard question, but I don't know if you saw uh OpenAI has a deal with Mattel to make a chat GPT enabled Barbie potentially.
2:21:33
We don't really know the shape of the deal.
2:21:38
How many LLMs can fit in a Pudgy Penguin? Yeah.
2:21:39
Yeah, that's what we want to know.
2:21:41
Uh I mean a lot of LLMs in this I mean you're you're you're in some ways in the toy business and what is the next generation of toys look like?
2:21:47
What are you optimistic about? What will change?
2:21:52
Um what technologies do you think will be integrated?
2:21:55
be integrated? uh what considerations will drive actual success instead of just like yeah we slapped AI on it like we tried that with uh with every toaster having Bluetooth didn't really work out but the Nest thermostat pretty f pretty pretty popular so yeah what how does this play out it's a great question I saw this coming two years ago and I
2:22:15
basically jumped for joy when I thought of it and I told the guys because for those who are not familiar the toys that we sell have the QR codes so it's not just like selling toys that are you know IP related we're actually putting NFTs in the hands of regular people because you buy this toy, you scan this QR code, you redeem these NFTs through a gas experience. So, it's like a whole crypto
2:22:31
So, it's like a whole crypto funnel, which is basically our product, our physical toys.
2:22:35
But, I jumped for joy cuz I was like, dude, what if these these characters that you bought and these NFTts that you created then became like your AI homeworker helper, right?
2:22:45
And this was an idea I had two two years ago.
2:22:47
And Lorenzo, our president, was like, you know, dude, you know, if there's a 01% chance this homeworker helper tells a kid to kill himself, it's not worth it because we'll get dessheld and it'll create a bunch of chaos.
2:22:57
And we weren't, you know, AI guys.
2:22:58
But we actually are doing a toy with Curio who recently raised pretty big over in uh the whole, you know, uh AI crypto or AI, you know, startup Twitter land, X. com land. And and place. Yeah.
2:23:13
With that, we're familiar.
2:23:15
We're from We've dabbled. Um that's great. Yeah.
2:23:19
How who's the um who's the musician that's involved with that? I don't know.
2:23:26
I mean, I Yeah, I'm I'm super bullish on this.
2:23:29
Uh you know, I have kids and uh and I mean, I was thinking about this with the Rabbit R1.
2:23:34
Like yes, the safety and security stuff is really really important, but safety interpretability research and the ability to run an extra LLM just filtering is this a harmful uh response from the LLM that seems to be getting better and better and and I'm pretty optimistic about having a a watered down LLM that can't explain everything but is really good at identifying plants and animals.
2:23:58
That's going to be a big product for kids, I think.
2:23:59
So, the trick to CPG is all about providing value for the customer.
2:24:05
Like Walmart will tell you if you're trying to stock like they're always optimizing for the best deal.
2:24:08
And so my mind frame is like how much bang for your buck can you get for this purchase.
2:24:14
Well, if you can buy it and redeem some NFTts that could be worth some digital dollars, maybe you recoup your cost on the physical item and then it's a net free good and then maybe you can go in and ask it questions and now it's a value ad.
2:24:25
And so I think this stuff is just changing.
2:24:27
It's a complete paradigm shift.
2:24:29
I think I think no one's really ready for how I think the world is going to change.
2:24:32
But in terms of a value ad, you can plug in an LLM to a lot of physical products, just alpha for the room, you know, and and create a ton of value for your end consumer, which is ultimately your goal as an entrepreneur. That's awesome.
2:24:44
Well, thank you so much for stopping by.
2:24:45
Always appreciate your perspective.
2:24:47
We got to have you back when there's more breaking news.
2:24:50
Uh, and we'll keep we'll we'll keep catching up with you. Good to see you, Luka. We'll talk. Cheers, Luka. Have a good one. Bye.
2:24:56
Really quickly, let me tell you about adqu. com.
2:24:58
Out of home advertising made easy and measurable.
2:25:00
Say goodbye to the headaches of out ofome advertising.
2:25:03
Only adqu combines technology out of home expertise and data to enable efficient, seamless ad buying across the globe.
2:25:08
If you're going to be in Walmart, if you're going to be selling toys, you got to get on adqu. com.
2:25:12
Anyway, our next guest is here, Julia. Hey, how you doing? Doing well. I like the intro music. Yes.
2:25:19
Welcome to the We are professional newscasters, as you know.
2:25:25
You're one of a handful of people in the entire world that has actually visited the Temple of Technology.
2:25:28
You've been into the Temple of Technology, the Capital of Capital. Yes. Yes, that's right.
2:25:34
Uh give us the latest and greatest in your world.
2:25:37
Uh new uh new piece in uh Palladium. What was the thesis? What inspired it?
2:25:43
Take us through the high level of the piece.
2:25:48
Yeah, so something I've been thinking about for a while.
2:25:49
It was actually sort of inspired by this freshman at Stanford who I was talking to in the library, very very sweet girl, and she was like, "What are you doing postgrad?"
2:25:58
And I said, "Well, I'm working for this magazine called Arena.
2:26:01
They want to do books, so I'm going to be helping with that. I'm super excited."
2:26:04
And she said, "That's a 1930 ass job."
2:26:06
And I was like, "Well, why is a 1930 ass job?"
2:26:09
I was like, "You know, most people are like consultants or software engineers, but you're making like this very real thing that's like you can touch."
2:26:18
I was like, "Yeah, like maybe it is a 1930s job."
2:26:20
I guess people a lot of people in the 1930s didn't have jobs, but there were some people who were writing books and publishing books.
2:26:28
So, I was sort of thinking it's like, well, how do we get to this point where it's like you it is sort of like this weird thing for people to be doing something real and tangible.
2:26:38
And then I think it's also just, you know, I sort of feel like graduating from Stanford in 2025 in some ways I'm like out on the last chopper out of NOM like CGBT was only like super ubiquitous really my senior year.
2:26:53
really my senior year. Like sometimes you just walk into the library and everyone had chat GBT on their screens and it was just you'd go to section and the section leader would be like please like I know it's really easy to use AI
2:27:04
but it would really make me sad if you used AI to write your weekly reading response and sort of just like all these things that were just like you felt like you're more practicing like acting like a college student because just AI was just making a lot of the dayto-day like obsolete. And it's also just like
2:27:22
And it's also just like there's a pretty good chance that like AI is a better writer than I am, there's a really high chance AI will be a better writer than I am in 5 years.
2:27:30
And I think it's like well if what I wanted to be is a writer and that's going to be obsolete or AI is going to be better than that me or like if my friends who want to be lawyers are just going to be replaced by Harvey and my friends who want to be software engineers are going to be replaced by cursor like what does the future look like? Mhm.
2:27:51
And I think I was also guided by this conversation I had with a wonderful professor at Stanford, uh, Jennifer Burns.
2:27:56
She's in the history department.
2:27:57
She has written biographies of Milton Friedman and Anne Rand.
2:28:00
I took a really great class with her in the fall called Thinking about Capitalism.
2:28:03
And I had a lot of conversations with her just because I found her work really interesting.
2:28:09
And she asked me, she's like, you know, you're sort of like pretty like imshed in like the current, you know, right-wing intellectual moment.
2:28:17
who is like the defining in like economic thinker.
2:28:20
And this was back in like October, November.
2:28:22
And I was just totally stumped.
2:28:24
And when I when I sort of published the piece with Palladium, I sent it to her and I was like, this is sort of my answer to like what what is the economic thought of Gen Z or really of the 21st century.
2:28:35
And it it's economic nihilism. Nihilism. So yeah, are we cooked?
2:28:43
Where are you most bearish?
2:28:43
It seems like you're you're worried about a lot of different trends.
2:28:47
Um is is is capitalism cooked? Is democracy cooked?
2:28:53
Is um technology I'd be curious to get an anecdotal job market kind of update?
2:29:00
It sounds like we got breakfast last week and you were saying that sure somebody uh some of your classmates got offers from what was it like Teeu or something somebody to work in like a warehouse and they were all like super Yeah.
2:29:14
And it was like become this like Shien warehouse manager.
2:29:16
And I was like actually that sounds like a really interesting job.
2:29:19
Like if I didn't have something lined up maybe I would apply to be like a Shien warehouse manager.
2:29:25
And there's this app called Fizz which is sort of like Yakyak 2. 0.
2:29:27
is sort of like Yakyak 2.0. it's an anonymous forum for college students at particular campuses and people got so mad that she like dared ask Stanford students to stoop to their level to become a warehouse manager and I think there's it's it's sort of like a paradox because Stanford students believe that
2:29:47
they are they are entitled to certain jobs and this is true of sort of like all IVS or elite universities but there are a lot of jobs that people aren't willing to do And I sort of wrote in the article for Platium, it's like the line used to be, well, the hardest part about Harvard is getting in. And now I would
2:30:03
And now I would say the hardest part about Harvard is getting a job that your peers deem elite enough to not feel any like social backlash or to feel comfortable in the rat race.
2:30:14
And I think it's Peter Turchin who's this phenomenal historian has written a lot about how conflict rises in societies when there are more elite aspirants than there are elite positions.
2:30:26
And I think especially as class sizes at these schools are expanding and the internet is democratizing access to education and democratizing access to power in some ways.
2:30:38
I think that these schools are in a really tough place because there it is not is no longer a guarantee that you can go to Stanford, you can go to Harvard, you can go to Yale and be upper middle class.
2:30:53
Is uh are are there really not that many jobs at McKenzie, Bane, BCG, the Wall Street banks, the big tech companies, the PMs?
2:31:01
It feels like we're we're hearing a lot of nervousness about job loss and not hiring, but it feels like also Mckenzie's printing a bunch of money just writing, you know, slide decks about AI transformation in the Fortune 500 or has the actual shift in what is a high status job changed and and and if it has, what what is the high status job?
2:31:24
Is it entrepreneurship now? Is it YC?
2:31:25
Where are we sitting with that?
2:31:28
I think it's I think it's VC.
2:31:28
I think it's entrepreneurship just regards to whether they want to go straight to venture capital.
2:31:33
Yeah, this is this is the golden ticket. Uh just kidding.
2:31:35
We support venture capital. Yeah.
2:31:37
I think like anecdotally it seems that big tech is like that used to sort of be like okay everyone told you to study CS and if you study CS you'll have like a 200k postgrad salary and now that's not really happening anymore.
2:31:52
Like yeah, it's I think there's definitely a shortage of or especially in like more big tech firms which offer more security than like startups.
2:32:03
It's a lot harder to get those positions now than 5 or 10 years ago.
2:32:07
Do you think that's more of a hangover from COVID and kind of Elon showing that he could run X at 1/5if a headcount versus truly AI disruptive to like replacing workers?
2:32:18
I think it's probably both.
2:32:21
I think there was like the uh zero um well there was a time in 2021 where just people were throwing money at anything that they could throw money at.
2:32:31
And I think we're in very different times now.
2:32:33
But I think I I think about it less in terms of just like oh what's the market like now and more in terms of like how has the social contract changed?
2:32:40
And I think for a lot of elite college students, the sort of social contract was for a time like okay, if you study like these three things, then postgrad you will have like and you don't up in college, you'll have like pretty easy access to these really important highpaying jobs.
2:32:58
But I think we've seen now that it's like maybe a job at Goldman Sachs as an analyst isn't that important.
2:33:05
You're just like changing font colors for 110 hours a week.
2:33:09
but also just we're seeing undermployment rates in these sectors.
2:33:12
I think the computer science underemployment rate is 1 in6 which is insane considering that pretty much every authority figure I had from ages like 8 to 18 told me to major in computer science. Yeah.
2:33:31
Do you Sam Alman wrote this post about the soft singularity and his kind of thesis was that if you went back to 1930 they would think we all have fake jobs uh and so we will maybe create new fake jobs.
2:33:44
Are you optimistic about that at all?
2:33:47
Yeah, I don't think the fake jobs thing is necessarily going away.
2:33:49
Like I'm in LA right now.
2:33:51
That's where my parents live and every third person is an influencer.
2:33:54
Um, so it's not hard for me to say let's hear for the influencers, you know, the hardworking content creator content creators.
2:34:01
It's very hard to go slaves to the algo. Yeah.
2:34:06
But I think what we're going to see is just there's a lack of what we're sort of seeing it's like there are there have been fake jobs for I don't know maybe probably not eternity but for like recent history and what we're going to see is just the fake jobs it just the work becomes more and more meaningless.
2:34:31
Like I don't know what the fake jobs of like 5 years, 10 years from now are going to look like, but I think what like a key thesis in my article was that like Gen Z is simultaneously like hyper careerist, but like there's no career at all.
2:34:44
It's just sort of floating around from one job to the other.
2:34:46
And because there's so much like hopping over from one company to the another, one firm to the next, there's never really any like real expertise that's developed and there's this shallowess in the work.
2:34:58
I think it's like my parents um especially my dad who just retired, he worked at one firm for like the vast majority of his career.
2:35:07
He was like an expert in that in that field and that wasn't anomalous for people of his generations of boomers at all.
2:35:15
But I think now with Gen Z, it's just like, okay, you have one job for six months to two years and then you move on to the next thing.
2:35:21
And these big companies have like incentive structures built in like McKenzie that has search time.
2:35:27
It's like, okay, well, we're going to after one year, we'll give you six weeks off so you can go and look for another job.
2:35:32
And I think in sort of doing that, it's admitting that it's a fake job because it has no re real tie to its employees and no real tie to their employees sticking around. Yeah.
2:35:44
for some for some of the new grad roles, at least the big three, they they'll very much encourage you to go take two years off to do an MBA, which feels like I don't know, it just feels odd to me to like lose an employee that you could be just nurturing internally and growing and stuff.
2:35:58
Uh what what's the reaction on campus been to the uh the big AI trade deals, these massive offers?
2:36:06
Uh are people most of those folks who are getting $und00 million offers, they have PhDs, they're published papers.
2:36:11
Uh maybe the new advice should be don't just get a CS job and then a CS degree and then try and go um work in big tech.
2:36:19
Maybe it should say actually stick in academia longer. Yeah. I I don't know.
2:36:25
I can't say just because like I'm not a CS major.
2:36:27
In fact, I never took a single CS class at Stanford. Yeah.
2:36:31
But I also think the the people who go into CS PhD programs just like love the subject.
2:36:39
And I think like the Stanford particular is very good at getting people to major in computer science because they just teach the intro courses very very well.
2:36:47
But I think that's a very different demographic.
2:36:51
The people you just sort of want to check out for college have like an easyish major that a lot of people around them do so they can have a cohort versus the people who are like I'm really into systems.
2:37:00
Like this thing is like if you met someone at Stanford who was like a systems concentrator rather than an AI concentrator that they were like really in it. Um, I don't know.
2:37:09
Oh, you're a systems guy. Name every system.
2:37:15
This is symbolic systems.
2:37:15
No, symbolic systems is a fake major.
2:37:18
It's like the systems track industry. I also did a fake major.
2:37:22
So, I'm not That's so funny. It's fake.
2:37:24
The realest major is art history.
2:37:26
Higher employment than CS.
2:37:26
You know, Symbolic Systems is like like literally every first employee at all the big tech companies who's worth like hundred million dollars now. They're all Yeah.
2:37:35
It's like they probably have a higher average income than the average comparative literature major, which is what makes it, but Oh, well.
2:37:39
Uh, well, thank you for so much for stopping by.
2:37:44
Jord, do you have any other questions?
2:37:46
Let's make this You're new our You're our Gen Z. Oh gosh. Okay. We need you. We need you.
2:37:52
I I want more updates, reporting on the Stanford class of 2025. Where are they landing?
2:37:58
We got a fake job for you.
2:37:58
Gen Z correspondent on a live stream. Yes. New fake job. New fake job alert. New fake job.
2:38:06
I had a funny idea when I was editor but never got around to doing it.
2:38:10
And so one last thing was that we would compile the names and photos of everyone who got a job at Goldman Sachs and it was just black and white portraits of all of them listed like vertically and it's was titled our best and brightest.
2:38:22
So something like that for you guys. There we go. Let's see it. Love it.
2:38:28
Well, very very excited for you to be at Arena.
2:38:30
Who who's gonna take the companies public if not Goldman Sachs? I don't know.
2:38:34
Who's going to issue the the massive uh debt to back these corporate titans if not Goldman Sachs?
2:38:39
Well, they are the best and brightest.
2:38:42
They're doing important things. They are.
2:38:46
Anyway, this has been fantastic.
2:38:46
Thank you so much for stopping by. Great to see you. Thank you. Cheers. Talk soon.
2:38:50
Uh up next, we have Joel from Thompson Reuters coming in the studio.
2:38:52
Uh he was supposed to be on AI day. He's buddies with Swix.
2:38:56
was introduced to me by Sean Swix who's been on the show earlier.
2:38:58
Uh we're going to dig into how Thompson Reuters is using artificial intelligence and what's going on in the enterprise.
2:39:07
Welcome to the stream Joel. How are you doing today? Guys, how's it going? It's going great. Good to meet you.
2:39:13
Uh would you mind kicking us off with an introduction on yourself and then I'd love to know a little bit about the shape of Thompson Reuters business.
2:39:19
People are obviously familiar with the brand, but uh on a day-to-day basis, what's actually uh going on the technology side? Yeah, you bet.
2:39:28
Um, so I've been at Thompson Reuters a few years.
2:39:30
I joined via acquisition uh as a CTO of an AI company, TR Acquired a few years back and have been with the company since then.
2:39:39
I think a lot of people know Thompson Reuters from the news business.
2:39:44
The obviously that's the most public facing part of the business, but from a from a revenue perspective and a technology perspective, it's really I would say a relatively small percentage of of what we do.
2:39:55
You know, we're one of the top two or three software providers in the world in the legal industry, in the tax industry, compliance, audit, and risk.
2:40:06
So we really cover I would say a pretty broad expanse of the information services uh industry.
2:40:12
Um for a lot of years we were really seen as a as a content provider or an information provider within these businesses.
2:40:21
But uh software and in particular like search has been sort of a foundational piece of the technology that has kind of underpinned delivering that information to customers.
2:40:34
again across across the sectors legal, tax, compliance, etc. Yeah. Okay.
2:40:39
So, walk me through uh of those verticals, which one is seeing the most acceleration from the current suite of AI tools?
2:40:46
Which ones uh kind of maybe next up at bat and not really uh the models aren't really having as much impact as you expected so far, but where are we getting the most leverage in AI?
2:41:00
Yeah, it's a great question.
2:41:00
You know, if you look at all of those industries, you wouldn't necessarily look at anyone and and immediately say they are uh particularly eager to adopt technology or aggressive in in that sense.
2:41:13
But, you know, for the last two two and a half years, the legal industry has been on fire and I'm sure you guys have talked about it.
2:41:21
There's a a ton of capital coming into that to the legal tech industry in general. Yeah.
2:41:25
Um but really from the customers too have been extremely hungry and eager to use this technology and it's having profound impact on uh the way those professionals think about uh doing their work and operating their businesses and and the underlying businesses business models themselves.
2:41:44
I think uh you're starting to see more um more progress being made in industries like tax though.
2:41:52
I think sort of the quantitative capabilities of the models kind of lagged the the language capabilities of the models. Mhm.
2:42:00
Um but I think that's improved.
2:42:00
But I also think if you look at like the tax or the audit industry that particularly workflow kind of heavy and I I think as agents have really taken off in the last you know six to nine months and the tool call and capabilities of the models have improved uh it's really opened the door to you know deliver like real true automation and productivity uh with some of these agentic frameworks and reasoning based models. Yeah.
2:42:30
Talk to me about I mean I imagine you're not focused on pre-training a massive model to get to the frontier.
2:42:36
You're probably a buyer of that capability, but we're seeing a lot of deals get done.
2:42:40
Open AI's fine-tuning models for clients above $10 million.
2:42:46
They have they have even cheaper offerings.
2:42:48
Palunteer is doing fine-tuning work now.
2:42:50
We just heard Thinking Machines Mir Marudi is doing some fine-tuning reinforcement learning for businessto business applications.
2:42:58
um what are you a buyer of right now?
2:43:01
What are you focused on uh building internal applications around?
2:43:07
Where do you want the various uh pieces of the building blocks of a great artificial intelligence enabled product to sit? Yeah.
2:43:16
Um this might be a little bit of a long-winded answer.
2:43:19
long-winded answer. I would say first and foremost when we look about look at shipping product into the market quickly we're certainly a buyer of these models and and we're close partners with Anthropic OpenAI Google and others in that regard you know when those
2:43:37
companies are working on their next versions of models a lot of times they come to us to to help them evaluate and test certain capabilities of the models because the the type of work that's done in in the legal industry in particular is is is challenging. It's not a simple
2:43:51
It's not a simple like rhetorical response.
2:43:53
It's it's uh a deeper level of reasoning that happens in terms of the the types of problems that we're solving.
2:44:02
So I would say first and foremost like we're we're we're buyers of the technology.
2:44:06
We have done you know fine-tuning engagements with several of these providers.
2:44:10
super excited about the sort of RFT uh trend that may be coming with with many of them in terms of being able to apply more RLbased methods to to that fine-tuning.
2:44:22
Um but we're also I I would say builders in some sense.
2:44:26
We're we're not you know pre-training models of the size of of uh GPT4 or or or claude uh sonnet but um but we do engage in that.
2:44:39
So about a year and a half ago, we acquired a small company with a couple folks from from Deep Mind who have come and and really built up our deep learning expertise and deep learning training expertise within our team.
2:44:50
You know, we've got about one and a half trillion tokens of our own proprietary content uh in in legal and tax and news and and like I said, that's been sort of one of our defining assets for a long period of time. Yeah.
2:45:05
uh as RL really becomes kind of the the main mechanism for driving further improvements in the model, uh we see our talent, you know, our 4500 domain experts across legal and and tax and otherwise uh really being pivotal in that process of driving performance into the model.
2:45:26
So, uh, we're certainly a buyer, but we're definitely trying to exercise, uh, the data and the content that we have proprietary to us, but also the the human experts that we employ to drive that.
2:45:38
Can you give me some extra color on on all the different value kind of on the table with one of these deals between a big company and a foundation model provider?
2:45:49
Because it can seem so simple.
2:45:51
It's like I'm I'm I'm buying an API, but there's actually a lot more to it.
2:45:56
And we were talking about this in the context of uh Apple potentially partnering with Anthropic or OpenAI.
2:46:02
And there's actually kind of a, you know, an odd relationship there where if Apple's sending a bunch of users to chat GPT, that could improve the model and that could make their business stronger.
2:46:15
They could wind up servicing ads in the responses like Google does.
2:46:19
And you could wind up with a situation where, you know, Google is paying Apple, not the other way around.
2:46:24
You could see ChachiPT eventually paying Apple for the right to be on the Siri button.
2:46:29
Um, but in your business, you have over a trillion tokens.
2:46:31
That's pretty valuable data.
2:46:33
There's a question of, you know, what what value are you bringing to that side of the deal?
2:46:39
And so, how are you thinking about doing a great deal that sticks and is aligned economically over the long term with a foundation model provider?
2:46:47
Yeah, it's a really it's a really hard question to answer and certainly you have to kind of like have some crystal ball to project out what the next two years are going to look like and it's it's hard to project two weeks to to do that effectively.
2:46:59
We got one here in the studio for this exact reason. Helps quite a bit. It helps a lot. Yeah, it's important.
2:47:08
I'll borrow that next time I'm out there.
2:47:10
Um, so you know for us I think like I said our people and our experts and the data that they create ultimately I think is is sort of core to our identity and core to what we believe is valuable to us and we try to really hold on to that and extract as much value from it into our IP as we can.
2:47:32
I think where where there's mutual benefit like I said before is um we provide a lot of insight and direction to these foundation model providers in terms of like what the models are failing at sort of what kind of capabilities.
2:47:47
you know, they have these public benchmarks that say, you know, every week one model's better than the other, but they don't really represent the real world.
2:47:55
And in particular, they don't represent the real world of like how professionals do work on a day-to-day basis.
2:48:01
And we spend a tremendous amount of time building real world evaluations into those kinds of things.
2:48:08
Uh, and not only that, but we spend a tremendous amount of time and energy grading and evaluating those on like real world rubrics for how a lawyer would interpret this.
2:48:18
So, it's not like it's a right or wrong answer. It's is it helpful?
2:48:23
Is it harmful in some way to a lawyer?
2:48:25
These are types of things that like create a require a tremendous amount of just domain understanding to to do well.
2:48:32
And so that insight is super valuable to the foundation model providers to understand how they need to adjust their data sets and it benefits us on the flip side because we get more performant models uh out the back end that that are that are uh better suited for our tasks.
2:48:48
H how much time are you guys spending just on preventing hallucinations when you talk about domains like law and tax?
2:48:59
You know, it's not uncommon for a a lawyer or a tax attorney to hallucinate themselves, put, you know, put the wrong number somewhere, you know, have have the right but but usually there's, you know, hopefully checks and balances in terms of someone else reviewing it, the client reviewing it, making sure that it's getting to the right place.
2:49:16
Um, but that feels like the number one of the bigger issues kind of holding back LLMs from, you know, unlocking the full value.
2:49:25
So I imagine it's it's a big focus for you guys, but but what are your thoughts?
2:49:28
Yeah, I mean all all the time.
2:49:31
I mean like I think that that's one of the core things that we do, you know, and you're right that hallucinations existed before AI was involved here and they exist now and they no matter how good these models will get, they will exist in the future.
2:49:47
I think there's there's some element of like um uh that that that will be the fact.
2:49:53
I mean foundationally the way the models operate like it it is a propensity and it can happen.
2:49:58
Uh so but we spend a lot of time trying to build guard rails trying to build uh methodologies to to prevent that or at least mitigate it in in as as many ways as we can.
2:50:08
Uh I would say this is where though the the UX of the applications really becomes like more and more critical like creating the environment where a lawyer like knows what they need to validate knows how to do that validation uh and can do it quickly and efficiently so that they're still saving time in whatever they're doing I think is a critical part of the the design of the applications and how they operate.
2:50:35
And as much time as we spend on the science side trying to build algorithms that prevent that, we spend just as much time on the design side trying to make sure we build experiences that allow lawyers to to build that that confidence and trust that that they need. Very cool.
2:50:52
Well, thank you so much for stopping by.
2:50:54
We'll have to have you back when there's more news from your world.
2:50:55
And I we hope you have a great day and have a have July 4th, too. Yeah.
2:51:00
Appreciate the We'll talk to you soon. Cheers.
2:51:02
Uh, Jordy, how did you sleep last night?
2:51:04
I think you might have me beat. I was a little rougher.
2:51:10
I got an 81, but how did you do? Go to eight asleep. com/tbpn. Get a pod five. Actually, unbelievable. Did I beat you? Did I beat you? Play the sound effect.
2:51:21
You know, you know the reward I get. 81. What do you get? I can't believe it. 52. 52. Oh my god. Wow.
2:51:30
I I'm said I said for what it's worth, you have put on a fantastic performance this show. I would have guessed 99.
2:51:36
I thought you were locked in.
2:51:39
I think you've done I think you've done fantastically this show.
2:51:40
I I I like need to hire a new head coach.
2:51:43
I'm going to like have a call. Whoa.
2:51:44
We We got to have Brian Johnson back on the show.
2:51:46
He He's down to come in the studio. He can give us a coach.
2:51:51
We'll bring Andrew Huberman back, too.
2:51:52
He was Andrew Huber was saying there's a drug that can help you sleep better. Something like that.
2:51:56
I might need to get on peeds to to actually beat you.
2:51:59
But then we're going to have to drug test because if you're if you're cheating and I'm and I'm natty sleeping, this is not going to fly. We sleep.
2:52:06
We focus on our sleep to make the show better.
2:52:08
We should be able to do any type of And there's no better There's no better place to sleep than a wander.
2:52:13
If you're on vacation, find your happy place.
2:52:17
Book a wander with inspiring views, hotel great amenities, dreamy beds, top tier cleaning, and 247 concier service.
2:52:23
It's a vacation home but better.
2:52:24
We are gonna be launching a video of us in a wander. Get ready for that. I can't wait. Get ready.
2:52:28
We have our next guest in the studio.
2:52:31
I'm gonna let you take the intro.
2:52:34
Coming in from Ambrook announcing some news. What's going on? Welcome to the studio.
2:52:39
Hey, thanks so much for having me.
2:52:41
Uh quick intro on yourself and the company and then let's get into the news. Okay, sweet. So, I'm McKenzie.
2:52:48
I'm the CEO, uh one of the co-founders of Ambrook.
2:52:51
Our mission is to help American family-run businesses become more profitable and resilient.
2:52:59
And uh yeah, the way that we do that is we're starting in agriculture.
2:53:03
So building out financial management software.
2:53:05
So accounting like the full bundle.
2:53:07
So accounting, payments, um banking, cards for America's farmers and ranchers.
2:53:13
How did you land on this idea?
2:53:16
Where did it where did it all start?
2:53:20
Yeah, it was a pretty winding path, but we actually started during the pandemic helping a lot of producers try to get access to working capital.
2:53:26
Um, and if you talk to a lot of producers about their problems and sort of what, you know, they think software can help with, it almost always has to do with the all the way up the Mazda hierarchy.
2:53:36
So, you know, getting access to capital to take more risk.
2:53:39
Um, but the more that we dug into it, we helped we helped um a couple thousand producers access a couple million uh during during the pandemic, but when you really dig into it, uh it's just kind of it kind of gets down to like accounting and financial recordkeeping.
2:53:54
Like it's just accounting all the way down non-existent was one of the issues.
2:53:58
you're like trying to underwite a business, but there's they're like, "Well, I have this piece of paper and I have this piece of paper and and you know, maybe they have a it's a huge range, like you can imagine."
2:54:10
Um, but a lot of um, you know, you have farmers that are running multi-million dollar businesses on pen and paper.
2:54:16
Um, and of course you have folks who are also, um, using, you know, standard F&B accounting software, too.
2:54:24
Um, but a real the real issue is just that farming is actually so complicated.
2:54:30
Like, you know, I don't know if you guys have ever been on a dairy or how to do or how to do FD.
2:54:34
I don't know if you guys have ever farmed.
2:54:38
Not a lot of content farming. Content farming. Uh, no.
2:54:40
I I I grew up in I grew up in the country in wine country, so I had some exposure.
2:54:46
I also have some portfolio companies that that have some end farms as customers.
2:54:51
So I feel like I've had some exposure but certainly not an expert.
2:54:56
But from from my understanding I think that the challenge is one you're you're actually on a farm and historically some farm you know maybe I'm sure there's a lot of like Starlink adoption which maybe is a is a catalyst farmers about but at the same time you're not at a desktop computer.
2:55:12
So mobile mobile the explosion of mobile helped in farming because you could be updating a database or even sending an email.
2:55:19
Um, but all that is like relatively new and and I and I don't think the like big accounting software providers of the world were necessarily like actually on the ground talking with farmers being like what software do you want to run your business? Totally. Yeah.
2:55:33
We we ended up so a lot of accounting and FPNA software is built for for like office workers.
2:55:38
So it's built as if they assume that you're going to be in an office, you know, eight hours a day.
2:55:44
And um we actually ended up flipping that very early on.
2:55:49
and building workflows that were built for folks who spend more time in the field than in the office.
2:55:52
And so um from the very beginning we had like crossplatform across mobile and uh and desktop like sort of web- based.
2:55:59
Um, but over 50% of our users use the mobile app as their primary device to do pretty complex bookkeeping, accounting, payments, uh, receipt management and uh, you know, understanding their their analytics and reports and and I think that speaks to kind of what you were talking about there.
2:56:21
How consolidated is the industry?
2:56:23
I feel like there's this narrative of like private equity is rolled up every farm and there's like a few mega farms and the certain like billionaires own all the farmland in America.
2:56:33
But then at the same time it sounds it seems like there's still like I personally know folks who have family farms that are still in this like single digit million maybe tens of millions of dollars business like they're definitely SM there's there's a little under two million farms in the US and only uh only only something like 12 to 16,000 of them do more than 5 million a year an annual very long tail. Yeah. Okay. So super long tail. Yeah. Yeah. So I mean that Yeah.
2:57:00
That's the market opportunity for you.
2:57:02
talk talk about the the new financing you guys announced today and maybe what was the inflection point that you that that made this round kind of possible. Yeah, totally.
2:57:12
So, uh we just announced uh 26 uh. 1 million series A. Let's go. Who's it from?
2:57:20
Uh so it's led by uh Josh at Thrive and Dentures. Congratulations. He's on a tear today.
2:57:30
He got a trade deal done.
2:57:30
He got he got It's been a busy busy day for for busy day for Thrive. All three of you.
2:57:35
So yeah, busy deal for Dylan, too.
2:57:37
Um yeah, I don't even think you you caught this, but they they released S1 just like an hour ago. That's right. Yeah.
2:57:43
Um more breaking news, but but anyways, back to Ambrook. No. Yeah. Um no, super.
2:57:47
Ambrook. No. Yeah. Um no, super. Uh yeah, I just I so the inflection the inflection point for us was that we just that you know I think once we realized that we actually did need to I mentioned earlier that we at first were just trying to solve this paperwork problem uh with getting access to working
2:58:04
capital and once we realized we actually had to just rebuild QuickBooks from the ground up that that is not that is a long build non-trivial so non-trivial uh and so we ended up sitting with a group of pilot customers a couple dozen farms across the US for about two and a half years before we ended up releasing it uh and actually starting to market it more to the public. And so it wasn't until
2:58:27
And so it wasn't until last year that we uh that we um launched and early last year and that that growth basically catalyzed uh catalyze this round. So that's awesome.
2:58:39
And and will is the goal for working capital to still be a core part of the business in the long run, but you just had to solve the underlying kind of ledger first? Yeah, totally.
2:58:49
And actually, I was I was the what pushed me to this decision too was just I ended up talking to a lot of these act big act players uh that a lot of the executives there that were uh doing a lot of a financing and I asked them whether or not they could standardize it by pulling on QuickBooks and they just said it was a garbage garbage in garbage out.
2:59:09
um problem and so they couldn't even like they weren't they themselves were going in and doing pretty specialized underwriting uh that was oneoff basically per operation.
2:59:18
I figured if we were going to have any shot at getting a lot of these folks our customers access to uh not just commod you know commoditized financing because a lot of these a lot of a bankers actually drive to the farm to put together boots on the ground boots.
2:59:34
We need boots on the ground. Yep. Makes sense. Yeah.
2:59:37
sense. Yeah. which is I mean the faith-based relationships are really important but what that also what the cost is that it means that more capital can't flow into this ecosystem right and you it's really a lot harder to to do
2:59:48
really interesting niche financing as well if you don't have that level of standardization and so a lot of what Ambro has been focused on is first just solving producers problems so making easier to file taxes making easier to
3:00:02
manage receipts but now what we're starting to do is a lot of our producers for example are for the first time understanding their managerial cost in real time and able to make a decision like one of our producers, you know,
3:00:14
realized that actually replacing heers was a much more profitable enterprise than his hay enterprise and being able to make those decisions in real time and either, you know, just be but because hay equipment is pretty expensive. And
3:00:24
And so he basically he needed to increase the amount of um income that was flowing through the hay enterprise to to offset the cost or essentially, you know, stop growing his own hay and instead buy it.
3:00:36
And like those are the types of decisions that actually in real time affect profitability in a long run and build a more resilient operation.
3:00:42
And that's just not something that again it's you know these farms are pretty complex.
3:00:48
It's one of the only industries where uh oftent times for manufacturing you're usually assembling things.
3:00:54
Actually with agriculture you're disassembling it.
3:00:55
So like what are the cogs on that?
3:00:57
Yeah, I it's exciting because to me it's like you can level the playing field and like if you're a farm and you don't have access to capital in the way that these big, you know, sort of, you know, mega corporations have like really efficiently run businesses, plenty of access to capital because they're running sophisticated financial operations.
3:01:18
is just so much harder to compete and and uh with a business like farming where I think a lot of you know there's a bunch of different subcategories but so much of the revenue comes in waves seasonally based on harvests and things like that and so it's if you're getting paid a couple times a year and you can't really have consistent access to capital.
3:01:39
It's like very challenging to actually run a business and stay in business and not be forced to just sell your land or or or do something else.
3:01:47
do something else. So it's very yeah totally there's I mean and there's a huge the big trend that everyone talks about in a is just the generational transition right like 70% of land is supposed to change hands in the next 15 you know 10 15 years the average age of
3:02:01
a farmer in the United States is 57 58 it's only increasing right and so a lot of these operations are looking at generational transitions and there's a knowledge gap there as well because if everything is in you know if everyone's in someone's head about how all the finances work and how the operation runs, right? Like that's fine, but what
3:02:19
Like that's fine, but what happens when you know what happens when you need to pass on to the next generation?
3:02:24
Uh, and so a lot of what we think about is democratizing access to that knowledge.
3:02:29
Um, democratizing access to capital, to knowledge, right?
3:02:31
I think this is also our thesis on AI, which is AI should actually be the great, you know, it should be actually the like the um democratizing wave that the computing revolution promised, but ended up just actually hollowing out the middle class.
3:02:47
I think you're actually seeing that come back in a real way where you're able to tap in.
3:02:51
You know, folks who don't necessarily have, you know, specialized skill sets or like have a, you know, are a CPA or something can actually understand more about their books in real time.
3:03:02
You don't need to necessarily have that yeah background.
3:03:03
You can get at least access to the intelligence of the average Mckenzie associate, but for but for you know the cost of Amber, you know, you don't you don't have to go hire McKenzie. So that's great. Yeah, awesome. Well, super exciting.
3:03:17
Thanks for coming on the show and uh come back on when you have news. Yeah, congratulations.
3:03:22
Much like we'll talk to you soon. Awesome.
3:03:24
The uh breaking news of course is that Figma has filed publicly filed their S1 with the SEC today on the printer.
3:03:34
Oh, is the printer's going okay to list on the Oh, don't don't touch. There we go. Boom.
3:03:43
Uh they have applied to list on the New York Stock Exchange under the symbol fig. Love it.
3:03:48
It's got a great ticker fig. Congratulations.
3:03:51
The New York Stock Exchange.
3:03:54
One of the best stock exchanges if you ask me. Preliminary. I said I said gone.
3:04:01
Uh John, I I called you gone. John John Gonk.
3:04:04
Um anyways, the other interesting data point Arfer Rock called this out.
3:04:10
There's 70 million held in Bitcoin ETFs on the balance sheet with board approval for another 30 million BTC purchase via USDC. Whoa.
3:04:22
Dylan, a little bit of a BTC treasury action going on. Wild. Um, what else?
3:04:28
Would not have expected that. He just sat.
3:04:30
Bill McDermott uh who was formerly at Xerox transformed SAP uh and is now leading service now is joining Figma's board of directors. Congratulations.
3:04:42
Bill is a legend in addition to his wisdom.
3:04:45
I especially appreciate his humility and authenticity.
3:04:48
So great addition to Figma's board and well you're heading to the New York Stock Exchange.
3:04:53
You got to get a watch on bezel. Go to getbbezzle. com.
3:04:56
your bezel is available now to source you any watch on the planet. Seriously, any watch.
3:05:01
Uh yes, this is very good news.
3:05:02
Um should we run through some timeline? Get out of here. What do you think? Some timeline.
3:05:05
Uh we got another trade deal, another massive trade deal.
3:05:09
Jeff Burkovichi is joining the Wall Street Journal.
3:05:12
Starting July 28th, he'll be heading up Wall Street Journal's San Francisco team as deputy tech and media e editor. Absolutely massive news.
3:05:21
Congratulations to Jeff on the trade deal.
3:05:24
We had another trade deal that we posted on the timeline.
3:05:26
Uh Sam is moving to Thrive Holdings to build exceptional businesses designed to compound over many decades.
3:05:34
Massive signing from Thrive, our friends over at Thrive.
3:05:38
I mean, we're hearing the rumors maxed out probably four years.
3:05:41
Guessing a one-year cliff, but we'll wait for more details.
3:05:46
A veteran of RAMP, an absolute dog. Congratulations, Sam.
3:05:48
And in other news, Public, our sponsor, has the IRA triple play here.
3:05:53
1% match plus 1% BL match plus 1% match. That's 3%.
3:05:56
So go check it out at public. com.
3:06:03
The absolute dogs over at public do it again. Yes.
3:06:06
Uh Will Manitis has uh a new law. New law alert. We got a new law. Oh, he's coining. He's coining. He's coining. Citadel's law.
3:06:14
Any industry with sufficiently high stakes will end up mirroring the culture of hedge funds.
3:06:20
Massive cash comp for top performers, bida away dynamics where entire teams walk together, extreme litigation of non-competes, hardo work 247 culture, short 10 years.
3:06:32
We are at most six months away from PE style midnight recruiting coming to AI.
3:06:36
at most 12 months away from undergraduates pre-ining for the job they want after a two-year stint at a frontier lab because before because they even start.
3:06:45
There's a scrambling now Hawaii sagoponic parade here.
3:06:51
Uh it's fun to be a ghast when Bali or whoever pays hundred million to bring some PM uh an unknown quantity of their loyalists.
3:07:00
That's portfolio manager, not my favorite is when portfolio managers have their their squad. No, no, no.
3:07:06
On the the LinkedIn gap where they'll add like I'm gardening.
3:07:10
It's like gardening for months.
3:07:13
Legally couldn't go work at another uh at another hedge fund.
3:07:14
Uh after a six-month cool down, of course, over uh but this is actually much pure expression of the kind of thing that is being attempted in the Wong FB Nikita Z deal.
3:07:26
But this is actually much pure expression uh at this at the margin I think makes the capital environment for funding independent big AI things much worse.
3:07:34
Not worth underwriting a team if you know they can do R&D on your dollar and then walk.
3:07:38
In the same way, my sense is fund staking terms got worse, not better postpod shop.
3:07:45
He's getting in the weeds there, but it's a fun fun post. Cut.
3:07:50
But yeah, but this is the long-term impact of the AI telling I I do think it's accurate to say that, you know, a lot of people over the last 24 hours have been saying, "Oh, nerds are being treated like athletes."
3:08:01
I would just say this is tech is this is AI researchers being treated like PMs in in the hedge fund world. Yep. Totally.
3:08:10
So, um and the last piece of uh breaking news, I think you mentioned this briefly yesterday, but Andrew and Horowits is back in the mix to acquire Tik Tok joining Oracle and Blackstone in talks to buy the United States business. That'd be fantastic.
3:08:24
I would love to see that. Be great.
3:08:27
Well, would be absolutely fantastic. End of the show, folks.
3:08:32
Please leave us fivestar review on Apple Podcast and Spotify.
3:08:34
And stay tuned for tomorrow's show.
3:08:37
If you leave a review on Apple Podcast, we will read it on the show.
3:08:40
You can put anything you want in there.
3:08:42
You can put an ad for your business. Yes. For your venture fund. Ria on the show.