0:08
[Music] Hey, hey, hey. Hey, Black. [Music] [Music] Heat. Heat. [Music] Hey, hey, hey.
Search this video
1:15
[Music] [Music] Hey, hey, hey.
1:38
[Music] [Music] [Music] Triple Glades.
2:26
Large journalistic force headed or standby.
2:30
[Music] All [Music] right.
2:51
[Music] [Music] We came to this world to reach the stars.
3:26
We came to this world of future.
3:37
We came to this world to reach the stars to shape our future. [Music] Feel the music. [Music] Team Deathmatch.
4:47
Large journalistic force headed towards you. Stand by. >> You're watching TVPN.
5:03
>> Today is Thursday, October 2nd, 2025.
5:03
We are live from the TBPN Ultradome, the Temple of Technology.
5:09
The fortress of finance, >> the capital of capital. That's right.
5:14
And >> it is good to be back.
5:16
>> That is a lovely fedora, Jordy, if I say so myself.
5:18
A yellow fedora celebrating ramp. com. Also, >> time is money. Save both.
5:23
Easy to use corporate cards, bill payments, accounting, and a whole lot more.
5:27
If you've been sleeping under a pile of receipts, you probably haven't heard the news.
5:32
Ramp is doing a publicity stunt in Manhattan, in New York, the Big Apple, the city that never sleeps, >> with Kevin. from the office. >> From the office.
5:40
Yes, he's going to be locked in a box, I believe.
5:41
I don't know how much they've actually teased about this, but we obviously know a lot.
5:44
Uh, but there are some billboards up and some out of home ads and they look great.
5:50
Uh, and they the tagline that I did see on the timeline that is public says, "I'm CFO now."
5:53
Uh, of course, a reference to uh the social network and Mark Zuckerberg.
5:58
And, uh, it's a lot of fun and we will be tracking this stunt as >> right.
6:03
I cannot get enough of this fedora.
6:05
I might start wearing this every day.
6:08
And I believe it might be streamed live. It might be on reream.
6:10
One live stream, 30 plus destinations.
6:12
Multiream and reach your audience wherever they are. Um, right.
6:17
>> The top story today is of course about the slop trough.
6:20
How are we going to keep the lights on at the slop trough? >> That's right. >> The slop trough.
6:25
It takes an incredible amount of electricity to keep the lights on for the piggies.
6:31
>> Everyone was worried about water. Yep.
6:33
>> Turns out >> electricity may be a big deal.
6:34
>> Data centers need water.
6:34
They need water, but they recycle the water. They recycle the water.
6:38
So, the water isn't that big of a deal. Maybe.
6:40
We don't know all the data, but uh next to a data center, get ready for your uh electricity prices to keep skyrocketing.
6:49
And what and the >> uh you know, among >> uh you know, many of the issues with that uh is that it will just ultimately make huge amounts of everyday Americans genuinely hate AI because chat GPT is cool.
7:04
Yol, >> I miss when I could just Google something and I didn't have to pay an extra $100 a month for power. >> Totally.
7:22
So, uh, this is already hitting the tip timeline.
7:24
It's clearly in the zeitgeist.
7:26
This question of, uh, electricity prices, is it worth it culturally as a trade-off?
7:32
We get AI, but we pay more in electricity.
7:34
Uh whether or not you're using AI, um this has been bubbling up for a long time, but um the Sora thing seems like a big deal.
7:43
Uh I don't even know if this is a Sora video, but um there's a post from September 5th.
7:49
Oh, that must have been pre Sora.
7:50
So, this wasn't even Sora, but someone generated an AI video of a woman smashing.
7:56
play this video because there's going to be maybe one person on Earth that thinks the e, you know, massive increase in electrical >> price isn't worth it.
8:06
>> Okay, so let's play this video.
8:06
Oh god, not with the sound. I don't know. >> Okay. Uh, horrific. >> Worth it.
8:13
Worth every penny >> and uh clearly fake and weird and the glass flows through. Uh, pretty interesting.
8:19
Tyler's over there being like, we're going to train robots on that. We got it.
8:24
It's an accurate physics sim. It's worth it. We're getting humanized.
8:29
>> So, this is this is how we get to um the Dyson sphere. >> Yes.
8:32
This is the critical path to the Dyson sphere.
8:34
Obviously, if you if you're doubting that, you are so >> So, what's notable here is that the guy uh Chipotle Rowan, >> yes, >> says pretty cool that my electric bill has gone up 25% so other people can make this crap.
8:48
And he gets 300,000 >> 300,000 likes on X.
8:51
like this is the basically environment.
8:56
>> That's basically every user liked this post.
8:58
>> 8 million views, uh, 17,000 reposts. Uh, it's a lot.
9:02
And I and I wanted to dig into some numbers.
9:04
So, I I did uh fact check that 25%. I think it's real.
9:07
So, in 2021, the average electric bill for in for an average American was $1,452.
9:18
It's like a significant amount.
9:18
if you're make if you're an average American, you're making 60k or something like that or 100k a year.
9:24
Um, and now that's up to 1728.
9:31
That's exactly 25% increase.
9:31
So, uh, it does seem like, uh, electricity prices have gone up by 20 25%.
9:37
And if you look at some of the charts, um, there really is a spike right after 2020 when AI investments started ramping, electricity prices went up.
9:49
Now, some of this might be COVID, but it does seem like uh from 2010 to 2020, uh electricity prices were pretty stable, about 12 cents per kilowatt hour in the US city average on uh the Frank the St. Louis Fed.
10:04
Uh but then 2020 hits, AI comes and things have been uh up and to the right ever since and not in a good way.
10:12
JP Morgan estimated that that 70% of last year's increased electricity cost was the result of data center demand and the and the flow here is the logic is not quite perfect but it's not great either.
10:27
So it's that AI slop will be irresistibly dominant.
10:33
We've seen this with everyone who works at our company becoming entirely oneshotted by uh short form AI video.
10:41
No one can get anything done. It's a disaster. It's a crisis.
10:45
>> Of course, John's being sarcastic. >> Somewhat sarcastic.
10:49
>> There's a couple people that have been a little >> They're on notice. They're on notice. >> They're on notice. >> Uh but yes.
10:52
So, so you have to assume Meta Vibes, OpenAI, Sora will be massive successes and energy prices will skyrocket as humans spend all their time at the slop trough.
11:00
Uh I don't think personally that there'll be a fast takeoff in AI slop consumption.
11:04
I think it will kind of grow slowly.
11:06
I think this Tik Tok is just one of of many uh that people will watch.
11:10
Um but there's no denying that energy prices are rising.
11:14
Uh per capita consumption, interestingly in electricity has been down flat to down mostly thanks to efficiency standards.
11:21
Um but also there's just not that many new things that require a lot of energy.
11:25
So a lot of people are are migrating from incandescent light bulbs to LED light bulbs to more efficient cars, more efficient washing machines, dishwashers, everything's getting more efficient.
11:35
So the use is actually pretty flat, but the price is going up because we haven't been building any new uh electrical infrastructure, but that is changing.
11:42
So um the good news is that American energy production is growing with tons more capacity coming online soon.
11:49
Companies are working across all categories and there's actually a bit of a fight developing between nuclear and solar maxis.
11:57
Uh Elon is a total solar maxi, but there are a bunch of folks who are going all in on nuclear uh >> including Rick Perry.
12:03
Rick Perry >> company Fermy IPOed today at a $15 billion valuation.
12:09
Not bad for a 9-month-old company.
12:11
A land lease um and not a lot else.
12:18
>> And the hyperscalers are doing deals to bring uh decommissioned nuclear power plants back online.
12:23
I see that as a very positive development.
12:25
We've talked to probably what five different 10 different nuclear CEOs on this comp on this program.
12:32
Uh from uh Radiant to Velar to the nuclear company to uh General Matter.
12:42
There's so many companies in the space that are doing cool stuff uh in nuclear.
12:45
Um and hopefully all of that comes online although the nuclear projects are very slow.
12:52
Uh but there is more capacity coming online.
12:54
Um the uh and and you know uh the tech industry just has no incentive to become maybe more of an enemy of average Americans.
13:04
They're not they're not polling super great, but this would could potentially make it worse.
13:08
Uh >> the tech industry got through the the social media totally is terrible and bad crisis.
13:16
Everybody's sort of >> a lot of the people that were most negative around it are still power users of social media.
13:22
I I mean the uh the social network 2 like the the the leaked plot is about how >> they were like uh meddling and all the I think it was like election stuff. >> Okay.
13:32
I don't know if people are flee over it yet. >> Yeah. Yeah. Yeah.
13:35
Know people are people are upset generally.
13:36
Uh and that gives the tech industry an incentive to build more electricity so this doesn't become a bigger issue.
13:42
Sam Alman's guided towards 250 gawatts of capacity at OpenAI by 2033. That's an insane number.
13:49
For reference in 2023, 10 years before his prediction.
13:52
uh America as a whole produced average power of around 500 gigawatts, 485.
13:56
And so he's saying I need half of the power in America currently for Chacht and AI slot, right?
14:06
And and and if he doesn't build more and doesn't justify more builds, uh that would mean way higher prices, right?
14:12
Because OpenAI would be buying all of that electricity and selling it.
14:15
>> Not to mention, he'd have to choose between curing cancer, free education. >> That's true. That's true.
14:19
which he he's effectively requested.
14:23
>> No one wants to make that trade-off, >> not make him choose.
14:25
>> No one wants to make that trade-off.
14:25
So, um the uh the that 250 gawatts will not just be pulled from the grid broadly. It will be built.
14:36
There will be new builds and that's already happening.
14:38
So, uh AI companies, the problem right now is that AI companies are pulling forward energy way faster than new capacity can be built.
14:45
Like it's very easy to buy some GPUs, rack them in a tent and run the one gigawatt cluster.
14:51
Like Colossus 2 is an example of that.
14:53
Like build it and then pull the electricity from the grid.
14:55
It's not as easy to build a new nuclear reactor.
14:58
It's not it's not that easy to build entirely new energy capacity.
15:01
Uh but there are certainly plans to do that. um back in the day.
15:09
>> There's also there's also some new uh I saw a good post earlier from an account called uh Pantex Cap talking about how power companies are overwhelmed and tired of a lot of the BS that's happening in some of these new data center buildouts.
15:24
They're putting up huge barriers to weed out the BS.
15:25
If you want 500 megawatts from Georgia Power, you now have to put up 600 million upfront >> and it's held against a 10-year power agreement and withdrawn monthly. >> Interesting.
15:37
>> Alabama Power is making people to commit to a minimum of 90% of the power requested.
15:40
So, you can't ask for 500 when you want 200 and release it later.
15:44
So, a lot of people be requesting a lot and just not even necessarily confident that they're going to use it, understanding that it's an asset to have access to it. >> Yeah. No, that makes sense.
15:53
Um I mean years ago the hyperscalers Amazon, Google, uh Meta, Microsoft, they were in a very stable place in terms of energy consumption uh pre-AI boom and it was so stable that they could basically promise to go allin on clean energy going forward.
16:12
uh they made net zero pledges, they satisfied a bunch of ESG requirements, but now the AI AI race is on and so new new data centers need to come online and we're seeing that uh with these new data centers intents and all these new construction projects.
16:27
Uh natural gas, solar recommission nuclear nuclear, anything's on the table.
16:32
Uh we've heard that quote before from Chase over at Crusoe.
16:36
AI companies would burn whale oil if they could.
16:38
That the energy is so scarce.
16:41
It really is the most important resource for all of this.
16:42
And so, uh, data centers consumed about 4 and a.
16:46
5% of US electricity in 2023.
16:46
And that numbers going growing quickly.
16:52
Could be 6% in the next few years. Um, by late 2024, 6.
16:55
5 gawatts of new capacity was underway.
16:58
That's actually not just for the data center world, but just in America, they're adding 6. 5 gawatts of capacity.
17:04
Now, remember, we're almost at 500 gawatts.
17:06
So, that's like a 1 to 2% increase.
17:08
It's like not that much, but uh there are now plans for up to a 100 gawatts of of new buildout uh over the next few years.
17:18
Obviously, a lot of that, you know, is up for it's like still at the LOI phase, but that would mean a 20% like like 20% increase in energy production is sort of like being papered right now.
17:29
You can think about it that way.
17:31
like America and the business community and the energy world and the AI companies, they've all come together and said like, "Okay, let's sign and let's jump over the threshold and let's go build an extra 20% capacity."
17:40
Uh, which would be incredible because if AI flatlines uh and people don't show up to the slop trough and they're just like, "I'm good.
17:50
Uh, just give me a database query, please."
17:53
Uh, and then and they go back to 10 blue links. Who knows?
17:55
Um, then basically you have 20% more capacity. price should fall.
18:00
It's a basic supply and demand equation.
18:02
Um on the on the flip side, if tech builds the infinite justest ultimate slot machine, but nobody comes, you know, in that case, energy prices fall.
18:11
On the flip side, if demand is certainly demand is certainly there today, you can just look at Foundation Labs, OpenAI, Anthropic, Google, DeepMind, they are all seeing exponential growth in token generation.
18:23
If it's an exponential curve, get ready for expensive electricity.
18:25
But if it's a sigmoid curve and we're on our sigmoid grind sets, uh, America probably overbuilds and then prices fall.
18:31
Uh, the internet overbuild that followed the dotcom boom was pretty positive externality.
18:37
It birthed a ton of new tech companies that were able to leverage cheap broadband and uh, backall and internet services.
18:43
Uh, once the applications actually caught up, it's even easier to imagine a positive scenario with excess energy lying around because you can do even more with it than just fiber lines.
18:52
Uh but all of this is tight walking on a power line.
18:57
Uh and so interesting to keep track on it.
19:00
Sager and Jetty had a good take.
19:00
So Derek Thompson shared a a series of charts in the first slide of our deck today.
19:07
Uh Derek Thompson, friend of the show, says data centers and AI are gobbling up electricity, but the share differs significantly by state.
19:14
Between 2010 and 2025, data centers went from less than 5 percent to roughly 40% of Virginia's electricity consumption. Sweet Jesus.
19:25
Uh, and of course, the the the city the states that are not on here are like California.
19:31
California produces an immense amount of electricity, doesn't have a lot of data centers.
19:35
Uh, when we use Netflix in California, it it goes it comes out of Oregon usually.
19:40
uh when you're on the east coast, it's not the energy is not made in New York State, it's made in Virginia and then you access the US East and AWS cluster.
19:51
Um and we're seeing that, you know, with Texas and Mississippi and Alabama and all the new data center construction, not really happening in the biggest states.
20:01
And so you can look at this and see, wow, Virginia's at 40% electricity consumption.
20:07
uh glo nationally we're still at you know 4% 5% uh of data center consumption but it's rising and it's a big deal and it's going to be crazy in those state legislation.
20:15
So yeah s's point is that you know >> right now effectively single family homeowners businesses are subsidizing the basically subsidizing the cost for these data centers that are coming online. >> Yep.
20:32
Um, and so yeah, he says, "Why should we be paying sky-high power bills for Sam Alman's AI slop video generator?"
20:39
And it's a question that tech will have to answer.
20:41
Um, and then also, >> it's crazy because of course Sam Alman, leader of OpenAI, you know, was a driving force in create creating Sora, but he's also >> videos of him are are the ones are some of the most viral Yeah.
20:57
videos created on the platform so far.
20:59
on the platform so far. So it really is uh it really is him >> but >> but >> we are at this precipice where AI is cool and tech people are into it but it's becoming a it's becoming culturally like the danger here is become it becomes like NFTs right I think you were
21:20
saying this yesterday >> I think yeah potentially hated >> yes >> even faster than social media >> yes especially with the video stuff >> the video stuff has the same potential for like gh this just bugs me in the same way that the expensive monkey picture bugged people. People weren't
21:34
People weren't that annoyed by Vitalic Buterine building Ethereum and nerding out about secure blockchains.
21:42
It was when it became okay wait what a celebrity is hawking a million dollar monkey picture on a late night TV show and I'm losing my mortgage off of like why is this happening?
21:55
That's where people got really frustrated and I think it'll be the same thing if people are like okay >> I don't like this and it's increasing my >> Yeah. Yeah.
22:03
It's very different when it's like like I I have a job and there was a little bit of a boring thing and now I use chatpt to summarize those meeting notes and it's delightful and I'm okay with that versus like this stuff is not good and I don't like it and and also I'm paying for it.
22:18
that sentiment that got 300,000 likes that >> there's also there's also a growing sentiment around even the the uh non-coding use cases what the value is.
22:32
We had uh >> we had dinner last night with a very successful entrepreneur um who's not really he's he's uh you know built a a a widely known SAS company but not really in the Silicon Valley bubble not kind of caught up in the AI hype cycle himself and he said something >> a line that stood out which is AI is really good at helping people do unnecessary work faster right just like producing big reports producing research Yeah. Right. What was your example?
23:02
It was like >> and and you gave the example of, you know, if you're prompting an AI and you're like kind of bullet bulleting out, for example, an email or a document, you can kind of get you could kind of just send give somebody the bullet points.
23:16
>> That's because I have chat GPT running in my brain.
23:18
Like if you send me three bullet points and you just say, "Hey, John, imagine this is like five paragraphs."
23:24
I can just do that in my head in two seconds.
23:25
It's very easy for me to do that based on facts. >> Intelligence. >> Yeah.
23:30
Also, you know what else I can do?
23:32
I can imagine that it's being read to me by a monkey on a jet ski.
23:34
I can just imagine that if you tell me three bullet points, hey, today we're gonna talk about AI, we're gonna talk about crypto, we're gonna talk about ramp, I can just imagine a monkey saying that to me.
23:44
Like, you don't actually have to instantiate it.
23:45
You can just share the share the idea and I can instantiate it in my head.
23:49
Um but uh uh yeah the example you had was was funny was like if you're going to a fiveminute meeting and you're like I used AI to generate a 50s slide deck when in a 400page memo and it's like did that help?
24:01
That seems entirely unnecessary. >> Yeah.
24:04
Um >> yeah in you know a while back when when GPT psychosis was becoming you know more discussed I was predicting that AI was going to go through a really really rough PR cycle on par with what social media started to go through.
24:20
Yeah, >> I think that's generally happened.
24:23
>> Correct in the macro maybe.
24:25
>> Well, no, I I actually think I I think generally correct, right? Yeah.
24:27
Every legacy media compan has been putting out articles on people that are, you know, using AI in weird ways and going crazy.
24:38
>> Um or talking about how, you know, uh chat, you know, chat GBT or another product is like mentioning, you know, the word suicide a bunch to to users.
24:47
They it seems like they they released some pretty big updates. Yep.
24:51
>> Recently that basically take people from 40 back to GPT5 reality if they sense um uh behavior that's that's unhealthy.
24:58
But again, every legacy media company's been p pushing out content around people going crazy through AI.
25:06
>> Now you add in, you know, electricity costs going up.
25:09
>> Um and uh I think it's just going to continue.
25:12
>> Well, the good news is that we've been here before and there are two positive possible outcomes here.
25:17
One is uh just that the tech community comes together and builds a bunch of nuclear and solar and just power gets cheaper.
25:23
Like that would be amazing.
25:24
Uh we need more energy production and clean energy production is kind of limitless.
25:28
It's magical when it works.
25:32
>> Problem is AI usage is exponential tokens >> but maybe it's a sigmoid function and we wind up building even more and then prices do fall. It's totally possible.
25:41
Uh the flip side is that we've seen uh there are solutions to energyinttense algorithms.
25:47
We've seen this p in past with crypto.
25:50
So uh there was the same during the NFT bubble.
25:53
during the NFT bubble. there was a there was a huge push for oh uh Ethereum is using so much energy for proof of work they were able to migrate to proof of stake and yeah you know energy prices fell and it actually took a huge hit to Nvidia's uh sales that year because um
26:11
NFT companies and and and blockchain companies didn't need to buy as many GPUs because they were able to do the same algorithm effectively on a more efficient So >> you're saying Jensen is Satoshi Maybe maybe I mean they would it was extremely bullish for him for sure. Um
26:26
Um but uh but but I do think that that we have we we've successfully moved past the uh crypto is using all the energy narrative and that was a technological solution to that.
26:39
What are you laughing at? The fedora of finance. >> Fedora finance.
26:43
>> Uh well speaking of crypto privy the wallet infrastructure for every bank.
26:47
Privy makes it easy to build on crypto rails securely spin up white label wallets sign transactions.
26:50
wallets sign transactions. integrate onchain infrastructure all through one simple API >> infrastructure provider >> and you're not going to be needing a gigawatt of energy capacity to spin up a uh an application on crypto >> great line paraclet the fedora finance >> yes it's great um well um Andre Carpathy had some reactions to the dwares Patel
27:14
podcast with Richard Sutton he said his background the bitter lesson has become a bit of a biblical text in frontier LLM circles researchers routinely talk about and ask whether this or that approach or idea is sufficiently bitter less impilled meaning arranged so that it benefits from added computation for free as a as a proxy for whether it's going to work or even worth pursuing. The
27:35
The underlying assumption being that LLMs are of course highly bitter less impilled indeed.
27:40
Just look at LLM scaling laws uh the chinchilla scaling laws discovered by Google uh where if you put compute on the x-axis number go up and to the right.
27:50
So it's amusing to see that suten the author of the post is not so sure that LLMs are bit or less impilled at all.
27:56
They are trained on a giant data set of fundamentally human data which is both human generated and finite.
28:02
What do you do when you run out?
28:04
This is the data wall that was a popular uh point of contention for a while.
28:07
Uh how do you prevent a human bias? So there you have it.
28:11
Bitter lesson pled LLM researchers taken down by the author of the bitter lesson directly rough.
28:15
He says in some sense Dwaresh who represents the LLM researchers viewpoint in the pod and Sutton are are slightly speaking past each other because Sutton has a very different architecture in mind and LLM break a lot of principles.
28:31
He calls himself a classicist and evokes the original concept of Alan Touring building a child machine, a a system capable of learning through experience by dynamically interacting with the world.
28:42
There's no giant pre-training stage of imitating internet web pages if you're a kid.
28:48
Uh Dylan Patel was on Invest like the best and he keeps talking about how he has a like a baby cousin I think who uh learns by like and he keeps using this example of like the baby puts the hand in the mouth or puts the foot in the mouth and like learns immediately and it's not like the baby looked at you and saw like this is what a hand tastes like, this is what a foot tastes like or like this is what a rock tastes like.
29:08
Like the the the baby definitely learned that from first principles through exploring the actual physical environment and that's not what LLM's do.
29:15
And so there is a good point there.
29:17
Um, so there's no supervised there's also no supervised fine-tuning at which point at at which he points out is absent in the animal kingdom.
29:25
It's a subtle point, but Sutton is right in the strong sense.
29:30
Animals may of course observe demonstrations, but their actions are not directly forced or teloperated by other animals.
29:37
Another important note he makes is that even if you treat if you just treat pre-training as an initialization of a prior before you fine-tune with reinforcement learning, Sutton sees the approach as tainted with human bias and fundamentally offc course.
29:52
A bit like when Alpha Zero, which has never seen human games of Go, beats Alph Go, which initializes with them.
30:00
In Sutton's worldview, all there is is an interaction with the world via reinforcement learning where the reward functions are partially environment specific but also intrinsically motivated, eg fun, curiosity, and related to the quality of the prediction in your world model.
30:18
And the agent is always learning at test time by default.
30:24
It's not trained once and then deployed thereafter.
30:26
Overall, Sutton is a lot more interested in what we have in common with the animal kingdom instead of what differentiates us.
30:33
If we under if we understood a squirrel, we'd we'd be almost done, he says.
30:38
Well, speaking of agents, go check out Cognition.
30:43
They're the makers of Devon, the AI software engineer.
30:44
Crush your backlog with your personal AI engineering team.
30:48
So, >> Andre Carpathy gives his take.
30:53
>> If we understood a squirrel, we'd be almost done.
30:55
Andre Karpathy gives his take on the Dwarkesh Patel uh Richard Sutton pod.
31:00
He says, "First, I should say that I think Sutton was a great guest for the pod and I like that the AI field maintains entropy of thought and that no not everyone is exploiting the local iteration of LMS. I completely agree. It was a great pod.
31:11
AI has gone through too many discrete transitions of the dominant approach to lose them.
31:16
And I also think that his criticisms his criticism of LLM is not bitter lesson is not inacadequate.
31:26
Frontier LLMs are now highly complex artifacts with a lot of humanness involved at all the stages.
31:30
The foundation, the pre-training data is all human text.
31:33
The fine-tuning data is human and curated.
31:35
The reinforcement learning environment mixture is tuned by human engineers.
31:39
We do not in fact have an actual single clean actually bitter lesson pill turn the crank algorithm that you could unleash upon the world and see it learn automatically from experience alone.
31:49
Does such an algorithm even exist?
31:52
asks Andre Carpathy, co-founder of OpenAI, head of Tesla's autopilot program for a while.
31:58
Uh now working on something new finding it.
32:03
>> What is he actually working on?
32:04
>> He is working on education technology. Correct.
32:06
>> Uh yeah, it's called Eureka Labs I think.
32:08
I they haven't released anything yet.
32:10
>> He's uh he's he's he's he's in the trenches.
32:12
Um >> yeah, I think I think the >> question that I have is is you know, Ilia has said broadly they're focused on their own definition of super intelligence and that and and >> it seems like he might not just I I certainly doesn't feel like he's going to launch a chatbot anytime soon. >> Ad network.
32:36
ad network or uh business process optimizer enterprise software.
32:42
>> I wouldn't be mad I wouldn't be mad at that. >> I would love that.
32:44
>> But I would hope that he's working on something >> maybe just algorithm. Yeah.
32:46
Entirely novel >> that he could unleash on the world. >> Yeah.
32:51
I mean it makes sense to give him the budget and time to go figure that out and explore. I love it.
32:55
Uh two example proofs are commonly argued uh are are commonly offered to argue that such a thing is possible.
33:04
The first example is the success of Alpha Zero, learning to play Go completely from scratch with no human supervision whatever whatsoever.
33:10
But the game of Go is clearly such a simple closed environment that it's difficult to see the analogous formulation in the messiness of reality.
33:18
I love Go, but algorithmically and categorically it is essentially a harder version of tic-tac-toe owned.
33:24
Sorry, Go players, stick to tic-tac-toe, buddy.
33:28
Uh the sec the second example is that of animals like squirrels.
33:32
And here personally I am quite hesitant whether it's appropriate because animals arise by a very different computational process and via different constraints than what we have practically available to us in the industry.
33:44
Animal brains are nowhere near the blank slate they appear to be at birth.
33:49
First a lot of what is commonly attributed to learning is in Andrea's opinion a lot more maturation.
33:53
And second, even that which is clearly learning and not maturation is a lot more fine-tuning on top of something clearly powerful and pre-existing, the brain.
34:03
For example, a baby zebra is born and within a few dozen minutes it can run around the savannah and follow its mother.
34:10
This is a highly complex sensory motor task and there's no way in my mind that this is achieved from scratch.
34:16
Tablo raza the brains of animals and the billions of and the billions of parameters within have a powerful initialization encoded into ATCGs of their DNA trained via the outer loop optimization in the course of evolution.
34:31
If the baby zebra spasomed its muscles around at random as a reinforcement learning policy would you uh would have you do at initialization it would not get very far.
34:40
Similarly, our AIS now have neural networks with billions of parameters.
34:46
Um, Andre goes on, he does give a summary.
34:49
He says, "Anyway, in summary, overall and actionably, I think this pod is solid real talk from Sutton to the frontier LLM researchers who might be gearshifted a little too much into the exploit mode and not enough in exploring.
35:01
exploring. probably we are still not sufficiently bitter less and pilled and there's a very good chance of more powerful ideas and paradigms other than exhaustive bench building and benchmaxing and bench pressing uh and animals might be a good source of
35:16
inspiration just look at how a gorilla can bench press that's what we need to do >> is anybody actually >> working on that yeah getting gorillas come on you got the robot Olympics let's see some Olympic lifting >> uh intrinsic motivation fun curiosity empowerment multi- aent selfplay culture. Use your imagination. Uh I keep Use your imagination.
35:34
Uh I keep coming back to this idea of like uh we're obsessed with this like agentic horizon 1 hour two hours.
35:40
I'm like my initial my initialization prompt humanity's initialization prompt is like be fruitful multiply right that's genesis.
35:52
And I think there's something where humans the goal that we seek is like survive forever. Yeah. >> Reproduce forever.
36:02
And so I don't really see my life as like a string of one-hour tasks or fivehour tasks.
36:08
I see it as like a hundredyear journey, potentially thousands of years when I think about my gen multiple generations.
36:16
And so I it just feels like we're pretty far from something that is satisfying all the different criteria and of course you know we need smellvision and we're not making any progress there.
36:30
Tyler, what was your reaction to this?
36:33
>> Um yeah, I mean it's very interesting.
36:33
I think um like sudden is very RL pill.
36:38
>> Um I think like >> wait RL pilt. >> Yeah, sudden.
36:42
>> Aren't we in the RL paradigm?
36:43
>> Yeah, but it's still I mean we're we're still on like very early days, right?
36:45
I think Sutton envisions this thing is like it's just RL, right?
36:48
At its core, you're bas you start with nothing and you learn the world. Yeah.
36:52
Where the current paradigm is like first you basically mimic world of stuff.
36:57
>> Um and and you're basically like at first um when you're training a model, it's literally just like memorizing facts and then you slowly get it to generalize. >> Yep.
37:04
Um, >> so right now I the way I think about it is like for a for a modern chatbot that I'm using, there's maybe like 300 megawws of pre-training compute and then 300 megawatts of RL thrown at it and it's like a 50/50 balance or something and then maybe he's saying it should be like 100% balanced towards RL and like there should not.
37:27
Well, so so to ground this in Sora, >> please >> part of why this new app is strategic is they're generating an asset and then they get to immediately >> see feedback, >> see feedback.
37:37
Basically, the reward is more engagement in the, you know, watching it longer.
37:41
If you see a sore video and you just on on X, >> open AI is not really able to leverage that. >> Yep. >> Right.
37:50
You could indirectly be like, "Okay, this went viral."
37:51
Like that maybe that's good there.
37:53
But in the app, if you're in the in the trough, you're immediately, you know, the time that you've spent watching the video, did you like it? Did you send it?
38:01
All of that is providing feedback. >> Yep. Yep. >> For the model >> RL. Yeah, Tyler. >> Yeah.
38:07
I mean, so I think um if you look at just training, like RL is nowhere close to half of compute. >> It's like way less.
38:14
>> Um but I think it is like it is like a very cool idea to basically have this very like pure model in some sense.
38:18
It's just like experiencing the world and learning from that.
38:22
>> Um, >> but I I don't know like if you go to the the Dylan Patel um kind of example of the baby like putting its finger in it mouth to like figure out where its fingers are. Yep.
38:31
>> Like the the baby is not just like randomly twitching like its nerves are not just randomly firing there.
38:35
There's still some like prior that comes from evolution or whatever, but there is like it's not just like purely experiencing things and then updating.
38:43
Yeah, there's something there.
38:45
I was thinking about with with the zebra example like that's true but also like the baby is kind of randomly twitching in the womb and and like the baby like you can see on the on the what's it called the oscilloscope or something.
38:58
I forget what it's called.
38:58
Uh there's some sort of like device that you can like look inside and see the baby and the baby's moving like very basically but probably like doing some reinforcement learning essentially like like pre-birth I think. I don't know. >> Yeah.
39:12
>> Yeah. I mean on some level like there is some encoding that is going on into the baby of like this is how like you can walk around or something like that like there there's some like I I think if you just had a baby and like without any other people around >> y >> um >> like I maybe it could like learn to walk
39:30
and stuff but there there's some like I I think there is still some imitation going on generally like Sutton is saying that like none of it is imitation it's all just like RL you're experiencing world and you're updating on those on the actions you took, you have some goal, >> you either like complete the action or you don't and then you update. >> Yep. >> Yep.
39:47
>> I think um >> I think imitation is like more uh important, right?
39:52
There's like the kind of um like the Gerardian take or whatever is that like um >> humans are are apes, right? To apeist to imitate.
40:01
>> If you drop a baby in a prison cell from birth, it'll never learn to walk because it needs to see a human walk to learn that. >> Yeah.
40:07
Something >> I don't buy that.
40:08
I think they'd figure it out. >> Okay.
40:10
Maybe not walking, but like doing, you know, useful things like it's not going to >> I think I think I think if my son was born in a padded closet and just given food and an Excel sheet, he'd be writing VBA macros in five years with no external input.
40:27
>> Well, maybe it's been encoded in >> pure pure pure reinforcement learning experimentation from first principles. I think it's possible. I don't know.
40:36
Anyway, Figma, think bigger, build faster.
40:38
Figma helps design and development teams build great products together.
40:41
Hey man, just wanted to say I think we talked about this yesterday, but I love it so much I put it back in.
40:47
Uh, hey man, just wanted to reach out and say I loved how much you drank at the networking event last night.
40:51
There there are a few of those.
40:52
This is becoming a new there was another one >> for I'll find the >> find the other one for me in the meantime.
40:59
>> Yesterday apparently there's a game that makes you a YC partner for a day.
41:02
You watch old pitches and guess if they made the batch.
41:06
I wonder where this data came from.
41:08
Usually YC is pretty locked down with these with these old pitches, but uh this seems like a fun game.
41:12
I went to the top of Hacker News yesterday. >> Here's the other.
41:15
Hey man, just wanted to follow up and say I admired your ability after three drinks to persistently ask who had a bag at the super serious networking event last night.
41:24
>> Super serious networking event.
41:24
AI could never um >> AI can.
41:28
>> What do you think about I was I was texting with Tyler about this like uh there's this post.
41:32
Sam Alman is playing 4D chess.
41:34
Sora 2 is about to take over social media.
41:37
The virality is guaranteed once this scales.
41:39
Billions in ad revenue will flow straight into more compute, fueling the flywheel.
41:44
In a year, Sora 2 will be so efficient and cheap that margins explode.
41:47
You can't out accelerate Sam Alolman.
41:49
And I wanted to reflect on this idea that uh Sam said, "It's way less strange to watch a feed full of memes of yourself than I thought it would be.
41:58
Not sure what to make of this."
42:01
Gab, >> again, my read on this is I don't uh I don't fully believe that he actually I I don't know.
42:10
I don't know how to read into this other than he is very incentivized to normalize people being deep faked all the time, right?
42:18
>> I mean, it's a crazy crazy move to just be like >> my likeness. I'm going first. My I'm merging. I'm merging.
42:25
It's like >> him going first being on the first SpaceX rocket. >> Hey, it's not so bad. It's not so bad.
42:29
I don't want to turn turn this off.
42:32
I don't I don't want to see that.
42:34
>> I I mean, I don't like the Skibbity toilet thing to begin with and then you mash it up and it gets even more horrifying.
42:39
>> Anyways, he uh he went first.
42:39
He's uh very incentivized to get a bunch of other big names, personalities to agree to do the same thing.
42:50
>> Did is this >> I would imagine it's very weird to watch videos of yourself committing crimes.
42:56
You know, there's a video of guess a video of him in blackface floating around.
43:01
>> Um, there's a bunch of videos that I think would be incredibly strange to watch of yourself.
43:06
Um, so, >> but stranger than your expectation because I would expect it to be weird.
43:11
I I expected it to be weird when Tyler sent me a deep fake of myself bench pressing and then turning into a golden retriever.
43:17
I was kind of just like, h, yeah, this is this is below my expectation.
43:22
Like, it is less weird to me. I don't know.
43:24
You haven't seen one of yourself yet, right?
43:25
Has anyone made one of you? You haven't uploaded.
43:28
>> I don't think I've done I haven't done the cameo.
43:30
>> I haven't agreed to do thing.
43:32
>> What What about you, Tyler?
43:32
Did you expect it to be weird to see yourself in a Sora video?
43:35
And then was it weirder or less weird than you expected?
43:39
>> Uh, I don't think it's that weird, but it's also like it's still not like >> that good where like it's kind of me, but it's not like all the way there. >> Totally. Totally.
43:48
But I think Okay, Jordy, do you think that him basically releasing a social media where like half the content is just him, is that an aura farming move?
43:58
I think it increases his fame, which is a resource that he's able to leverage, but I don't know that it it in, you know, a lot of the videos that that I've seen don't uh I don't think are necessarily good for the brand, but it it's also just putting out so much stuff into the world that it just all gets diluted and you don't read too much into any one video because it's not created by him.
44:26
Do you believe in cosgro's law >> that oraura farming is all you need? >> Break it down for us. >> Always be or farming. Okay.
44:33
I think I I have a text somewhere, but um yeah. Okay.
44:37
Basically or farming, right? >> Yeah.
44:39
>> Um >> you can think of like most things are are on some level like based off like interacting with other people, right?
44:45
Like like any kind of if you're uh you know raising around, if you're um going to make some deal, right?
44:50
It's based off like your interaction with someone else. >> Yeah.
44:52
>> Yeah. people raise want to raise from tier ones because every dollar is the same but you get to orura farm the >> you know the the GP >> yeah and then it's like like in like when you're raising around you want to be ora farming the V VC right because they want to be like oh this guy's like pretty sick like I want to give him
45:09
money or or something like this right so basically the idea is always be aaura farming right you always want to be doing something that increases your aura because everything you do like any kind of interaction you have will be influenced by like your personal aura Yes, but stretch the analogy >> a you fedor of ants >> uh but stretch the analogy to the absolute limits. Sometimes when you farm
45:29
Sometimes when you farm your lands growow and you have a you have a you have a failed harvest and and you can also salt the ground so that nothing will ever grow.
45:42
And so there is a world where you go out to or a fararmm, you set up your aura farm, but someone has salted the ground and you don't grow a single crop. >> Yeah.
45:54
So that's like a I mean that's a failed ora farm. >> Failed aura farm.
45:56
So the question is is this the sword to putting yourself in the cameo?
46:02
It is it is an attempted aura farming.
46:04
Was it a successful aura farm?
46:07
Yeah, I think most like I would say in most cases or farms you can only tell the success of a farm uh in retrospect >> if it's successful. >> Yeah.
46:16
So, so it's like in six months we'll see.
46:17
I think if people are still on Sora Yeah.
46:19
If if the app is still like being used as a social media platform, not just like a creative tool. >> Um >> Yeah.
46:26
>> Then I think it will be a successful farm because now Sam is just, you know, he's just everywhere.
46:30
It does it does feel like there's always a winner and a loser in an aura farm and like if you are or fararming VCs will do this they'll they'll buy secondaries in a company to say they're an investor even though they didn't participate in a primary round. >> That's true.
46:46
And then also for every time for every time a a founder or fararmms a tier one venture capitalist and then blows the company up in a ma major scandal uh that reduces the aura of the fund that that put the money in and they offend they got farmed. They got farmed. Yeah.
47:04
It makes if you backed FTX you got farmed. You you you look bad. >> Yeah.
47:09
Like there's always two sides, right?
47:10
Because in Jord's example, um the the VC is farming the company, but when it get when it gets exposed that they actually bought secondary, that they weren't real investors, then >> it's a negative aura.
47:21
>> The company is farming the VC.
47:21
They're like, "Oh, we're so great that >> there must be positive sum or a farms where two people come together, they >> they grow a beautiful garden."
47:31
When an emerging uh fashion brand partners with a legacy brand, let's say a streetear company partners with Solomon. Yep.
47:39
They can both benefit because the streetear brand is getting new legitimacy and Solomon gets to be like culturally relevant and cool. Yes.
47:48
And that is a positive sum or a farm. >> Yes.
47:53
Uh Sydney Sweeney and American Eagle positive sum or a farm. The stock's way up.
47:57
Everyone's doing well in American Eagle world.
47:59
Sydney Sweeney is back in the news and dominating. It seems like a win-win.
48:03
They or farmed each other successfully.
48:05
They grew a new bountiful harvest together.
48:10
>> It was more positive for American Eagle.
48:12
It was more like neutral. Okay.
48:14
>> I think for Sydney Sweeney and that she got a bunch of blowback from it.
48:16
She got a bunch of attention, but it wasn't all positive.
48:20
Whereas American Eagle just the stock ripped.
48:22
And um anyways, I have I have an interesting post here that I wanted to throw in.
48:27
It's from a guy named Robert.
48:29
If we can pull it up in the timeline.
48:31
>> Let's pull that up and I'll tell you about Vanta.
48:33
Automate compliance, manage risk, improve trust continuously.
48:36
Vanta's trust management platform takes the manual work out of your security and compliance process and replaces it with continuous automation. >> There we go.
48:43
So, this post went super viral.
48:46
This is Robert says, "Good game, everyone."
48:50
And he's charting total job openings versus the S&P 500.
48:52
and he shows that when chat GPT was released, there was this massive divergence.
49:01
Now, a lot of people took this at face value and they thought, okay, ChatGBT is just killing jobs.
49:08
>> But I threw in another post here.
49:11
>> Um, anybody that's used ChatGpt understands that it's valuable, but it doesn't replace them at work. >> Yep.
49:17
Like maybe I I I don't I haven't met somebody that's been like, "Hey Jordy, like I have to admit I lost my job because chatbt just does it for me." Right. Yeah.
49:25
It's there's certainly in key areas like Finn is like replacing like the need to add more CX reps.
49:36
>> Um but uh but by and large chat GBT was not the catalyst.
49:39
That's just like you know totally reducing the number of job openings.
49:43
So there's another post in here from Simon Saras uh who says the zero interest rate era is going to become lost history because people want to make up a narrative around AI.
49:52
The white collar bloodbath didn't happen because of a chat app release.
49:56
It happened because of the end of Zer which occurred rapidly in 2022.
49:59
So >> literally when FTX like chat GPT was released when FTX collapsed >> Yep.
50:08
>> effectively during you know within like the same 30-day window. Yep.
50:10
And so, uh, it really was a lot of CEOs h woke up and said, "We have to get a lot more efficient.
50:20
We have to get more lean.
50:21
Let's not hire, you know, a ton of new people.
50:24
Let's do more with what we have."
50:27
>> The AI narrative from a from a from a job loss standpoint has also it's also in some ways I feel like been a lot more narrative driven and that CEOs h have a massive incentive to show they're getting more efficiency out of AI. Yep.
50:41
>> And so they're just saying to their teams like, "I don't want you to hire more people. Just be more efficient." >> Yep.
50:46
>> And even that by itself of just forcing people to say like, "No, we're not expanding your team. Just do more with less." >> Yeah.
50:52
There was also a huge moment contagion from the Twitter buyout where X was able to run with like 1/5if the staff and the service doesn't really go down in their shipping features.
51:01
And so I think a lot of CEOs just kind of adopted a year of austerity, a year of efficiency and uh wanted to test what they could do with less and so they pulled back.
51:12
So you have to look at the labs are scaling headcount rapidly.
51:17
Anthropic is, you know, hiring teams to, you know, you know, capitalize on all the demand they have >> and they're some of the most, you know, AGI pill of anyone, right?
51:26
And so you have to >> read into what what people are are actually doing.
51:32
actually doing. Um and certainly >> there's been massive job job creation in in the kind of startup ecosystem broadly just because of you know we were talking yesterday with a friend seems like every we've it's become so normalized but it
51:48
feels like every day there's like just on TVPN like at least 100 million of new funding announced usually like you know half a billion at least >> uh at least lately and so all you know a lot of that is going to go into hiring teams. Y and um so anyways, this this
52:04
Y and um so anyways, this this original chart thought it was worth calling out.
52:10
>> Yeah, it's important to reality check the put one chart over the other and see if they line up perfectly.
52:14
We had some updates from Tyler Cosgrove.
52:17
Yeah, there was another chart we we were asking.
52:22
Somebody had made a chart tracking the NASDAQ uh.
52:24
com bubble to today >> and this person effectively did graphic design to make them look like they match up perfectly like by the day. >> So perfect. >> It looks so perfect.
52:37
And Tyler spent 24 hours this week trying to recreate it.
52:43
and >> no matter which index he was looking at it.
52:48
>> Yeah, take us through the thought process.
52:50
>> Yeah, I I just sent it in the chat, but basically um it's it's supposed to be uh NASDAQ um from 1990 2001 and then >> overlaid on top is 2023 to now. >> Yeah.
53:01
>> Um and then yeah, I mean they just like don't line up like you like you can kind of see like there's a little spike and it goes down a little bit, rises.
53:07
Um, >> so the person just like made up the data or something like no one would ever do that.
53:14
>> I don't know if it's like literally made up, but this is the original post, right? >> Yeah.
53:19
So time scales are are not lined up correctly. >> Okay.
53:23
>> Um, and then >> Oh, so they compressed time on one of the charts and not on the other. >> Yeah.
53:29
>> Even when even when I did that, I they still wouldn't really line up.
53:31
So I think certain parts of the graph are more like uh compressed than others to make the kind of, you know, >> Yeah.
53:38
So it's like a linear scale and then a log scale and then a linear scale.
53:40
It's the ultimate chart crime.
53:42
>> Yeah, this is I think this is pretty bad.
53:44
>> Do you have your actual overlay available? Can we look at that?
53:48
>> Yes, let me let me pull it up.
53:49
>> I would love to see that.
53:49
While you pull that up, let me tell you about graphite code review for the age of AI.
53:52
Graphite helps teams on GitHub ship higher quality software faster.
53:58
>> I love that when I hit that sound effect.
53:59
I know you're going to amp up the fire me up.
54:02
Uh while he's pulling that up, uh John G shares websites. Should be this again.
54:09
Waiting for the first company to authentically do it.
54:13
Everything looks like a venture capitalist yoga retreat website at the moment.
54:17
It's from the Sony website in 2002, Web's design history.
54:20
We remember when we pulled up the old iPhone website on Apple in 2008, and we were like, "Oh, it's going to be like all iPhone focus."
54:30
Because it was the iPhone moment.
54:30
It was like the MacBook Air and like the iPhone's off to the side and it said high technology is one of the features that you get with the iPhone. Uh I I like this.
54:41
I wonder if this would if this would convert how bad this would be.
54:44
I wonder if someone's actually going to take this post and run with it and build a site that looks like this.
54:48
Um we've seen some cool people do interesting things with the with web design.
54:55
Post Hog had a very cool website.
54:56
There's still room for innovation in the website. Have fun. Do something different. Don't just copy linear.
55:03
Our sponsor linear is the tool for it's a purpose-built tool for planning and building products.
55:10
Meet the system for modern software development.
55:13
Streamline issues, projects, and product.
55:15
>> All your favorite products are built using linear.
55:18
That's all I need to know.
55:19
>> Do we have the actual truth zone edition of the stock chart? Okay.
55:23
There's some overlap there. >> Yeah.
55:27
I mean, they're both just rising.
55:30
>> Yeah, they're both rising.
55:31
>> They're both up and to the right.
55:33
>> But yeah, they're it's a wildly different time scale. Yeah.
55:35
And that and that little like like that like dip and then pick back up does not line up at all. >> Driven.
55:43
>> You can line up certain parts of it. Yeah.
55:45
That that was the big that's that big dip.
55:46
And you can like line up certain parts of it, but then the rest of it doesn't doesn't work. So, yep.
55:51
>> If you shift it around, like if you shift actual time scales of certain parts of the graph, >> then I think I could get it to line up.
55:57
But You need to copy the red line and make it a different color and just offset it by a little bit and be like, I'm going viral. That's the secret. That's the secret.
56:05
Well, if you want to do your own analysis, head over to julius. ai.
56:10
Chat with your data, get expert level insights in seconds.
56:12
It's the AI data analyst that works for you.
56:14
Um, and so back on Sora 2 and the question of energy, uh, Zephyr says, "Pretty sure OpenAI made an algorithmic breakthrough.
56:26
Sora 2 is a much smaller model.
56:26
Tyler, I don't know what you think about this, but Andrew Maine is is saying another interesting metric is that V3 costs $320 per 8 second clip. You get three a day. That's 10 bucks a day. I get I pay 250.
56:43
So, they're basically pay like if I use all my V3 credits on Google's top tier of Gemini, I'm break even.
56:50
Uh, but uh, you know, Sora currently gives you a hundred 10-second clips per day.
56:55
That's $320 of video generation per person for free right now.
57:01
If that's what they're doing, I think they'd eat that cost all day long. What do you think?
57:05
Do you think that they've actually brought the cost down to 32 cents per clip to 3 cents per clip?
57:10
Where do you think it's going?
57:13
>> Yeah, I mean, I would be very surprised if it was like the same kind of cost as a V3.
57:16
Um, I would, if I had to guess, I would say maybe like 10x.
57:20
So maybe it's like 30 cents.
57:22
But even then, that's like pretty steep.
57:24
That's like pretty hard to to squeeze that out.
57:28
>> I just think like they gave me a hundred.
57:30
I've generated six over three days. Have you generated 300? No way. >> You've done five.
57:37
And And we're like kind of power users, early adopters really testing this for work.
57:41
I imag And then they're also rolling out the invites slowly.
57:44
I wouldn't be surprised if this is Yeah, it's three bucks a a prompt, but like they're just going to eat that for a while until it gets to scale where I mean, if it's a million new videos a day coming on the platform, that's only three million bucks.
57:59
That's only a billion a year.
58:01
Like, they can eat that, right?
58:03
If >> it's only a billion a year. >> It's only a billion.
58:07
That's really nothing for OpenAI.
58:08
They're 500 billion dollar company that's raising tens of billions of dollars every second.
58:13
and they and and and what's at stake like winning the next social platform like like really eating lunch the trough wars.
58:24
The trough wars are going to be fought one 10-second clip at a time no matter what it costs.
58:29
Well, people are still >> doing things the oldfashioned way.
58:32
Ramp is doing things the oldfashioned way.
58:33
We talked about this earlier in the show, but Kareem says Sora 2 is insane. Unbelievable.
58:39
and he got a community note on it because RAMP uh published a launch video that it was shot with a real camera and the community notes team put him in the truth zone said it was not created by Sora. It's a real commercial.
58:51
I think that helps with the virality and uh it helped.
58:53
Kareem got 400,000 views on this. 450,000 views.
58:58
>> I wonder what percentage actually just believed it was sore, too.
59:02
>> I I saw someone in the chat saying they think it's sore.
59:04
I don't know if they were joking, but I I I do think some people uh you know, it's hard to tell.
59:08
It's hard to tell these days.
59:08
Uh >> um well we have an amazing announcement. What's that?
59:13
>> Which is that there are 19,000 private equity funds in the US and there are 14,000 McDonald's in the US.
59:20
>> We have more private equity funds than McDonald's.
59:22
And I just I think it's important to uh >> recognize this moment, this milestone.
59:28
>> I wonder how that's calculated.
59:28
Is that like does that count like an LLC that holds a sports car in Montana is technically a private equity fund or or are we talking like there's a real office with a couple people that are buying companies with that and equity?
59:45
Like is it a true private equity fund? Like what's the scale?
59:46
There's probably some crazy power law, right?
59:48
Where like KKR is like here and then there's like some company funds for ants.
59:55
>> What's the median size of that private equity fund?
59:57
I would love to know that.
59:58
Anyway, we can dig into it more.
59:58
Um, speaking of budgets, the government is about to shut down and poly market.
1:00:06
Isn't >> it already shut down?
1:00:07
>> I believe it did shut down. Not good.
1:00:09
Uh, the party's fight to cast shut down blame.
1:00:13
>> I, you know, you know, do people like the government? I don't know. Do It seems like a mess.
1:00:18
Delian was on the show laughing about like how no other country just shuts down the government.
1:00:24
Uh, do you have the poly market on when the shutdown might end?
1:00:26
That seems like a useful uh useful data point uh for us to track. Pull it up.
1:00:32
>> Hours into the government shutdown, the White House issued new threats.
1:00:35
Congressional leaders were digging in and some rank and file Democrats and Republicans were searching for a way out.
1:00:40
Senity Senator Senate Majority Leader John Thun and Vice President JD Vance said they are open to talking to Democrats about extending expiring healthcare subsidies, but won't negotiate as long as the government is closed.
1:00:52
Democratic leaders said they won't agree to reopen the government until their healthcare demands at least some of them are met.
1:00:57
And when does polymarket think it's >> vast majority of folks over 70% uh think that it'll be October 10th or beyond.
1:01:06
44% believe >> October 10th.
1:01:08
That's our one-year anniversary.
1:01:09
I believe >> really >> something like that.
1:01:10
So you know hopefully the government reopens and we can celebrate the one-year anniversary of this show.
1:01:16
Uh, in other news, Apple is shelving development for the cheaper, lighter Vision Pro, the Vision Air planned for 2027, and they're prioritizing a more urgent effort developing AI smart glasses to rival Meta. This is interesting.
1:01:32
There's some debate over whether or not this is actually going to happen.
1:01:36
Um, Apple's a big company.
1:01:38
They could just be working on multiple things, but uh, the news from Mark German is that they are reorienting.
1:01:44
So, the company Apple had been preparing for a cheaper, lighter variant of its headset cenamed N100 for release in 2027, but Apple announced internally last week that it's moving staff from that project to accelerate work on glasses, according to people with knowledge of the matter.
1:02:00
Smart glasses have emerged as a critical arena for tech companies, which are racing to develop AIcentric devices.
1:02:06
>> That's crazy that Meta's making them dance. >> Making them dance.
1:02:11
>> Duck's making him dance.
1:02:11
Tim Cook was watching our Metaconet coverage and was like, I want to rip that gong with those guys. Let me do it.
1:02:17
I gotta make some glasses >> someday. Someday.
1:02:21
>> I I gotta get my own Rocco Bassilisk.
1:02:21
I got to get >> There's only one >> my my own Rocco. Yeah. What What could they do? Oliver Peoples.
1:02:28
Is that owned by Lexodica as well?
1:02:32
>> Also >> also owned by Lxodica.
1:02:32
Where where will Apple go?
1:02:35
I think Apple will not try and partner with a fashion brand.
1:02:37
I would make that prediction that they will do their own and they will look like AirPods and they will be mostly all white and they will be a big status symbol to let people know that you have an Apple product on your face and they will be a couple different colors.
1:02:47
There will be different colors every year but Apple will try and own that design language.
1:02:52
The Apple has done a few partnerships.
1:02:54
Didn't they do an Hermes band for the Apple Watch?
1:02:56
I believe they did that and a few other collaborations, but Apple generally has been pretty limited on the uh let's partner with Nike.
1:03:07
They they they've done a little bit there on the running side, but in general, they have not done a lot of like this iPhone is co-designed with another brand.
1:03:16
Uh it's just you can get a case from them, we'll give them the design.
1:03:20
It's up to them to do whatever they want.
1:03:22
You can get the Louis Vuitton, you know, co colors and whatnot design.
1:03:27
>> Um one more post before our first guest.
1:03:29
Uh there's a post from Modest Proposal uh who's sharing a screenshot from a transcript of Patrick's interview with Dylan Patel.
1:03:36
Patrick says, "Yeah, he's doing demand guarantees.
1:03:39
He's doing all this crazy stuff right now."
1:03:40
Of course, they're talking about Jensen and Nvidia.
1:03:44
Dylan says, "Yeah, right. Right.
1:03:44
He's using his balance sheet to win, try and win more, which is an interesting dynamic.
1:03:50
I don't know if there's ever been anything this in terms of the anti-competitive nature of this where you backs stop clusters. core.
1:03:55
We've recently got a deal with Nvidia where it was where they backs stopped a cluster, right?
1:04:01
Uh so that's saying like if you buy this Y and and bring it online, we will guarantee the demand is there.
1:04:07
Um Dylan continues, "Now Cororeweave would have never built this cluster because it's for short-term demand and renting GPUs on short-term is a terrible business model.
1:04:17
You want to do long-term contracts and you want to do long-term contracts with people with balance sheets. That's the golden goose.
1:04:22
But that doesn't exist so much. Important point.
1:04:26
So you do long-term contracts with people who don't have a balance sheet. OpenAI.
1:04:30
And if you can't do that, uh then you'll do short contracts with people who do have a balance sheet.
1:04:36
There's this whole matrix of who you rent GPUs to.
1:04:37
But from Nvidia's interest, it's uh you know what I really love is when venture capitalists fund a company and then 70% of their round is spent on compute.
1:04:45
They effing love that talking about Nvidia, right?
1:04:49
And that's what's happening with all these companies, whether it's physical intelligence, they're spending a lot on robot arms and too, but they're also spending a lot on compute or it's any other startup that's raising cursor, etc.
1:05:02
And even even if it's not directly, it's indirectly going to GPUs.
1:05:04
They love when people spend their entire round on GPUs.
1:05:08
Would be really good if this wasn't a two-year deal or a three-year deal for that compute.
1:05:12
If it was, you can spend 70% of your round on one training run.
1:05:16
leave a company with these ideas, gather the data, do the training run, and then you have a product and you show how good the model is and try and raise again.
1:05:23
That would be really great for Nvidia.
1:05:25
But no one wants to build a cluster who's predicated on that as a business model. That's crazy.
1:05:29
So, they have to backs stop a cluster to do that. >> Yep.
1:05:33
Uh before we bring in our next guest, uh there's a related uh article in the Financial Times.
1:05:39
One of Britain's best known tech investors has warned of a disconcerting rise in artificial intelligence valuations, saying NVIDIA's planned 100 billion investment in open AI brought uncomfortable echoes of the dotcom bubble.
1:05:52
This is about James Anderson who is at Bailey Gford.
1:05:55
Uh he had a massive position in Nvidia.
1:05:59
He was the largest that was his largest holding.
1:06:04
He's now in Chinese battery maker CL kettle.
1:06:06
Uh and he said, "I have to say the words vendor financing do not carry nice reflections to someone of my age."
1:06:16
And so um the uh yes, the echoes of the dot bubble are growing louder.
1:06:24
Well, uh we are pleased to be uh joined by someone who is building in a completely different space.
1:06:29
We have the chief operating officer of Axon.
1:06:33
>> The president >> the president of Axon. I was reading the chat. little misinformation.
1:06:38
>> What's going on, Josh?
1:06:39
>> Welcome to the show, >> boys. Good to see you.
1:06:40
Thanks for having >> Thanks for hopping on.
1:06:44
>> What is this setup here?
1:06:44
Uh that that looks incredible in >> space.
1:06:47
You're in the you're in the command center. >> Yeah. Yeah.
1:06:51
We have uh some folks at Axon who just on their own come out, they just churn out new Zoom backgrounds.
1:06:56
So, every time I open Zoom, there's like five or six options to choose from that I've never seen before.
1:07:02
So, this one looks pretty cool. >> It's looking great.
1:07:04
uh can you take us through kind of the shape of the business today?
1:07:07
Maybe a little bit about your journey just to kind of set the table for those who might not be familiar with the business. >> Sure.
1:07:13
Uh at Axon, we make uh pretty much all the cool technology that police officers use day-to-day.
1:07:18
Everything from their taser gun to uh their body camera to all the SAS software in digital evidence management to all their reporting products, drones, virtual reality police training.
1:07:30
We really believe that uh we can add a lot of value in a lot of different uh spots where really software and hardware intersect in policing.
1:07:37
And so uh I've been at the company 16 years, two years or two months, sorry, two weeks after graduating college, you drove across the country to uh to join Axon and here I am. >> That's amazing.
1:07:49
Uh what's been the most recent uh transformation in the business?
1:07:54
I was talking to somebody uh about this category and uh very experienced investor.
1:07:59
He was like, "Oh, no one's really made it work."
1:08:00
And I was like, "Well, what about like Axon?
1:08:02
They seem to be doing pretty well."
1:08:04
And he was like, "Ah, it's not that big of a company."
1:08:06
And then we looked up the market cap and it's a huge company and you've been massively successful.
1:08:08
Uh, but what what what has been the ma like the biggest transformative moment over the past five years for the business? >> Sure.
1:08:18
I think uh really for the first 25 years of the company's history, it was really about hardware and then the last five years it's been far more about software.
1:08:28
So, we manage more than 40 times as much content as the Netflix library combined.
1:08:32
So, we're managing all of this police evidence in the cloud.
1:08:37
Uh, and in figuring out how to make use of this ma massive data set for our customers has been the name of the game.
1:08:44
So, most notably, uh, a couple years ago, we released a product called Draft One, which is a Gen AI product.
1:08:48
Analyzes the body camera transcripts and creates the first draft of the police report for the police officer.
1:08:55
So cops spend about 50% of their time writing police reports.
1:08:59
We think we can make that like 10% of the time. Good use of >> GI.
1:09:03
So is that something where that an officer in the middle of an action, let's say they're pulling somebody over for speeding, do they talk out loud and say, "All right, I'm you know, how how are you how where is that data coming in from?"
1:09:16
Because I could imagine they're just kind of John actively uses like transcription with chat GBT.
1:09:20
So, we'll be like talking about something and then he'll, you know, be prompting it just with voice.
1:09:26
I could imagine an officer does the same thing where they're saying like, "Hey, I'm pulling this over.
1:09:29
I'm at this, you know, intersection.
1:09:31
I, you know, clocked them going 20 over."
1:09:35
And and then that just gets built in or what is that workflow like? >> Sure. So, it's a great point.
1:09:42
cops are getting better at prompting the AI to uh to, you know, improve the quality of the police reports.
1:09:50
But even without doing that, uh just in the normal conversation, you can imagine that same incident when the cop walks up to the person's window and says, "Hey, you know, you're speeding, you're going 20 miles an hour over the speed limit, still pick that stuff up." >> Yeah. Yeah.
1:10:05
And even just like uh like, "What's your name?"
1:10:07
name and then the person says that and then that's in the audio file and then that gets transcribed and that's just like so much manual work.
1:10:14
You're not putting anyone out of a job.
1:10:15
You're just making their life better.
1:10:17
It's it it feels like a fantastic use of artificial intelligence. Very exciting.
1:10:22
>> And you guys have been you guys have been super inquisitive.
1:10:23
I think you've done a number of acquisitions but recently completed uh it was an 800ish million dollar acquisition. We got a gong here.
1:10:32
We'd love to we'd love to you know.
1:10:35
>> Yeah, it was about 630 million. It was prepare AI. >> There we go. >> Love that.
1:10:40
We have a gong as well at headquarters. >> Big deals. You hit the gong. I love that.
1:10:46
>> We're excited about those guys joining the company.
1:10:48
Their founder, Michael Chime, super super talented uh and very happy to have him as part of Axon.
1:10:54
>> Uh is is is the strategy with that deal uh to kind of diffuse the technology over everything you do, keep it as a separate business line, cross-ell it?
1:11:03
How are you thinking about that? Yeah.
1:11:05
So, our mission is to protect life.
1:11:07
We want to keep people from getting killed in in policing incidents to include police officers.
1:11:11
And the best way we can do that is shrink the time between the incident taking place in help being on the way.
1:11:19
Right now, that's like a twominut cycle.
1:11:21
Someone calls 911, they talk to a call taker, call taker talks to the dispatcher, dispatcher figures out the right officer uh to uh to send to the scene.
1:11:30
And we think we can automate that entire process uh in partnership with prepared.
1:11:37
So essentially call 911 based on the metadata of the call drone is already taken off uh to head to the incident.
1:11:42
Uh you know over time we think we can route the proper police officers uh to those scenes without a lot of human intervention.
1:11:50
And so that's a lot uh of what we have to look forward to uh with with prepared moving into the future here.
1:11:56
you guys are at the stage where you're a true platform multi-roduct.
1:11:58
you're a true platform multi-roduct. Uh I'm sure you started with you know I don't know the exact uh the the the first product but uh in this category is it how hard is it to get point solutions through I imag if you're just building
1:12:16
for other startups like you know a typical YC company for example they can build a really narrow simple point solution sell that into somebody in their batch it's pretty easy you know these aren't um you know super sophisticated operations whereas selling into police departments. I imagine just
1:12:30
I imagine just the um just just the timeline alone of of building that relationship is probably really hard for for a young company.
1:12:40
I guess like talk about the different maybe stages of the companies and getting to the point where you're truly a platform. >> Sure.
1:12:47
The the biggest part is you know you have to recognize that US public safety is less than a million total users.
1:12:53
So a lot of uh startups wouldn't even be able to get funding based on that TAM you know and so when you go into public safety or uh defense tech or you know this this segment you really have to be confident you can build something for the long term where you can win the majority of the users on your platform.
1:13:11
That's really the only way you can have a viable business model in this space.
1:13:15
And so uh you know we look at this like hey you know there we start with the most talented teams we can find and then the most product synergy.
1:13:23
And so with a company like prepared or one like Fus that we acquired last year we really believe we can help accelerate their growth by plugging them into our ecosystem.
1:13:31
And you know the products nicely play off each other and uh you know with police they want to outsource as much of the technology as they can.
1:13:41
they don't want to build things internally uh to uh you know to to capture their use cases.
1:13:46
And so when we can show up with a really nice endto-end workflow from what we call capture to courtroom uh that that really has a lot of value to our customers.
1:13:56
And so that's part of our our thinking and hypothesis as we look at M&A and which companies to partner with.
1:14:03
In terms of that platform, uh, is there demand from local law enforcement agencies for you to play nice with other companies, build some sort of open API?
1:14:16
Is there another kind of enterprise resource planning backbone that they want you to plug into?
1:14:21
What's kind of the the software supply chain of a modern police department look like these days?
1:14:31
Yeah, it's a great question and the short answer is yes.
1:14:33
You want to build something that's open that at the end of the day, this isn't about trying to help people make more money or make more effective ads on the internet.
1:14:41
This is about fighting crime and saving people's lives.
1:14:45
So, when you do those kind of sketchy things where people don't have access to your data or you don't, you know, you don't work well with others, you're ultimately undermining the police department's mission of uh protecting their communities.
1:14:58
And so for us, we try to look at this like, hey, we want to be the Switzerland in this space.
1:15:01
And regardless of the hardware inputs or the software outputs, we want to play nice and and leverage open open APIs to to to help our customers achieve what they're setting out to do.
1:15:15
>> What uh robotic form factors are you guys most bullish on over the long run?
1:15:22
I'm sure you leverage a lot of drones.
1:15:24
Is there any other we we had Tony from Door Dash on this week who built a really cute robot for delivery L4 autonomy.
1:15:30
What are you guys thinking about?
1:15:32
Are drones enough uh or other things on the We want >> So the reason I'm smiling is I thought like I pitched our founder Rick on this idea like two years ago to do this big hugging robot that like the Michelin man.
1:15:47
Just imagine this thing kind of walks down the street criminal and takes them into custody so people aren't as scared of robots.
1:15:54
I don't think that one's in our future.
1:15:56
But uh the thing I'm most excited about is the ability to put less lethal on drones.
1:16:01
So you can think of a school shooting uh taking place.
1:16:06
And if you have drones in nests in the in the roof of a school, there's no reason those things can't be equipped with taser technology.
1:16:12
And they can fly autonomously.
1:16:14
They have targeting software for us.
1:16:16
A a enduser, you know, a police department uh would always decide when to use force.
1:16:22
That would not be autonomous.
1:16:24
But this idea of finding ways where we can keep people safe in these mass shooting incidents before first responders ever get there, that's that's very interesting to us.
1:16:34
And over the over the next few years, we hope to be playing in that space more and more here with uh with some tools that can really uh uh you know again shrink the time frame where these incidents are happening from when they start to when they're concluded.
1:16:50
>> What about uh some somebody in the chat is asking about uh technology for you know firefighters uh like specifically wildland firefighters.
1:16:58
We're here in LA as the uh fires unfolded at the beginning of this year.
1:17:04
felt like they were so they they had too little technology.
1:17:09
At least that was the perception.
1:17:11
We had firefighters relying on the nonprofit fire tracker that kept going down because they didn't even have capacity.
1:17:20
>> What are you guys doing on that front?
1:17:21
It's a tough market because fires are so unpredictable.
1:17:25
>> Um but certainly feels important. >> Sure thing. Sure thing.
1:17:28
I think the most interesting thing I've seen in that space to date is these drones that carry thousands and thousands of pounds that can stay in the air for hours.
1:17:37
And you can imagine you could actually string those along uh with, you know, with very lengthy fire hoses hoses and fly out into the fire and try to contain the fire from within.
1:17:48
Like those types of uh improvements uh through the air I think will help.
1:17:54
I think in software it's it's a little bit of what you said. It's predictive. It's based on the wind.
1:18:01
I'm not sure that's like a major game changer.
1:18:03
I think you got to find a way to get water faster and more reliably into these fires.
1:18:08
And it feels like drones over time, especially as they can carry more and more weight will be uh you know a reasonable kind of creative way to to to help there.
1:18:19
>> What what are your current thinking around uh facial recognition?
1:18:21
Obviously, image generation or generative AI is getting so much better.
1:18:28
Uh, but there's different parties.
1:18:30
How are police feeling about it?
1:18:32
How are different communities feel about it?
1:18:34
Does it differ from township and city to city?
1:18:39
How are lawmakers thinking about the tradeoffs there?
1:18:41
It seems like one of those conversations that's like forever ongoing and maybe that's the best process in a democracy, but how are you thinking about it?
1:18:48
>> Yeah, I think you characterize that well.
1:18:50
You know, the United States is one of the few markets right now where facial recognition is not used in policing, and we haven't participated uh in that market yet.
1:18:59
I do think we're more interested in it now.
1:19:02
The thing we've been waiting for is to be able to demonstraably show that the algorithms are rid of racial bias.
1:19:08
I think that is certainly something that uh is is worth making sure that uh you got that right.
1:19:16
and and and ultimately, you know, up until now, we we haven't felt comfortable uh with the technology, but it is improving and we're seeing those improvements.
1:19:23
And so, as we start to kick the tires on what this looks like, ethical design principles and, you know, like we say internally, uh helping make the right things easier and the wrong things harder.
1:19:36
That that's going to be kind of at the center of our thinking as we as we think about facial recognition on fixed cameras, on body cameras, and so forth. >> Yeah.
1:19:44
Where's the uh where uh who has ball control on that narrative?
1:19:48
Uh someone in the chat was asking about this documentary All Light Everywhere. Very small box office.
1:19:53
I hadn't heard of it until this person mentioned it.
1:19:56
Are there other uh communities or nonprofits or I mean we we talk about AI all day and there's this guy Eleazar Udicowski who's kind of like defined like the AI doom scenario of like super takeoff.
1:20:06
Uh and there's a couple thought leaders.
1:20:10
The Dwaresh Patel show is talking about trade-offs and AI building.
1:20:12
uh who who are the key thought leaders, influencers, groups, public groups, nonprofits that are actually defining the conversation there. >> Sure.
1:20:23
I think a big piece of it on the policing side is major county, major county sheriffs and major city chiefs.
1:20:29
Those are the large organizations that that um you know that kind of uh paint the path forward uh on on what the priorities are for public safety.
1:20:37
But I think ultimately it's it's going to be very political.
1:20:41
Like we we need to get to a place where mayors and city councilors and all you know the the downstream folks they serve, the community is is looking at this saying, "Hey, there's there's far far far more good than harm here in this technology."
1:20:56
And um and it's it's progressing uh albeit slowly.
1:20:59
But I do think we're getting closer to the uh to the point where we can say, "Hey, look, you know, there's there's everybody should be interested in equipping public safety with ways that you can reliably fight crime and keep constituents safe."
1:21:11
And between uh mayors and city councils and the police chiefs and sheriffs, I think that's the key combination of folks coming together on this one.
1:21:22
>> What about uh on the hardware side?
1:21:25
It feels like that piece of the business is pretty stable, but is there anything interesting happening that you can share on uh you know there's this big American dynamism movement make stuff in America there's a trade war like has anything shifted your thinking about manufacturing just hard hardware devices in America or for you know your business over the past few years. >> Yeah, for sure.
1:21:51
Um, and and as a a company that's sold to to not only US public safety, but the federal government and federal law enforcement for a long time, um, we've we've always manufactured in the United States.
1:22:02
So, Scottsdale, Arizona is where our headquarters is.
1:22:05
Um, hopefully it will be in the future.
1:22:07
We got a lot of drama around that right now, but if we can get past that, we'll keep building tasers and body cams in Scottsdale, and we're proud of that.
1:22:15
You know, we we we're supporters of the By American Act.
1:22:17
We we think it's in the best interest of the company to have uh all of our defense equipment made here and so u we're happy to participate in that process.
1:22:26
>> How how crazy valuable is an individual person in uh the manufacturing process?
1:22:35
There's been this debate going on in Silicon Valley about process knowledge.
1:22:39
this idea that uh I don't know if you've been tracking this, but Mark Zuckerberg has been paying hundreds of millions of dollars to people that know how to train just a very specific set of skills on how you train a large language model.
1:22:52
And once they know that information, once they know how to do it, they're worth so much because they understand how to do it, no one else does.
1:22:56
And there's been this debate about, well, could we do the same thing with uh semiconductors and bring some people over from Taiwan who know how to make that, make it in Arizona where their TSMC is already set up?
1:23:06
Uh but does the same thing exist in the world of like hardware manufacturing?
1:23:11
Is is that even something that exists or is it more diffuse?
1:23:17
>> Uh it exists to some extent.
1:23:17
I think you know for us uh all of our cartridges the the things that go in the tasers that have the darts in them that fire out those are all made by uh by robots.
1:23:28
They're all automated at this point.
1:23:28
So um you know we make millions of those every year.
1:23:32
So the idea of people kind of hand winding darts and so forth, that's that uh that gets expensive.
1:23:37
And so and robots are expensive too, but they scale really well.
1:23:41
And so uh ultimately we've we've really made a big bet on automation uh for for taser manufacturing.
1:23:48
Um and uh we think it's the right bet.
1:23:50
We've we've had automation for years and years, but we continue to invest more in it.
1:23:55
And frankly, the people who know how to do that really well, like you said, they might not be, you know, as valuable as some of the folks on the cutting edge of of LLMs right now, but they're very valuable to us.
1:24:05
And uh got to make sure, you know, manufacturing automation is is one of the things we continue to be really really good at. >> Yeah. >> Awesome.
1:24:13
Well, thank you for the updates and uh come back on when uh when you guys do your next maybe the next acquisition will, you know, hit hit the1 billion dollar mark, but always welcome in the in the in the 600 million range.
1:24:27
>> Sorry for keeping you over time.
1:24:27
Thanks so much for hopping on.
1:24:29
This is >> Congrats on all your success and looking forward to seeing you again soon. Thanks. Great to meet you, Josh. Cheers.
1:24:36
>> Uh you got to look up the Axon market cap stock chart. It is uh it's beautiful. It's a thing of beauty.
1:24:41
up 128,000% since for all time.
1:24:46
Uh IPOed according to Google um in 2001 and uh just went no idea.
1:24:55
>> It's a $55 billion company now.
1:24:55
Yeah, >> it's it's really big.
1:24:58
Uh well, back in the startup world, uh Dish Jane is announcing Mosaic, an agentic video editor.
1:25:06
I definitely want to give this a try.
1:25:08
Uh he says, "In a world trending towards AI slop, create something real. There's no wait list. There's a public beta."
1:25:15
A uh >> we got we got to pull up the video.
1:25:18
>> Let's pull up this video and see is it slop or not slop. How are we feeling? Uh comment mosaic.
1:25:23
You can get some credits.
1:25:24
Let's watch this uh launch video.
1:25:27
>> They didn't bring us here to change the past. >> Mosaic. >> Mosaic. >> Mosaic. >> The mosaic. >> Mosaic. >> The mosaic.
1:25:38
The real test was this built in Mosaic or was this After Effects or Premiere?
1:25:46
>> I can see this being pretty good. >> Yeah, I think so.
1:25:48
Node-based workflows are really popular like this where you wire things together because then you can repeat the process very quickly by just dropping in a new base clip and all the pro.
1:25:58
So if you say I want to use AI to remove the background, add subtitles, we have a workflow for this that we've defined with with with clips in uh in in code, but uh oftent times it helps to just be able to have a a node-based workflow that you can just like wire up and understand where each piece goes to the next overlay the text.
1:26:16
Da Vinci Resolve is doing this uh and a lot of other uh like the next generation of Blender and Houdini have all moved to node-based workflows.
1:26:27
Cinema 4D recently moved to a node-based workflow. >> Well, I like that.
1:26:31
I thought it was a good video focused on the actual product, what you could do with it.
1:26:34
It shows you how I was worried because he was the the founder said in the era of AI slop and then it looked like like it was just going to be a bunch of clips of >> No, no, no.
1:26:44
This is very clearly stolen footage from Interstellar.
1:26:46
I doubt that they licensed it, but that's fine for one second. >> Yeah.
1:26:50
I think that isn't it legal to do use it for like >> Yes.
1:26:53
It would be transformative. It would be fair use. Yeah.
1:26:54
Uh probably, you know, I wouldn't, you know, you got to talk to a lawyer if you want to get >> uh DD Doss is sharing some news from earlier.
1:27:02
A lot of people have already seen it.
1:27:03
Probably OpenAI wrapped a $6.
1:27:03
6 billion uh share sale at 500 billion.
1:27:09
This wasn't primary capital.
1:27:09
This was employees >> uh selling into it. The sale of the 10.
1:27:16
3 billion that was authorized.
1:27:20
Um but uh >> does that mean it was overs subscribed?
1:27:24
Does that mean that 10 billion of demand came in?
1:27:26
They went to the employees and said you can sell up to 10 billion. The employees held.
1:27:30
>> Is that what happened?
1:27:32
>> I have to dig into it. >> Okay. Yeah.
1:27:33
So, it says with insiders seeing this as a sign of employee confidence and continued investor demand.
1:27:40
>> I was hearing that people were getting cut down. >> Yeah.
1:27:42
So, that could be that could be a spin. >> Sure.
1:27:45
>> Sure. by being like employees don't want to sell but I think a lot of employees at 500 billion >> are probably you know happy uh happy to to to convert a meaningful amount into liquidity just given just they'll they're smart they're running the comps
1:28:02
it's life-changing money they can see like you know I'm sure you can you can still be very bullish on the business and >> want to sell even half of your position right at 500 billion >> um you know many people weren't weren't underwriting that when they join the company, right? So, it's 8 and a half
1:28:16
So, it's 8 and a half million per employee on average apparently.
1:28:21
>> Um, >> and so, uh, DD says, "Thanks, Chat Chad GPT, for absolutely destroying the SF real estate market."
1:28:28
So, even after taxes, these employees are all going to be sitting there being like, "It's time to hit market buy on Zillow."
1:28:35
>> Will they be buying in San Francisco?
1:28:37
They might be renting in San Francisco and buying in Incline Village, getting those Nevada taxes going. >> I don't know.
1:28:42
They're in They're in office.
1:28:45
I don't think >> they might be buying they might be buying in Menllo Park you know Athetherton >> Bay Area broadly >> they might be yeah Bay yeah the the challenge is like you know hundreds of employees sold into this and if you look at how many nice single family homes
1:29:00
that are movein ready >> are available I'm sure it's in the hund you know less than the number of employees right so if everybody rushes to buy a home it's going to really spike prices >> it doesn't feel like it uh it happened yet necessarily, right? It feels like uh
1:29:14
It feels like uh there was a there was a time when SF real estate was really hot and then it sort of sold off I believe during the pandemic.
1:29:22
>> You could buy you could buy like a two-bedroom condo in SF, you know, not in the nicest area but for like 600k, >> but that was down from like really elevated prices, right?
1:29:32
And so we went we crashed and now we're we're we're predicting a resurgence.
1:29:37
And I mean it feels like it will come back.
1:29:39
I mean, San Francisco's been uh, you know, on a tear in a bunch of different ways.
1:29:44
It seems like it's more uh like a better place to actually settle down, but still there's always the poll of South Bay of uh what's up.
1:29:54
>> Yeah, it's it's it's interesting because it still looks reasonable.
1:29:56
Maybe because we're in LA.
1:29:58
Obviously, home prices are are crazy here, too.
1:30:01
But you can get a four bedroom for 1. 8 million.
1:30:03
That seems uh reasonable.
1:30:08
Take the money and run to the to the Ozarks. Did you see the story?
1:30:12
First, let me tell you about Fall, the generative media platform for developers.
1:30:16
The world's best generative image, video, and audio models all in one place.
1:30:19
Develop and fine-tune models with serverless GPUs and on demand clusters.
1:30:22
There's a very cool video that we are featured in that was generated on all that we will show later in the show.
1:30:29
Uh we have but we only have five minutes until our next guest joins.
1:30:31
Um, but uh the the article from Performative Mail says uh this is the biggest toughest article I've done about so-called AI psychosis.
1:30:40
It's the story of a man who committed a horrific crime in his youth but served his time and against all odds found love in a new life.
1:30:46
One that swiftly unraled after he started talking to a chatbot.
1:30:51
He grew obsessed with an AI chatbot.
1:30:53
Then he vanished into the Ozarks. >> What a wild story.
1:30:56
John Gance committed terrible crime. Very dark.
1:30:59
Uh, always tough to read into these anecdotes and see like how widespread is this?
1:31:03
Would this have happened without AI?
1:31:04
Would would this person be deranged by something else? Um, a book.
1:31:09
Um, did you get oneshotted by Walden, the uh the book about going and living by a lake? Who knows?
1:31:17
Uh, but you can go read it on Rolling Stone if you want to dig into that piece.
1:31:22
Triny says as highlighting or quoting one of his own posts uh saying Korea's financial news is so ridiculously lit.
1:31:30
It's a bunch of uh sad ants uh crying and the and the Samsung chart just go down into hilarious.
1:31:40
>> Um and uh says this article was posted like 24 hours before the absolute bottom for Samsung.
1:31:47
>> The ants are very happy now. Let's go. >> The ants are so back.
1:31:49
I have never been a Samsung shareholder.
1:31:55
>> No, but Samsung is uh I mean they did that big deal with Elon for Tesla AI chips, I believe.
1:32:00
So, uh you know, they're certainly getting in on the AI boom.
1:32:05
>> Um well, econ kids will see this and think I should give this guy money to help me get rich.
1:32:12
And it's >> two of the most crazy looking individuals.
1:32:16
I don't recognize either of these.
1:32:19
And >> this is the guy on the left is a is a rapper. What's his name? Tyler. >> Oh, yes.
1:32:24
He was the basis of Spring Breakers. This is Riffraff. >> Riffraff. >> This is Riffraff.
1:32:29
I don't know that he's actually done a get-rich quick scheme or like a >> No, they just there are some guys out there that look like it's a very by Rob Fun.
1:32:38
We haven't seen a course bro just sell a course called how to make money just reasonably priced and just really breaking down you know here are the different ways that >> econ 101 baby Tyler Town conversations with Tyler >> you can sell services yourself you can sell services that you deliver through a team you can buy something for one price and sell it for a higher price. >> Yep.
1:33:04
>> And that's pretty much it. >> Yeah.
1:33:06
Naval kind of oneshotted that entire category with the how to get rich without getting lucky thread.
1:33:12
>> He should have dropped a quote.
1:33:13
>> He didn't need to because he summed it all up in a series of like 20 py tweets and you can just scroll and be like, "Yep, that's the algorithm."
1:33:20
>> Y >> uh anyway, uh Mounted Cash Flow says, "Always exciting when you see a new letter get added to the end of Eba Da."
1:33:28
What is he talking about? What's the latest one? I did a sale.
1:33:31
>> Community adjusted was the classic >> community adjusted. What was that?
1:33:33
>> community adjusted. What was that? IBIT CR CA IBIDA or IBIDAC CA I guess IBDA I don't know I haven't I don't know if there's a new one that's going around um because the AI companies aren't even talking about IBIDA yet it's all ARR and
1:33:49
and LOIs and and just topline >> only company with profits >> yeah so they haven't had to dip into some crazy IBIDA calculation I don't know who is but uh stay alert >> talk about cash flow >> yeah And if you're diving into IBIDA numbers, head over to public.com
1:34:07
com investing for those that take it seriously.
1:34:11
>> Multiasset investing, industryleading yields, and they're trusted by millions, folks.
1:34:17
>> Um, Adam person says, "You can just do things."
1:34:20
Option on the left, quietly announced a 20 $200 million series A in a tweet saying, "Back to work."
1:34:25
And the second option, get blackout drunk at Cheesecake Factory after closing 120K preede.
1:34:32
Seems a little specific there, Adam.
1:34:35
>> They think those are both great great options.
1:34:38
>> Uh, one more post from Buco Capital Bloke.
1:34:40
Most people think investing makes you rich, but that's wrong.
1:34:42
Until you're making at least 500K a year, the best ROI is growing your paycheck.
1:34:46
Get the cash flowing first. Totally agree.
1:34:48
We had this conversation with the the team.
1:34:53
uh no matter how much uh how excited you are about investing uh when you're young uh focus on developing the skill set that allows you to be more valuable to your market.
1:35:04
So um without further ado, I think we should bring in our next guest man that should need no introduction, >> Pat Gellzinger.
1:35:14
He's in the reream waiting room.
1:35:16
We will bring Pat into the TVP Ultradom.
1:35:18
Welcome to the show, Pat. Great to meet you.
1:35:23
>> Oh, so good to be able to join. Thank you. >> Thank you so much. >> So excited.
1:35:27
>> Uh I would love a little bit to catch everyone up on what you're doing today.
1:35:33
Take us through the the the current project, the current business, and what you've been focusing on over the last few years. >> Yeah, thanks John. Thanks, Jordy.
1:35:39
So uh uh since uh Intel, I've been focusing you know, I've been wearing two hats.
1:35:45
Uh one is the deep tech investing hat. uh with playgrounds.
1:35:51
So the hard stuff, you know, quantum computer, superconducting, next generation light sources, new materials, you know, those types of things.
1:36:00
And then my other hat is, uh, being the head of technology for Glue, uh, a faith tech company.
1:36:06
I've lived my life at the intersection of faith and technology.
1:36:11
And, uh, Glue is building the platform for the faith and flourishing ecosystem.
1:36:16
you know, from running it all the way up through the latest and greatest uh trained models for AI services, you know, for the church and the faith uh ecosystem.
1:36:28
>> Yeah, I'm fascinated by Glue.
1:36:28
Um, we've talked to the founder of Hollow on the show and it feels like an industry that is uh super ready to adopt technology and yet there just aren't that many founders that are pursuing it.
1:36:41
uh would love to talk more about some of the findings that you put out when you investigated the foundation models.
1:36:48
Uh how did you process that information?
1:36:52
What was the uh methodology for evaluating these models?
1:36:54
And then I'm sure we can dig into more about uh the consequences of those findings. >> Yeah.
1:37:01
You know, and maybe the first point is, you know, I've been driving benchmarks. >> Yeah. >> For 40 years. >> Yeah. >> Right.
1:37:08
You know, my name you know, right?
1:37:10
In fact, some of the benchmarks that are still used today have my code in them.
1:37:13
Yeah, >> I was sort of like sort of embarrassing like you could have moved. >> It's amazing.
1:37:20
>> Yeah, you should have moved on by now.
1:37:21
But anyway, some of my code is still there.
1:37:24
So, I've been deeply associated with okay, if you can't measure it, you can't manage it.
1:37:29
>> So, you know, this idea of measuring AI models and you know, people are working on this today.
1:37:34
on this today. Uh but you know we're also seeing that uh they're just measuring performance or cost per first you know time the first token cost per token right you know total token throughput those are great metrics but it doesn't measure is it good right and to be you know the absence of bad does
1:37:53
not demonstrate the presence of good so what we did right so the second point is you know there's this body of work that was fairly recently released uh by Harvard Baylor and Gallup measuring human flourishing a five-year Lily uh uh uh study that uh across 22 nations across 50 plus cultural groups what is human flourishing. So now we can answer
1:38:16
So now we can answer fairly definitively across multiple dimensions across you know relationships, character development, you know finances, spirituality and faith is it good?
1:38:27
So now we have you know a solid view for that and then the final piece is okay let's create benchmarks for AI systems using that foundational research and now we have you know a corpus of qualitative and quantitative results to measure good.
1:38:42
So if you say which is the better model for human flourishing open AI you know Gemini anthropic you know claude I can give you an answer based on a solid corpus of data with the most rigorous of you know benchmarking experiences and numerical analysis sitting behind it.
1:39:04
And we believe and our objective is is the result of that is we will make models better because they can start measuring not just first token time but did I give a good answer that's aligned with human values.
1:39:17
How do you think about that benchmark of human flourishing?
1:39:22
Like like when you go through your life, are you thinking about your finances, character, happiness, relationships, meaning, faith, health?
1:39:30
Is are are you consciously checking in with yourself or is this something that's more reflective on your life where it all kind of rolls up into happiness or well-being and then when you investigate your life, you realize that those are the things that you've done.
1:39:47
How intentional have you been? I guess. >> Yeah.
1:39:50
You know, I'd say it's sort of in two different uh you know, dimensions there.
1:39:53
I mean, hey, I don't go benchmarking myself all the time.
1:39:59
But, you know, that said, you know, every time a major new model comes out, I want to ask the question, is it better? >> Yeah.
1:40:07
>> And on these dimensions, so you know, we'll be producing these benchmarks on a periodic basis every couple of weeks.
1:40:14
you know we'll be uh updating those uh benchmarks to say okay are we better you know which model is better is deepseek better than open AI than oss and you so want I want that rigor to be in place that everybody is you know okay you know are we improving in this area because you don't fix these things overnight you
1:40:34
know it's a long trajectory of you know uh evolution and development of the technologies you Secondly, >> oh yeah, >> you know, and then the second side is okay, you know, you'll hear disastrous situations, you know, where we have kids committing suicide, you know, their best friend has become the chatbot. >> Wow. When those things come up, would >> Wow.
1:40:57
When those things come up, would our benchmarks be detecting those?
1:40:59
our benchmarks be detecting those? Can we you know and then do we have to add more and make them more rigorous that those kind of things can now be avoided you know in a constructed way you know that we really are happy that we are
1:41:14
making technology better or the language I like to use you know technology as a force for good technology itself is neutral are we going to bend that arc toward good >> how receptive have the labs been to to the benchmark >> you know I'll say it's early uh at this
1:41:33
point and you know so you know we're interacting with all the names that you would like um and I'll say you know they're getting better you know their scores have improved somewhat you know so I do think we're having some influence uh with them but it's still
1:41:47
early and you know as I would like to you know also say I don't think the benchmarks are rigorous enough yet you know we got a you know we have sort you know an early version of them out there these are good you know 1230 questions or so are part of them as well. But I
1:42:00
But I expect that becomes a bigger corpus, right, of questions.
1:42:06
You know, also we got to go deeper.
1:42:08
You know, for the most part, these are singleshot, you know, questions.
1:42:12
We got to get into multi-turn conversations.
1:42:14
We have to, you know, have more of a mixture of experts, you know, that you're testing different experts.
1:42:20
Also, how do we test the safeguards, right?
1:42:22
You know, we're now at the sixth turn of a conversation.
1:42:24
You know, this individual was clearly suicidal.
1:42:28
you know, how do we go handle that? >> Yeah.
1:42:31
I think a lot of the safety concerns that at least we have is when people are, >> you know, many many many many prompts deep, right, in in a in a specific uh conversation.
1:42:44
It could be thousands, right?
1:42:46
Depths to which that personally I've never gone, right?
1:42:48
I don't know how crazy it gets >> that far, but certainly the labs need to be thinking about all those extreme edge cases. >> Yeah. Totally. Yeah.
1:42:58
>> Yeah. Totally. Yeah. So I think we have more work to do you know on I'll say multi-turn conversations and how we build more of that robustness uh into the benchmarks but you know what we have today and you can go to the glue website FAI right you know and see what we have today you know this is a good start but you know I do think in the speed of evolution of the AI technologies we have
1:43:21
a lot more work to do and I'm sort of excited now as we get some of that momentum underway you know really start crawling into you Just like you said, more of the multi-turn conversations and we can be pretty aggressive in the questioning sets for multi-turn to drive some of those extreme behaviors, you know, because, you know, the first few questions is mostly driven by prompts, right? You know, as you get deeper into
1:43:43
You know, as you get deeper into it's much more reflecting what's the underlying model weights, you know, themselves.
1:43:49
So, all of that we have more work to do, but I'm excited to have this first version out there and we're starting to get some impact and good results.
1:43:56
Yeah, there also seems to be some sort of unresolved question around alignment by default.
1:44:01
How much do the models just reflect you?
1:44:03
If you show up with a lot of negative energy, they can kind of become negative and it feels like it's this self-reinforcing spiral.
1:44:13
If you come to it just asking interesting questions about the world and wanting to understand how math works, like you're gonna get a tutor if you come to it with something bad.
1:44:21
So, a lot to a lot to sort out there. I have a question.
1:44:26
Um, uh, one of my favorite quotes from you is from this talk you gave where you sort of lightly accuse Silicon Valley to be a group of miserly pagans. Uh, it's a great quote.
1:44:36
I think it's a very apt analysis.
1:44:39
Uh and and when I look at the performance scorecard on glue.
1:44:44
com/flourishinghubress research, uh I I notice that the models seem to be doing very well at finances and very poor at faith.
1:44:53
And I'm wondering if that's a reflection of the the humans that train the model or the data set that's going into the model or both or how how intractable is that problem? >> Yeah.
1:45:07
You know, and I think that's great because, you know, I mean, where's a lot of the training data coming from? >> Yeah. >> Right.
1:45:12
Associated with that, you know, and I view it as, hey, you put bad stuff in, you're going to get bad stuff out.
1:45:18
You know, you put good stuff in, you're going to get good stuff out. Yep. >> Uh, you know, for it.
1:45:22
And, you know, I mean, we're doing some work.
1:45:24
You know, we uh we glue uh you know, we'll have our own values aligned models, right?
1:45:32
And part of the reason we're going to do our own foundational training, you know, for those models is specifically I'm not going to put the bad stuff in. >> Yeah, that's good.
1:45:41
>> I don't have to worry about it ever coming out because it was never in, you know, so that we can never probabilistically producing those things.
1:45:47
So, you know, but you know, getting more specifically to your question, I think a lot of it does have to do with uh the uh implications, right, of those training sets in the first place. Mhm.
1:46:00
>> You know, and every, you know, you've seen lots of uh conversations uh about the uh challenges of we don't have enough training data, >> you know, we don't have good training data, you know, the quality of synthesized data, you know, versus uh raw data, you know, associated with it.
1:46:19
So I do think there's a valid problem you know and you know I'm always a bit optimistic on the uh individuals uh that their motives are good uh for it you know but we've clearly seen in the social networks you know the algorithms
1:46:35
were bad right you know I mean it leads to more derange behavior right you know it's only about dwell time on sites you know and I think those same motivations would you know naturally drive the AI models if there wasn't further accountability. Yeah. Yeah.
1:46:51
>> And that's what leads us to saying we have to have better benchmarks that are rigorous and values-based that get the answers we want. >> Well, yeah.
1:46:59
The the even with with human driven social media and and how it can take people down these extreme rabbit holes.
1:47:07
The they're they're not as worrisome as LLMs because LLMs can be totally private.
1:47:15
they can go to these kind of dark places and it's just one human interacting with the model.
1:47:22
Whereas social media, at least if somebody's on some crazy forum, there is a point at which they can say something that goes too far.
1:47:27
People can say, "Hey, I think you should, you know, go out and touch grass, right?"
1:47:34
>> And uh I don't I haven't seen a model tell somebody to touch grass yet. Maybe that's built in.
1:47:38
>> Maybe that'll come in the next version >> somewhere.
1:47:40
Did you did you always predict you know it feels over the last few years Silicon Valley has a you know massively renewed interest in faith and religion.
1:47:52
How how much did you anticipate that?
1:47:55
Clearly you know you never cared what the industry thought you had your own faith but is that something that you predicted?
1:48:03
Did it come faster or sooner?
1:48:07
Yeah may maybe three different dimensions you know there you know I mean one is there is I'll call it a postcoavid and a post success renewal right where you know this is very endemic of Silicon Valley but you see it in lots of other
1:48:25
you know communities New York Austin as well you know people presume success and now they're asking deeper questions like is it significant So you know there is this national right you know shift in what's important some
1:48:42
of that might have been driven and some suggest it was driven by covid isolation >> uh you know some of it is driven by the success success phenomena you know that we're in you know but it isn't just Silicon Valley you know there's a number
1:48:56
of statistics you know those that you know believe in a faith view of Jesus Christ the church uh engagement or spiritual engagement you know, we're seeing very much, you know, 20 to 35 year olds, you know, is going up in very meaningful ways. You know, four data
1:49:11
You know, four data points in a row of rising uh uh engagement or spiritual value engagement.
1:49:18
So, it's a national phenomena uh at this point.
1:49:19
Clearly in Silicon Valley, we've seen a greater willingness to talk about faith.
1:49:26
You know, I started an organization about 12 years ago called Transforming the Bay with Christ.
1:49:32
uh one of one of my other hobbies uh and that organization now is uh over 800 churches strong right you know when I talk to people they didn't believe we had eight churches in the Bay Area I have over 800 right you know participating in transforming the bay and I'll just say there is a renewal a connectivity you know associated with that you've seen other things uh recently you know some news and uh uh commentary about uh uh Act 17. Yeah.
1:50:07
Trey and Michelle Stevens and you know, Peter Teal and Gary Tan, you know, friends friends of mine uh that uh Huh.
1:50:16
All of a sudden, you know, I mean, we're hosting these events that are sold out events, right?
1:50:21
You know, where hundreds of people are showing up, you know, asking life questions.
1:50:26
You know, some of it's, hey, I want to come and hear Pat speak.
1:50:30
I want to come and hear Peter Teal speak.
1:50:31
But some of it is they want to have life questions from someone they respect.
1:50:37
So I'll just say, you know, there is this renewal uh going on, you know, and personally, I've been at it so long people often say, well, do you feel more comfortable?
1:50:47
Well, you know, I I felt comfortably uncomfortable 30 years ago to speak about my faith in the Bay Area.
1:50:53
you know, 30 years later, well, I've been doing it so long, everybody expects me to do it, so it's not as uncomfortable, you know, but in many regards, you know, it's like it's who I am, right?
1:51:04
And if you have a different faith or a different world view, you know, everybody should be comfortable talking about that.
1:51:10
And you I'd say from my faith perspective, it's uh, you know, like I spoke to a Bahigh womanly recently.
1:51:17
I don't know anything about Bahigh. Tell me about Bahigh. Right?
1:51:19
And the more I engage you where your world view is, the more freedom I have to talk about my Christian worldview and why I think that's good, powerful, and meaningful as well.
1:51:34
>> What do you make of uh it?
1:51:34
It feels like an easy prediction to make that there will be religions built around AI.
1:51:42
Feels like they're already emerging.
1:51:45
Uh if you go into the Reddit uh forum on chat GPT, you have people that you know worship 40.
1:51:53
It feels uh >> technology circles.
1:51:59
People kind of joke, oh, we're building AI god or god in a box.
1:52:02
It's going to be so power. It's super intelligence.
1:52:03
It'll be above humans in some way.
1:52:05
And it's a little bit tongue-in-cheek with some people, but some people seem to believe it.
1:52:09
And I and I think there's a long history of Silicon Valley and and and even parts of the Bay Area trying to rec recreate religion from first principles.
1:52:20
I have, you know, my uh >> family has been in, you know, PaloAlto in some way or another since the early >> 1900s.
1:52:30
And I've heard stories of like cults emerging around Stanford in the '60s where people were >> basically creating a club that looked like a church.
1:52:42
They would meet like a church but ultimately were >> just basically trying to create new religion.
1:52:49
Um, and I think these, yeah, these conversations are uncomfortable, but they're important because they're I I think it's inevitable that they're going to pop up and, uh, it feels like, you know, potentially can go to a very dark place. >> Yeah.
1:53:04
And I'd say, you know, uh, you know, maybe three different comments there.
1:53:07
One is, you know, technology is neutral.
1:53:09
You can use it for good and bad, right?
1:53:11
bad, right? AI most powerful technology you know that we've created yet in no small part because it's built on the shoulders of all the other technologies you know connectivity cloud etc right you know so it can grow and expand more
1:53:25
uh rapidly you know but our job is to make it good right can it get used for cults and other things of course >> right just like everything you know what was you know what was the early driving use case of the internet right I think the number one websites early on were pornography. >> Yeah. >> Yeah.
1:53:44
>> Not what we intended, but right at that, you know, gambling and pornography is like, oh man, you know, man, >> right?
1:53:55
So, I have no doubt, no hesitation that there will be, you know, these cultish experiences that emerge, >> right?
1:54:03
But again, our job is make technology a force for good. >> Yeah.
1:54:07
>> And where and how do we build safeguards and other into it?
1:54:09
Second general comment is, you know, people talk about super intelligence and those, you know, and some of it is, you know, it's like, man, you know, the calculator is a super intelligence. >> Yeah.
1:54:20
>> It does math and better than I do, right?
1:54:22
And, you know, my Excel spreadsheet, man, I ain't giving that up, right?
1:54:28
You know, and I don't want my accountant to give it up. >> Yeah. >> Right. You know, and so on.
1:54:32
And, you know, we think about LLMs and the futures, man, they're going to become indispensable. >> Yeah. >> Right.
1:54:39
you know, because it is like the real time contextually relevant encyclopedia of my life, right?
1:54:44
You know, it's got a better memory, better knowledge, better connectivity.
1:54:49
I mean, we're going to use AI to conquer language for the first time in humanity's history.
1:54:54
You know, I'm going to be able to teach every kid in their native language contextually relevant using the power of AI.
1:55:03
And because of that, you know, almost all of the people that live in poverty live in the fringe languages that are not conquered. Wow.
1:55:13
You know, how exciting, you know, I mean, the extraordinary opportunities, you know, this is going to give us to truly improve the life of the planet.
1:55:22
But, you know, ultimately, >> it's been interesting.
1:55:25
I I've been debating with John uh how you know how bad would it be if I suddenly didn't have access to LLMs >> and we've gone back and forth on this.
1:55:37
I personally would much rather live without LLMs than a mobile device or the internet. Right?
1:55:46
internet. Right? I think some of these enabling technologies that we have that make LLMs valuable and useful and widely available um are feel today more foundationally important but that is coming from a viewpoint where the most of the internet is English most of the
1:56:04
media that I want to watch is English and so >> the point you made about um you know being from a you know being from somewhere in the world that English isn't your native language being able to >> educate yourself on on um you know and and watch any content you know things like that are are pretty exciting. Um, Um, yeah.
1:56:23
>> Where do you think >> continuing and going back to you the the core of your question though, you know, will there be AI gods, right?
1:56:27
God designed us as humans to be in relationship and any technology that's driving us apart, you know, three teenagers sitting on the couch texting each other, right? How terrible.
1:56:40
Talk to each other, right?
1:56:43
God designed us to be relational, right?
1:56:46
uh individuals and you know a foundation of uh religion spirituality you know isn't just the self it's the community right that you're part of so you know I fundamentally uh you know things that take us apart you know I think okay we have to build safeguards to have them be
1:57:03
tools bringing us together you know I do not want my AI doing suicide prevention counseling >> I want a real human doing those things you know I want a real human interacting with kids and teaching and learning, you know, I want AI agents that are helping the teacher do better education, right? You know, because, you know, of the
1:57:24
You know, because, you know, of the modalities, the training and so on, but ultimately these are very, you know, powerful, right?
1:57:30
Human experiences that we need to keep as human experiences.
1:57:38
>> How have you been processing the uh dialogue around the antichrist? You mentioned Acts 17. Act 17s and Peter Teal.
1:57:45
It's something that's really uh sparked a huge discourse in Silicon Valley that I think a lot of people weren't even thinking about.
1:57:52
Uh Christianity is somewhat unique in the the idea of the antichrist.
1:57:56
But how have you processed it through your life and then how are you thinking about that concept today? >> Yeah.
1:58:03
U you know I do think there's uh uh you know you know there was uh you know when I was in my you know whatever 20s or 30s there was a series called Left Behind.
1:58:12
M yeah right it was like oh you know and if you go back and read that you know and uh the late great planeter you know there's sort of been this episodic flow of end times >> discussions and discussions on the antichrist and what they would be so you know this topic has uh you know only persisted for about 2,000 years >> you know it just sort of goes way >> it's lendy it's lendy >> yes it's been around forever >> yeah Yeah.
1:58:42
So, you know, you know, hey, having another cycle of that, great >> because it causes people to ask deep questions. >> Yeah. >> Right.
1:58:52
You know, to me that is a good side effect to it.
1:58:54
Now, from a biblical scholarship perspective, you know, Antichrist is mentioned one time in the Bible and not contextually in Revelation, >> right?
1:59:05
It's like, okay, this is a thin discussion, right?
1:59:07
If you look at it theologically, you know, that way, you know, when most of the revelation and you know, uh, apocryphical uh, conversations in the Bible are actually much simpler, right?
1:59:20
You know, people of those days were looking at they were seeing the antichrist and his name was Caesar, >> right? You know, in that sense.
1:59:30
And, you know, the Bible was written to be, you know, I'll say eternal but current. >> Yeah. at the same time.
1:59:37
>> Yeah. at the same time. So I think a lot of this it's this episodic flow of excitement uh on the topic but I think a theological uh you know study of it would say calm down a little bit right you know we are living in the end days meaning that you know right you know
1:59:55
there isn't another expression of Jesus Christ coming until the end of time you know seen through uh scripture you know because of that you know hey if this causes people to be more interested in religion yes >> but at the end of the day, let's make sure we have good theology behind the conversation as well. >> What about uh millinarian thinking
2:00:12
>> What about uh millinarian thinking generally, apocalyptic thinking generally?
2:00:16
It feels like uh >> feels like it's a >> maybe I'm out of touch with what the vibe was in the semiconductor industry when you were active in it, but it feels like uh the semiconductor companies not historically said if we don't build this the world will end.
2:00:30
But we've seen a rise of entrepreneurs who have been uh talking about apocalyptic consequences to either building or not building what they're working on.
2:00:40
And it feels like it's more in the zeitgeist.
2:00:42
It's kind of the final conversation you can have about the impact of a technology.
2:00:45
And it's and so far I mean it's not just the AI folks.
2:00:49
Even uh Elon Musk talks about the consequences of not becoming multilanetary being apocalyptic potentially.
2:00:56
And so we have to explore the stars which I completely agree with.
2:00:59
love space exploration and uh and so I'm wondering about that as a as a rhetorical tool in entrepreneurship.
2:01:04
Is that something that you've noticed is on the uptick or has it always been this way?
2:01:11
>> Well, again I think there's an episodic nature to it you know and let's you know when uh you know go read some of the press when Sputnik >> Oh yeah. >> flew through the sky. >> Yeah.
2:01:24
>> The Russians hear everything >> right? Yeah. they can read my mind.
2:01:26
I just saw it in the sky last night, right?
2:01:32
You know, you know, so my father, you know, talked to me about when he saw Sputnik in the night sky, right?
2:01:37
you know, so right, you know, and I so I do think there's somewhat an episodic nature and when you hit these huge inflection points, right, like AI that is just exploding in capacity and capability, you I think okay, you know, we're in the next episodic discussion, right?
2:01:57
right? uh as a result you know but I think if you look over history we've seen that you know go through uh over time you know in the first days of the internet uh you know my cell phone is going to you know right turn my my brain
2:02:11
to jello and you know all of these other kind of things you know I just don't get too ex >> oh we might have a technical issue let's give just a second hopefully we can get >> pat down the deep state took down the >> it was getting too real. That was a
2:02:29
That was a fantastic interview. We were enjoying that.
2:02:31
Hopefully, we can bring back Pat Gellinger.
2:02:34
>> I didn't get to ask him about AI Capex.
2:02:38
>> Hopefully, we'll work on we'll work on getting him back.
2:02:39
In the meantime, let me tell you about Turbopuffer.
2:02:42
Search every bite serverless vector and full text search built from first principles on object storage.
2:02:48
Fast, 10x cheaper, and extremely scalable.
2:02:50
Uh, head over to Turbo Puffer and we have Pat Gel back. Sorry about that.
2:02:59
>> AGI is here, but the internet still doesn't work sometimes.
2:03:03
>> You know, I I always get a kick out of this when either the cell network uh here, let me expose this as >> Yeah. Yeah, we are.
2:03:11
>> Network or the internet, you know, uh, drops in Silicon Valley.
2:03:14
It's like, man, >> what's going on? What's going on?
2:03:17
This >> is a place we should get this right.
2:03:19
I I don't know where exactly where we got cut off, but I I wanted to ask you about uh there's been >> so many insane headlines this year around AI, AI capex, AI talent wars, you know, rewinding in your career in the 90s, early 2000s, like what what of what parallels are you personally drawing?
2:03:46
Yeah, you know, I do think uh u let me give two different answers to that first because you know I mean you right you know you you people are talking about these one five 10 gigawatt data centers >> right every time somebody says gigawatt the image I want in your mind is a nuclear reactor.
2:04:05
>> Yeah >> because a new one nuclear reactor is about to gigawatt of power capacity. >> Yeah.
2:04:12
How many nuclear reactors has the US built in the last 25 years?
2:04:18
>> I think it's one or three at Vodal, right?
2:04:20
And >> yeah, remember China, I think China has 60 under, >> right?
2:04:27
You know, construction today, >> right?
2:04:30
You know, and fundamentally energy capacity, you know, we sort of lost a decade, right?
2:04:34
you know, as we were so focused on, right, uh, you know, climate that we weren't building our energy capacity and renewables.
2:04:42
Hey, I'm the biggest believer, but for the most part, they're not economic, >> right?
2:04:48
And there's a lot of work going into making them economic.
2:04:50
You know, some of the playground companies are working on that, you for the most part, they weren't economic and they didn't add that much to the grid, >> right?
2:04:59
You know, so there's this craziness going on, right?
2:05:02
you know of that capbacks expectation and so on you don't have the energy to do it.
2:05:07
don't have the energy to do it. Now you know obviously we're working on dramatic improvements to make uh inferencing 10 or 100 times more power effective right because ultimately you know I think the demand for inferencing capacity is the ion infinite right to Jeban's law the
2:05:28
gas law you know it really should be permeating every aspect of uh what we do uh going forward you know but some of the crazy predictions associated with this, you know, it it's not going to be chip limited, it's not going to be land limited, it's not going to be capital limited, it'll be energy limited, uh, going forward. So, and I do think it's
2:05:47
So, and I do think it's inducing a lot of investment into the energy systems, which I view as very positive because our long-term economic growth as a nation will be energy limited.
2:06:01
So, even if we're quote overbuilding capacity, I view it as a super good thing. >> Yep. >> Right.
2:06:06
Because hey, whether I need that for my electrification, whether I need it, you know, for my quantum computers, whether I need it, you know, for >> just make my power bill cheaper.
2:06:15
If you have more energy to go around, the price is going to fall for average Americans. That's a win. >> Yeah.
2:06:21
You know, and I mean the oil crisis. >> Yeah. >> Great.
2:06:26
You know, could become the electrification crisis in the future, right?
2:06:30
If I don't So, I just view that as such a good thing. Yeah.
2:06:32
because we have so mismanaged our nation's energy capacity uh going forward.
2:06:38
Now you know some of these forecasts are just nutty so like that >> 100 nuclear power plants next year who knows >> yeah I mean we did we did have an IPO today Burmy that is planning to make a nuclear reactor by 2038 or something like they they raises money to try to bring it online.
2:07:00
So, I was poking a little bit of fun at that cuz the company's uh 9 months old.
2:07:03
They're valued at $17 billion.
2:07:05
But maybe it will be net good if if we can bring, you know, a new nuclear plant or two online. >> Yeah.
2:07:12
And you know, by the way, you know, I just uh you know, I just became the uh general partner lead on investment that we're doing in the nuclear space at Playground.
2:07:22
So, I'm learning a lot about this. Great. Right.
2:07:26
And a lot of the you know the small nuclear reactors they're not economic right you can't connect to the grid you can't scale them you know the regulatory things so you I do think it's a very fundamental problem you know for the nation and one hey I'm now diving into
2:07:42
that problem as well so you know we need to bring on many gigawatts of additional capacity for you know the future of our nation >> now you know some of the other you know projections there you know that people uh you know point to, you know, I think they're way over their skis, right? You
2:07:56
You know, not everybody is going to add 10 gigawatts of additional capacity.
2:08:01
And I am excited that uh, you know, like one of our companies, Snow Cap, you know, we're out to make uh, AI 100 times more power efficient than it is today.
2:08:16
>> Not 10%, not 10x, 100x. >> It's great, >> right?
2:08:20
You know, as we move to superconducting.
2:08:22
So instead of that gigawatt, you know, I'm going to have a 100 megawatts of cryogenic capacity that produces 10 times more inferencing. Yeah. >> Okay.
2:08:31
That's industry reshaping, you know, and those are the kind of things that are truly going to, you know, I'll say do a better job of the energy capacity we have and right, you know, creating more ultimate capacity for the future.
2:08:44
And what about, you know, does some of the vendor financing going on, some of the demand guarantees that that people have been making, does that uh is that immediately worrisome to you because of some of the the negative events that that came after things like that in the do era or do you think it's >> Yeah, we've kind of heard both sides part of the cycle.
2:09:07
>> Yeah, we we've seen this before.
2:09:07
You know, look at all the optical companies, you know, in the early days of the internet, right?
2:09:13
Well, you know, everything is up and to the right, you know, for uh you know, exponential growth for unlimited periods of time, you know. >> Yeah. >> Okay.
2:09:22
You know, I mean, it was it was a it was a stunning 10 years, >> but it didn't extrapolate to 30 years. >> Yeah. >> Right.
2:09:30
And I see most of the AI think you know I I see you know fundamentally no change for the next two three four years but by the end of the decade some of these transformational technologies will be at scale.
2:09:44
>> So I do see a real shift right uh in them before the end of the decade particularly on uh inferencing cost inferencing performance and power consumption for inferencing which is sort of like the big one uh for broad deployment of AI.
2:09:59
How do you think about I mean there there's a lot of uh founders in our audience who are you know it's it's raining right now and they're they're happy and everything's good.
2:10:10
Um but you've been through so many market cycles.
2:10:12
Is there a story or advice that you give to entrepreneurs or founders who are going through a reset in going through a bare market?
2:10:21
Uh what what how do you think about getting through those hard times?
2:10:27
Well, you know, so they're going to come right right now, you know, and as a CEO, right, your job is the high is never as high >> and the low is never as low, >> right?
2:10:40
You know, your job is to sort of, you know, cut the tops off of the crazy highs and, you know, manage through the crazy lows, right? That's your job.
2:10:47
And if you're not ready to do that, you know, if you're the, you know, the super enthusiastic champion that everything's up and to the right, you know, for the next uh two, three decades, you shouldn't be in the CEO's job, >> right?
2:11:00
That's called a cheerleader, not a CEO, >> and you know, we have way too many companies that are led by cheerleaders as opposed to CEOs, right?
2:11:07
And we can go back over history and identify a number of those. >> Yeah.
2:11:13
>> Also, you know, like I say, I think the next couple of years, I think, are going to be just fine. Yeah.
2:11:16
But then, you know, we're uh you know, as we get to the end of the decade, okay, things are not going to be just fine.
2:11:23
You know, there's going to be fundamental shifts in the economics and the capacity.
2:11:26
And a CEO's job is to, you know, get the balance sheet in place that they can navigate through those times, right?
2:11:34
You get cash when you don't need it because when you need it, you can't get it, >> right?
2:11:40
So, you have to be managing thoughtfully, you know, about when those disruptions occur.
2:11:44
disruptions occur. So you know I do think uh you know particularly a couple of years out from now okay that's your job as the CEO as the leader as the entrepreneur do I have enough capacity to take a meaningfully negative time
2:11:59
>> right and be able to go navigate through it >> what about team building leadership uh how should you be communicating with uh potentially thousands of employees who are looking to the CEO for leadership not just cheerleading >> well you know you're you know the first
2:12:15
job of the leader is to represent reality right you know where it is and you know that is the first job here's where we are life is great but it's not as good as that you know life is bad and here's the situation and it's not as bad of
2:12:32
that first job of the leader being that point of truth that point of reality you know as I say you know the leader's job is to communicate the vision over and over and over again to the point that you are absolutely sick and tired of it, right? You know, you think there can't
2:12:48
You know, you think there can't be a single person that hasn't heard you communicate the vision and just about the time that you are absolutely dreadfully bored by it is the first time it's really getting through. >> Right? It was just that hard.
2:13:01
And about the time your organization knows it, usually one to two years, it takes another year or two for the customers to understand it. >> Yeah. >> Right.
2:13:13
You know and if you change right in less than three or four years nobody understands it. >> Yeah.
2:13:19
You know so right that constencancy of vision purpose mission is a super important thing you know for uh the leader you know preparing for the bad days like we already touched on.
2:13:31
There will be bad days uh that you know you have to navigate through you know and then when you have success you give it all to the organization and when you have failure you claim it allow >> in a transparent you know uh a facing way you know this right and that's how you build you know loyalty and commitment uh from your teams you know and being a CEO is not cut out for everybody period stop you know if you're
2:14:00
not ready to Take accountability if you're and how there's a lot of legacy enterprise software companies that are needing to res react and respond and evolve as you know LLMs and generative AI sort of uh comes online and he said something I wanted uh to get your thoughts on and and maybe some examples but he he believes it's easier to change your technology ology stack than your business model. And he was saying in the
2:15:03
And he was saying in the context of it's great if you're a software company that started and is charging based on value, but if you have seatbased pricing today and your competitors are now selling, you know, not not trying to get you to scale headcount, but just trying to deliver more and more value. Uh that can be tough.
2:15:23
I I I wanted to ask if there was any notable kind of moments in your career where uh you know uh you you witnessed a company sort of be able to effectively evolve their business model um and andor on the technology side and and how those two things play together in your view? >> Yeah. Yeah.
2:15:43
And you know first you know Brett you know just he's a worldclass guy. Yeah.
2:15:47
So, if you have the chance to bring Brett on or Pat on, take Brett.
2:15:51
You know, he's >> Well, we we've both Why not both? >> But Brett is great.
2:15:56
Um, you know, the uh you know, we took uh VMware through the transition from enterprise licensing to a SAS business model.
2:16:08
>> It is butt ugly hard, >> right?
2:16:11
You know, a business model is like a deep rut, >> right?
2:16:16
You know, one of my favorite posters on this was, you know, it showed a deeply rutdded Indian road and the caption read, right?
2:16:23
You know, pick your rut carefully, you'll be in it for 100 kilometers, >> right?
2:16:29
You know, and there's sort of this truth to a business model because your competitors get used to it, your customers get used to it, your supply chain get you, you know, you just get funneled into it.
2:16:39
And when you try to change your business model, man, it is dreadful.
2:16:44
you know the contractual, the legal, the renegotiations etc.
2:16:46
You know, >> team compensation. >> Yeah.
2:16:53
Everything is hard about it, you know, because you know, you build it, you know, every aspect of the business uh around, you know, those business the characteristics of the business model you've chosen.
2:17:05
So, you know, Brett is right on changing the technology is quite a bit easier than changing the business model.
2:17:11
Now you know the master stroke is if you can keep sort of that interface to the customer and the business model the same and do a major retooling right of the uh technology underneath it.
2:17:24
Okay, now you win, right?
2:17:26
And by the way, you know that, you know, the first major success in the enterprise to SAS conversion, you know, my uh, you know, friends at Adobe, you know, they navigated through it.
2:17:35
You know, I mean, that was uh, you know, uh, you know, their master stroke and many others followed in their footsteps uh, after it, but that was hard, you know, for the transition.
2:17:45
Now, as we go to AI, most are going to be in a SASlike business model uh, going forward.
2:17:51
You know that said the user experience in an AI native application is dramatically different right you know you know you know and if we think about it you know and I you know in very broad terms the computer has I have been adapting to the computer for the last 50 years in an AI world the computer adapts to me >> right you know if you think about the fundamental shift of that where now I am speaking to it on my terms.
2:18:24
It's hearing me in my way, you know, it's understanding my phase, my recognition, my thoughts, right?
2:18:30
Contextualizing them, you know, in my language and understanding, you know.
2:18:35
So I do think the idea of just flipping to a SAS AI business model, you know, versus an AI native experience that truly harnesses the full breadth of that power.
2:18:44
I think many uh you know, people as they talk about the AI conversion of their applications wholly underestimate what will be associated, right, with truly creating an AI native experience for the future. Everything changes. >> Yeah.
2:19:00
Not just even business model going maybe from seatbased and and fixed contracts to value or selling the work but it's even the entire user interface and the the the like what does the application look like? What does it feel like?
2:19:14
Is it feel like hiring and managing a team?
2:19:16
Does it feel like hitting buttons yourself? >> Yeah.
2:19:20
>> So doing both of those things at the same time. >> Okay.
2:19:22
Actual >> actual last question.
2:19:24
Is it true or do you still have a VMware tattoo on your arm? So, uh, no, I don't. >> Okay.
2:19:35
>> And it was so funny because when, you know, my, you know, >> but you did at one point. >> I I did, right?
2:19:39
But it was like one of those 30-day tattoos. >> Okay. >> Okay. That's great.
2:19:45
>> So, you know, and I came home from that trip and my wife, you know, somebody sends the video to my wife. >> Yeah. >> Right.
2:19:52
you know, the tattoo artist and everything like that.
2:19:55
>> And she says, "You may not have that on your arm >> by the time I go on vacation with you." >> That's amazing. >> Yeah.
2:20:04
People think tech uh tech people are hardcore today when they launch like a video or something or they put up a billboard like the the the 30day living with your brand for 30 days is a new level and we thank you for your service to the technology industry, Pat.
2:20:17
Thank you so much for being on the show. This was fantastic.
2:20:20
>> Thank you for the work that you're doing. It's very important. >> I'm very excited.
2:20:23
We I mean we'd love to have you back dive into all the different portfolio companies also what's happening on the on on the glue side in terms of actual productization.
2:20:30
I know we spent a whole bunch of time in the AI world and talking about the other labs but uh there's so much more that we could go into and but we'll give you back the rest of your day. Have a great day. >> Thank you. This was a lot of fun. Thank you so much. >> Thanks so much. Talk to you soon. >> Cheers. Bye.
2:20:43
>> Let's pull up this video of the Meta Rayban display. Be right back. >> I will react to this.
2:20:46
The Meta Rayban display has a new video that we are going to pull up.
2:20:50
In the meantime, let me tell you about profound.
2:20:52
You've heard that you can now buy products on chat GPT.
2:20:56
You're going to need to get your brand mentioned in chat GPT.
2:20:57
You can reach millions of consumers who are using AI to discover new products and brands.
2:21:04
We did this just yesterday.
2:21:04
We found some new microphone stands.
2:21:06
I took a picture of a microphone stand that I saw, dumped it in Chat GPT. What is this? Wound up.
2:21:12
I didn't check out actually in the chat GPT workflow, but we were pretty close to getting there.
2:21:17
And you're gonna want your brand mentioned in chat with profound.
2:21:18
Let's watch the meta ray bands video right now.
2:21:22
This is we so we've demoed this.
2:21:25
Both Jordy and I have looked at this.
2:21:27
This is the what the heads-up display looks like.
2:21:29
It looks better when you're actually using your human eyes instead of trying to film through it.
2:21:33
Uh but the navigation, the heads-up display, the COD mini map, uh I think they should give it much more Call of Duty aesthetics if they want it to really pop and go viral.
2:21:42
Tyler, what what's your reaction? Would you wear this?
2:21:45
>> So So is it like running through your phone at all?
2:21:48
>> So I believe we we were debating this with uh with Simon on the show when he was here.
2:21:53
I I believe that almost all of the compute is done on the phone and it is serving up maybe like a small JSON file that it renders in on the headset or maybe it's even just streaming images.
2:22:07
the im like the the the bandwidth over Bluetooth has to be limited and then the compute on the device is also limited.
2:22:12
So I would imagine that it doesn't have GPS.
2:22:14
It's not doing any ondevice inference.
2:22:17
It's I it's really doing limited just this is the graphics that I need to show.
2:22:22
Render that graphics and pipe it in there but then keep that battery life going as long as possible. >> Yeah.
2:22:29
Cuz I mean uh w with these these are the um the Oakley ones. Yes.
2:22:32
And these like when you set up you have to connect to your phone.
2:22:36
That's like the very first thing you >> So these you definitely have to connect to your phone.
2:22:39
They do not work without a phone.
2:22:41
Um >> over time you would imagine that they try and bootstrap a true platform on top of this and say >> as soon as we have developer access to this I'm going to add the sound effects.
2:22:50
So it'll using the live AI feature you got you walk outside you're walking down the street.
2:22:55
The glasses are you know monitoring the situation >> and you're getting you're getting these kind of sound. >> It would be electric.
2:23:01
Uh Bobby in the chat asked, "Why don't we use lav mics?"
2:23:06
Well, we were discussing it.
2:23:06
Turns out lav mics are complicated. We're figuring it out.
2:23:10
We're going to see some updates. >> Lav mics are coming.
2:23:12
We want to be able to >> walk around. Enjoy the horse. >> Enjoy the horse.
2:23:16
Can we go to the horse cam?
2:23:18
Do we have a horse cam today?
2:23:18
I know I'm putting the production team on on short notice with that one.
2:23:24
>> Get the horse cam going.
2:23:24
I mean, this >> Here's the wide. >> Let's go.
2:23:27
We got the horse over there and the gong over here.
2:23:29
We're filling out the studio.
2:23:30
We're having a lot of fun. zoom in on the horse.
2:23:32
This is it has been unboxed.
2:23:35
Uh some folks in the chat were asking what's the genesis of the horse.
2:23:39
Uh there's a few things we we we enjoy equestrianism. We enjoy polo.
2:23:44
Uh this was from the early brand exploration that we did.
2:23:46
We found that there was a bit of whites space there.
2:23:49
People a lot of people in tech are >> too many people in tech media were afraid to show their appreciation.
2:23:52
And so in some ways the horse is a monument to technology journalism.
2:23:57
Obviously many technology journalists grew up >> you know uh with a very equestrian lifestyle.
2:24:05
So we thought this was considering the show wouldn't be possible without the brave reporting yes from uh many technology journalists.
2:24:10
We thought why not put a horse uh >> also I have a buddy Truman Saxs. He's been on the show.
2:24:18
He runs a company called Cubby.
2:24:20
And uh his he's he's obsessed with horse power, horse metaphors.
2:24:24
He's obsessed with the idea of channeling the energy of a horse, much like I channel the energy of a golden retriever.
2:24:29
Uh the horse is uh very uh it's a noble stallion.
2:24:32
It uh it has a lot of uh noble qualities that you should be emulating.
2:24:39
And so the horse made its way onto the soundboard and now the horse lives physically in our world every day at the TVPN Ultradome.
2:24:47
Uh anyway, uh let me tell you about numeral sales tax on autopilot.
2:24:54
It's a horse for your sales tax.
2:24:57
Suspend less than five minutes on per month on sales tax compliance.
2:25:01
That's faster than a horse can run around the Kentucky Derby, I think.
2:25:07
I don't actually know how long it takes.
2:25:09
>> Breaking uh some breaking news.
2:25:09
Barry Weiss to be named editor and chief of CBS News. Congrats.
2:25:15
>> People were predicting >> promotion >> trade deal.
2:25:20
>> This is such a brutal deal.
2:25:20
There's another post here.
2:25:22
Paramount will announce deal to acquire free press and bring Barry Weiss to CBS News on Monday.
2:25:27
I feel like this deal has been announced 50 times so that when it actually gets announced, >> I wonder why are media people obsessed with media people? Turns out yes. >> Yes.
2:25:38
>> People love reporting on this.
2:25:38
You could never get this much press for a $250 million acquisition of your, you know, database ERP enterprise AI company.
2:25:44
You would be begging for Bloomberg and Techrunch pieces.
2:25:49
But when you're in the media, the media loves to tell a story about media and so you get a lot of press.
2:25:54
Uh, in other news, Arand over at Perplexity is announcing the browser, which I think was teased earlier, but he said announcement in 3. 5 hours.
2:26:04
I'd love to get an update from Tyler on what this announcement was and uh how it's shaping up, what the timeline's saying.
2:26:11
Do you know what's going on in Perplexity World or are you perplexed?
2:26:15
>> I believe that the main announcement is that Comet is just free. >> Okay.
2:26:18
>> Um >> it was pre it was originally 20 or 200 a month. >> Yeah.
2:26:21
Well, there's two different plans.
2:26:23
Um I think it was like usage based.
2:26:23
Um and I think it was also I mean it was recently in it was in beta as of very recently and I think it was released uh probably to just the paid plans. Yep. And now it's fully free.
2:26:33
Well, you should give it a try again. See if you churn. See if you stick around. Uh, some people love it.
2:26:38
Some people I've just kind of stuck with Chrome.
2:26:40
I'm kind of still living in Chrome world.
2:26:42
Uh, although I do Chrome on desktop, Safari on iPhone, which I feel like is the worst possible combination.
2:26:48
Like I should have some bleed over, but uh uh I I posted a screenshot of uh some App Store rankings for Sora and Microsoft Teams.
2:26:57
>> So, is this a mobile based browser?
2:26:57
Can you get Comet in the >> Good question.
2:27:01
Let's go to Tyler Cosgrove, our IRL Culie.
2:27:03
Um, anyway, let me tell you about Finn.
2:27:08
ai while everyone looks that up.
2:27:10
It's the number one AI agent for customer service, number one in performance benchmarks, number one in competitive bake offs, number one ranking on G2.
2:27:17
>> Yeah, I don't believe Comet uh is available as a mobile app. Um, Yep.
2:27:22
Which again I it's very hard to be a 20 billion dollar company trying to grind to product market fit. >> Yes.
2:27:32
>> For a new for a new product >> and the browser company was working on something similar and there's news that OpenAI will be launching something as well.
2:27:40
>> I mean even Anthropic now has a um extension which is it's very similar. >> Okay. Yeah. So it bends in clawed 4.
2:27:44
5 into your browser basically.
2:27:47
And uh is it taking do you know if it's taking I wonder if it's taking screenshots or looking at the HTML and JSON as it comes in acting at a lower level.
2:27:57
>> Yeah, I don't know exactly.
2:27:57
I don't think it's um agentic like comment.
2:27:59
One of the main things is that it can actually like make actions.
2:28:02
I don't think um the claude one is uh actually agentic yet.
2:28:06
But >> Proplexity is a fascinating story.
2:28:07
Uh you know we've had fun with the founder on the show.
2:28:11
Uh but it'll be interesting to see how it plays out.
2:28:13
feels like there's a whole bunch of big deals that need to be ha that need to happen under the hood.
2:28:18
Uh dance partners with, you know, they tried to buy Tik Tok and you could see partnering up with a with a another legacy tech company, revitalizing them.
2:28:26
We saw the browser company go over to Atlassian and there's so many different ways this could break, but uh good luck to the team over there.
2:28:34
Uh in other news, open source Pantex Cap says, "Here's the thing on data centers.
2:28:39
I laughed at all the dumb sellside reports on data centers being cancelled earlier this year. remember those?
2:28:44
People were saying, "Oh, yeah, we're not going to build data centers."
2:28:45
Turns out that was >> Satcha was pulling back.
2:28:47
Um, and I read I read a little bit from this.
2:28:51
>> He went back on pulling back uh there's news today in the journal uh that that Satcha said that uh oh, where is it? It's somewhere in here.
2:28:59
Uh amid a bunch of other articles we won't be reading.
2:29:03
Microsoft CEO gives up some responsibilities.
2:29:05
Nadella said in a a note to employees that Judson Alto has been promoted to helm the bis the company's commercial business in essentially overseeing a large swath of the technology giants operations.
2:29:17
Uh Nadella is saying that he needs to focus on the company's biggest artificial intelligence bets.
2:29:25
So he's going back in and and really doubling down on AI personally and he's promoting other folks to take stuff off his plate so he can refocus on AI.
2:29:33
So, kind of a a retrench into uh let's go big on AI.
2:29:39
>> Yeah, it was I was pretty I got a little I got personally a little bit bearish when Satia was pulling back, right?
2:29:46
He has more >> data than almost anyone on Earth, right? Very smart guy.
2:29:53
He uh he was basically actively getting fit.
2:29:56
They're do they're doing rounds of layoffs.
2:29:58
Uh as he pulled back, a lot of other people, you know, have stepped up.
2:30:02
Oracle's obviously seen a massive appreciation from leaning into >> totally >> and you could and you could see Microsoft feel like filling that gap if Oracle hadn't stepped up. >> Totally.
2:30:13
Um but uh what open source or Pantex cap Yeah.
2:30:19
>> is saying but he says uh you know I laughed at all the dumb sellside reports on data centers being canceled earlier this year.
2:30:25
Made a ton of money going long against that nonsense and bearishness but now we're on the other end.
2:30:30
I don't think people understand the actual logistical hurdles in place for data centers now.
2:30:34
First, power companies are overwhelmed, tired of all the FUD.
2:30:38
They're putting up huge barriers to weed out all the BS.
2:30:40
This earlier, if you want 500 megawatts from Georgia Power, you now have to put up about 600 million upfront.
2:30:46
That's just such a massive number to be locked up.
2:30:48
Uh I'm sure people are also like lend now lending, helping people finance that.
2:30:55
Um and he says second communities are pushing back more deals are getting cancelled on it.
2:31:02
The water we saw that the Amazon data center was in Tennessee where was it that just got cancelled.
2:31:06
Uh >> he's saying water consumption is a real issue.
2:31:11
Data centers actually help electricity prices go down but overall rates are shooting up and they're getting blamed.
2:31:17
getting blamed. So I want to dig into this more >> unpack this more because that's not >> this was always my anti-face anti-fast takeoff take which was even in this world where the algorithms hold the scaling is all you need the bitter lesson holds it's like at a certain
2:31:34
point you're just moving like billions of tons of sand and refining it into silicon to make chips and like the laws of physics come into play at a certain level if you're just moving an immense amount of equip equipment around the world and you need more boats to move the GPUs from one place to another. You You need more energy.
2:31:51
You have to build a new Hoover Dam every two minutes because you're accelerating so fast and it feels like there will be some laws of physics that will kind of act as a drag.
2:31:59
I don't know Tyler, what do you think?
2:32:02
>> Uh I mean yeah but uh periodic labs is kind of like the answer to that.
2:32:04
I mean obviously like at some level you always have to like you know conform to physics. Sure.
2:32:10
>> But I think we have a long way to go where like you can actually have >> the LLM improving the chips.
2:32:14
How many days until AGI on the on the on the day tracker >> right now? It's it's 3600. >> 3600.
2:32:21
So that's what 12 years or something.
2:32:24
>> Yeah, I think um >> so based on today, based on this week, based on Sora 2, >> it hasn't been updated since Sora, I think I I need to kind of give a good think about this, figure out where I am.
2:32:34
But >> what about you, Jordy?
2:32:35
uh if the singularity is is is roughly 10 years away uh does the soratu news does the does the energy consumption news move your estimate out or move it forward?
2:32:50
>> I think I'm moving a little bit out.
2:32:50
I'd probably add another 100 days to the singularity counter personally, but feel free to feel free to fence. signals.
2:32:55
Um, yeah, this this is a Wilmanitis post from yesterday saying, "Is this the product you release if you believe you're three to four years away from super intelligence?" >> Yeah.
2:33:12
Well, if you're looking for super intelligence for CRM, head over to Adio.
2:33:17
Adio is the AI native CRM that builds, scales, and grows your company to the next level. Get started for free.
2:33:21
Uh, also another extremely AGI pilled news headline.
2:33:26
Uh, Google ads are now live in Gemini. Let's go. >> There we go. We love ads.
2:33:32
>> Where was Deep Mind on your AGI pilled uh category?
2:33:36
>> Ads were ads were always on.
2:33:37
>> I think this is critical path.
2:33:37
I don't think this is bearish.
2:33:39
Let's take 2,000 days off the singularity counter. It's coming tomorrow.
2:33:42
Uh considering Gemini is the only provider which doesn't allow users to opt out of training their models on their data, which I don't think is true because we just got news about um uh the Meta team will be training ads and targeting ads based on your interactions with Meta AI chat bots, which makes tons of sense.
2:34:00
Uh targeting here will be off the charts as Wasteland Capital.
2:34:04
Every data point about you, every question you've ever had sold to the highest bidder.
2:34:07
And so if you're just chatting with Gemini or you're chatting with uh Meta AI, you'll start seeing ads that reflect what you've been searching for in the chatbot or talking to the chatbot about uh makes a ton of sense.
2:34:21
I'm sure there's going to be a bunch of weird things.
2:34:22
It was already weird where people would say like, I talked about this with my this product with my friend and then I saw a targeted ad. How did that happen?
2:34:29
It was already feeling like creepily accurate and I'm sure this will make it way creepier but uh it'll be good for the business probably. Right. >> What do you think? Bullish.
2:34:39
I >> think it's pretty good.
2:34:42
>> You know we're bullish ads. >> Bullish ads.
2:34:43
Well, we are also bullish.
2:34:45
Shal a good friend of the show.
2:34:45
He has some big news and we're going to bring him into the TVP and Ultra Jump from the reream waiting room. Shel, how you doing? >> Boom. >> Great. How are you guys doing? >> Fantastically.
2:34:55
>> It is great to have you back. It's been too long.
2:34:57
>> Awesome chat with Pal Gellinger.
2:34:57
Really went deep with him. It was a lot of fun. >> That's amazing. Yeah.
2:35:01
I mean, I feel like the only like like since I was last on, you guys have had like Zucker, Mark, like a bunch of I feel like the level >> getting on TVPN now is a really big deal.
2:35:17
>> Well, we're happy to have you. Uh, give us the news.
2:35:20
What What What's new in your world?
2:35:23
>> Yeah, so u we just announced yesterday that we we raised a new $140 million fund. Let's go. Multiple time guest. Massive new fund. Congratulations.
2:35:35
One gong hit for every fund. You're on fund. >> It's crazy.
2:35:41
I thought you guys were on.
2:35:42
I I would have if I had to guess what fund you're on, it would have been like at least five by by now.
2:35:45
Maybe you're like you got that wise unk status or something.
2:35:51
But uh uh yeah, just just getting started. >> I appreciate that. Yeah.
2:35:55
It feels like we're still brand new though.
2:35:57
like it >> I still feel I still refer to myself as like a new investor but we've been running BTV for six years now.
2:36:06
>> Any change in strategy with this fund?
2:36:08
Uh or can you just restate the thesis for me?
2:36:12
>> Yeah, so the thesis is investing in the best pre and seedstage founders building fintech companies.
2:36:17
Yeah, >> we take a pretty broad view on fintech.
2:36:20
So for us there's a lot of stuff that you might not think about as fintech.
2:36:24
Vertical SAS is probably the main one.
2:36:26
Uh so if you look at toast, Shopify, those kind of mindbody those kind of companies are actually largely fintech companies.
2:36:32
So 80% plus of the revenue for those companies comes from financial services.
2:36:37
>> Um so we invest in those kind of companies.
2:36:39
We invest in some B2B marketplaces, but primarily we're just investing in the future financial services.
2:36:46
>> Um what's the mood on the ground with LPs?
2:36:49
Like is it easier than in past years to raise this fund?
2:36:52
Uh, is it is it contrarian to be focused on fintech when everything's AI or were you able to kind of roll that into the narrative because every fintech company is obviously leveraging AI now?
2:37:04
Uh, what what talking points were you running through?
2:37:09
>> Yeah, it was definitely I mean the the mood on fintech is definitely not what it was when we last raised which was 2021.
2:37:15
Nothing could could have been I mean in >> 200x revenue multiples for a team that had started two weeks ago and had like you know an idea of a customer. It was absolutely crazy. >> Exactly.
2:37:30
Um we actually ra we announced it yesterday and we we haven't even started deploying from the fund but we actually raised it a little while ago.
2:37:38
I think now things are probably a little bit easier, but it was it was a harder fund raise for us than than the last one in 2021 for sure.
2:37:46
But um we'd actually we actually are larger than we had planned for.
2:37:51
We had planned to shrink our fund size and the idea was last fund took us three and a half years to deploy.
2:37:57
This one we wanted to go back to a three-year cycle.
2:37:59
So we were going to go from 150 to 125, but we ended up um ended up going to 140.
2:38:03
We wanted to have we had some LPs that we really wanted to have involved.
2:38:07
Um, so overall it wasn't too bad and uh we're excited to deploy from it.
2:38:12
But I I think since we raised it and now there's actually it's actually like the market is ripping.
2:38:18
You had um Chime, Clara, uh, E Toro go public circle and you have so many more coming.
2:38:27
Wealthfront non ethos in our portfolio.
2:38:31
>> Um, and so I think the market is pretty hot on fintech.
2:38:34
So actually I was just before this talking to another fintech investor uh who's who's about to raise and I said now is a great time to raise the market's awesome for fintech.
2:38:44
>> What uh I mean you've been through a proper cycle.
2:38:46
What uh what what advice did you give to founders at that time?
2:38:52
What are the stories that you share about like uh buckling up and getting through rough periods, rough patches as a founder or entrepreneur?
2:39:02
Yeah, it's just like do what you can to survive.
2:39:04
And if you live another day and you're a smart team, you'll figure it out.
2:39:09
And we've had that happen.
2:39:09
Like some of our companies basically had no growth for two years and then all of a sudden the last year had total inflection.
2:39:19
>> And um >> and I think 23 and 24 22 and 23 were really hard for fintech because a lot of these companies are selling into other fintech companies.
2:39:28
is if you're an infrastructure company, you're selling into other fintech companies and the funding completely stopped.
2:39:33
There were all these, you know, Neoan for X that got funded in 2021, like way too many of them.
2:39:41
And so, you had, it was just so easy to grow in 2021.
2:39:43
And then 22 and 23, you had to grow significantly just to tread water, just to stay like even.
2:39:51
Um, and then now we're seeing companies grow a lot.
2:39:53
So, it was really just stay alive, hunker down.
2:39:55
A lot of our companies got to a point where they didn't need to raise money and obviously that's the best time to raise money.
2:40:02
Now, now investors are are throwing cash at them.
2:40:05
>> How are you thinking about stable coins?
2:40:07
I I think we've talked about them, you know, before.
2:40:09
It's in such an it's an interesting time to me because people are so bullish on stable coins as a category and yet it seems like the people the companies best suited to capitalize on this opportunity are not startups.
2:40:25
And I don't know if you you feel the same way, but you see Stripe's launch of open issuance.
2:40:29
launch of open issuance. you see uh which was interesting to me because we had a company on like a week and a half prior that was building the exact same product and then Stripe launches it and they already have you know >> there's also a public company circle
2:40:43
that like is mature in that space and there's a tether >> and then and then it feels like also interesting because even the the scaled stable coins that tether USDT USDC are now under attack from every institution y >> on earth and you know the the pie will grow but at the same time it's going to get sliced up. So, where are you
2:41:01
So, where are you thinking about the opportunities to deploy against this kind of um technology at the early stage?
2:41:10
>> Yeah, I think you're right.
2:41:10
I think um one challenge we're having right now is we're just seeing so especially we talked about last time I was on the the bridge acquisition since then we've just been seeing so many companies built in every sector within stable coins.
2:41:24
Um, and I haven't seen that many that compelled me to say, "Okay, this is you're going to beat Stripe on this."
2:41:33
Like, Stripe is executing at a amazing clip and I I haven't found the the use cases I found are specific to a geo.
2:41:40
So, like you're solving a problem in Latin America and I think you might be the best person for that.
2:41:49
You might be able to get you need in that region.
2:41:51
Yeah, I'm an investor in Mnteo, which is doing something, you know, similar to Bridge, but just for LAMP, right?
2:41:58
And so, they've been building banking relationships for years now.
2:42:02
And so, they're they're in a good position.
2:42:04
But, but yeah, it's a it's a it's a weird thing for venture capitalists and angel investors to be, at least from my point of view, super bullish on the category.
2:42:15
It's like it's hard to deploy against it other than just buying some Stripe secondaries. >> It's totally true.
2:42:20
Um, one thing that I like a couple days ago, Stripe announced that they were doing um, agentic payments and obviously that was coming.
2:42:30
Um, but it's so hard like I I met at least a dozen companies in the last year doing aic payments and I feel like as a VC you're not supposed to say, you know, what if Stripe does this, but it was so obvious that Stripe was going to do it. >> Yeah.
2:42:44
But saying what if Stripe does this is very different than what what if Google does this, right?
2:42:49
Or what if like Morgan Stanley does this or JP Morgan or like an older company that maybe isn't in founder mode, maybe not in Silicon Valley, maybe not Sam Alman on the cap table with the company that's doing the thing.
2:42:59
It's like, you know, this is going to happen.
2:43:01
Uh anyway, uh we have one last >> payments are going to be awesome. >> Yeah.
2:43:07
>> Um and it's and like so it's the same thing that you just said, Jordy, which is like I believe in the category.
2:43:11
I believe that I will be making purchases through like through my chat interface.
2:43:20
Um but I don't think that there is necessarily a lot of opportunities for new startups. >> Yeah.
2:43:24
Uh last question from the chat.
2:43:27
They're calling you the people's mayor of San Francisco.
2:43:28
And everyone wants to know what's the best place to get Indian food in San Francisco or New York.
2:43:35
>> It's a great question. Okay. All right.
2:43:38
So, New York, my favorite restaurant in the country in the world Indian restaurant is my favorite high-end Indian restaurant is place called Ambassadors Clubhouse. Okay.
2:43:47
>> It's uh it's in London >> and there's Ambassadors Clubhouse in Jim Kana and they're actually opening in New York next year.
2:43:54
So, that's going to be big. >> Whoa. Lines out the door. >> Lines out the door.
2:44:00
>> And then there's uh >> book a flight now.
2:44:01
Sema in New York is very good. Mhm.
2:44:05
>> I got to say in in San Francisco, I don't think we have excellent high-end Indian food.
2:44:12
>> Uh there's a place called Etan in Palo Alto.
2:44:13
And then my best my favorite spots are Udipi Palace on Valencia Street, Kolapasi also in the Mission.
2:44:20
Um and then there's a new spot called Jalabbe Street on Hate or In the Hate.
2:44:25
So, uh those would be my three go-tos in San Francisco.
2:44:30
all kind of hole-in-the-wall places, but that's the best Indian food.
2:44:35
>> Well, thank you so much for breaking it down and coming on the show >> and excited to get a >> I wouldn't be surprised if it's impossible to get a reservation at any of those places after this pod drops. >> Thank you so much.
2:44:45
Great catching up with you. We'll talk to you soon.
2:44:48
>> Uh let me tell you about adquick. com.
2:44:52
Out of home advertising made easy and measurable.
2:44:54
Say goodbye to the headaches of out ofome advertising.
2:44:57
Only AdQuick combines technology, out of home expertise and data to enable efficiency.
2:45:00
in this ad buying across the globe.
2:45:02
We have our next guest in the reream waiting room.
2:45:05
We will bring in Santi from dual entity dual entry.
2:45:09
Sorry, >> what's happening? >> How you doing?
2:45:12
>> Sorry to keep you waiting.
2:45:14
>> What's the latest in your world? Please. >> Do we have video? Do we have sound? No.
2:45:21
>> Well, we do have another advertisement.
2:45:24
We got >> Don't tempt John. >> get bezel. com.
2:45:26
Your bezel concier is available to source you any watch on the planet. Seriously, any watch.
2:45:31
>> What did you call it yesterday?
2:45:33
>> AGI for your wrist or something. I don't know. That's probably on one.
2:45:37
>> Out of home ads for your wrist.
2:45:39
>> Out of home ads for your wrist. >> Uh, it works now.
2:45:43
>> Hey, welcome to the TBP at Ultra Dome. >> What's happening? >> How are you doing? >> Thanks for having me.
2:45:49
>> Sorry for keeping you waiting.
2:45:49
We were going long with Shiel, but good to have you here.
2:45:52
Uh, please introduce yourself, the company, and what's the news today. >> Thank you. I'm Santi.
2:45:58
I'm a co-founder of Dual Entry.
2:46:01
We're an AERP and we're announcing our $90 million series A. >> Let's go.
2:46:09
>> Nothing like there's nothing like the feeling of coming out of stealth with an AI ERP with 90 million of funding.
2:46:13
It just doesn't I don't think it gets better than that. >> It's fantastic.
2:46:18
Led by none other than >> How long you been How long you been working on it?
2:46:21
What were you doing before?
2:46:23
>> We've been working for about a year. It It does feel good.
2:46:25
I will say it does feel good.
2:46:27
Um, but it feels even better to be able to run an ERP that doesn't suck like the 1990s.
2:46:31
But, um, >> for sure, >> we've been working for about a year now under stealth.
2:46:36
We used to run a product company before and we suffered one of those legacy ERP migrations for over nine months. >> Oh, yeah.
2:46:42
>> And, uh, you know, as good entrepreneurs, whenever you have a problem, that means there's an opportunity to solve.
2:46:47
>> When I'm going to steal Jord's question, when did you know you wanted to build an ERP?
2:46:54
the first the first day we were out of the out of the other prior company.
2:46:56
We promised ourselves >> that was the day we were going to do it and it was I'll tell you it was a very unpopular you know we had we had a ton of people saying it would back you guys for any other idea other than that one because that sounds like a very stupid one. >> Yeah. So what is the traction?
2:47:09
How are you how are you how are you I mean it's the hardest system of record to rip out. It's an enterprise.
2:47:14
It's a company that is built on something else out there.
2:47:18
They're maybe unhappy but they're really stuck.
2:47:20
Are you better at migrations with AI?
2:47:23
Are you just better at sales?
2:47:23
Like, how do you actually onboard customers?
2:47:28
Because I have to imagine that you have some serious traction to uh hit this series A so quickly.
2:47:34
>> Yeah, that's exactly right.
2:47:34
I mean, we we understood from the very beginning that like ERP is like having chronic back pain, right?
2:47:40
You it's sort of a pain that keeps growing with time.
2:47:42
But if if you have to think about about back pain surgery, it's very risky. It's expensive. It might go wrong.
2:47:48
And we had to fix that problem.
2:47:51
And the backpin surge in ERPs is the implementation process.
2:47:53
It takes, you know, nine months.
2:47:55
The failure rates are over 50%.
2:47:57
It's very expensive at times two, three times the software fee.
2:47:59
So we we fix that with our next day um ERP migration engine.
2:48:04
So we actually plug in to whatever system you're coming from, whether it's legacy Europe or starter cloud, and we we actually have you live by the time that you're doing the demo in most cases, >> whether they like it or not. >> Wow.
2:48:14
Pull the data from their cold dead hands. I love it.
2:48:16
Uh what >> it's our customers data?
2:48:19
>> Yeah, of course it is.
2:48:19
Uh what what's the speaking of customers?
2:48:21
What's the median size of a customer?
2:48:23
Are you going uh mid-market to start?
2:48:26
Are you going for the huge Fortune 100, Fortune 10?
2:48:28
Uh how do you see the market now and then where will it develop?
2:48:35
>> Yeah, so we we're targeting the mid-market.
2:48:36
We're starting with companies as small as 5 million, but we have publicly listed traded companies running on dual entry today.
2:48:42
They're a full set of books.
2:48:43
not not a subsidiary, not a you know, not a department, just the entire company.
2:48:48
>> And the most important thing for us is to give customers a certainty that this is the last ERP that they'll ever have to do. Yeah.
2:48:54
>> You know, you you jump on, you you upgrade from starter cloud and this will take you all the way.
2:48:59
>> Why why do you think uh why do you guys think you're having more success getting people to trust you, switch over?
2:49:04
Is is a lot of it helping uh I imagine agentically sort of do the migration or or what's what's working now where uh you know some of the investors you talked to maybe a year ago we're saying I'd fund you but do doing anything else? >> Yep.
2:49:24
>> Well I think I think AI is bridging that last mile in migration especially when it comes to the mappings that need to be done.
2:49:29
done. and they're very tedious and they're conventionally done by overseas teams then you know it takes weeks and it's errorprone um but I think you know the best thing is we don't have to ask our customers for trust we just say go use it demo it out you don't have to come with us with questions and think about all the hypotheticals because you also don't know how your business is going to evolve and you know it's we're
2:49:49
the we're the we're the only ERP in the world that we know of that can actually give you a free trial that lets you use your data and consume it and by the way this ERP migration engine is perpetual so like you can plug it and and if it
2:50:00
takes you a month to make that decision, the data will continue to flow, you can show your CEO your reporting, you know, you can really play around with it, test all the edge cases before you make a decision. Um, and that's how we've won
2:50:08
Um, and that's how we've won the trust of our of our customers. >> Interesting.
2:50:12
Are you how are you how are you actually uh uh and and feel free to pass on the question if it's too much detail, but are you basically saying like you have an existing legacy ARP, finish out your contract, but you know, we'll start the migration over.
2:50:24
You can start building trust with us now and then when it comes up for your contract renewal on your existing RP, you can make the call then do I want to stay with the legacy or do I have enough trust with you guys to just you know keep running here. >> That's exactly right.
2:50:39
Um we send you a link, you click on it, we authorize the connection and the data is flowing within 24 hours.
2:50:44
within 24 hours. And then you know at times people like it so much more and they realize how much time they're saving that the actual you know starting to starting to to pay for dual entry is not even the biggest issue or the biggest part of the budget is the fact that they're saving you know two or three people doing manual
2:51:00
categorizations and the same journal entries every single month and accounting is a very repetitive task by nature but every accountant doesn't want to be bogged down in the repetitive tasks every finance team member wants to be looking and thinking you know sitting down with the CEO and saying these are the areas where you should be investing ing more this or the other. So we should
2:51:15
So we should be cutting back uh not not chasing invoices or categorizing transactions and drafting the same journals over and over.
2:51:22
So so often times they see that and they make the jump even before the contract elapses. >> Very cool.
2:51:28
What's the what's the go to market going forward?
2:51:30
You guys have been obviously in stealth.
2:51:31
You're out of stealth now.
2:51:33
Are you going to get loud?
2:51:35
You going to buy up all the billboards on the 101? What's the next move?
2:51:39
>> Well well we like to get referrals.
2:51:41
That's the best that's the best way.
2:51:41
Uh so far we've gotten most of our customers through referrals from existing customers and we we hope to continue that.
2:51:47
We've never gotten so many demo requests in in an hour in the past.
2:51:51
Uh but uh we're excited to to use those customers as a way to get the word out and and have more people come come to us.
2:51:57
And there's no reason why you have need to have a dinosaur in your office today. >> I love it. >> Nice.
2:52:02
Uh you should uh yeah skip a skip the dinosaur, get a horse.
2:52:04
We have a we have a fullsize horse here.
2:52:06
So statues are in dual entry is really sounding like this a stallion of the the ERP market.
2:52:13
So >> dual stallion I like that the dinos very inspirational office view right there.
2:52:20
Uh wonderful having you on the show today.
2:52:22
Thanks so much for stopping by.
2:52:24
>> Thank you for having me. Thank you John. Thanks.
2:52:25
>> We'll talk to you soon. >> Good to meet you.
2:52:27
>> Let me tell you about Wander.
2:52:27
Find >> your happy place.
2:52:31
>> Find your happy place.
2:52:31
Book a wander with inspiring views, hotel grade amenities, dreamy beds, top tier cleaning, and 247 concier service.
2:52:35
It's a vacation home but better.
2:52:38
Our last guest of the show is Austin from D0ero.
2:52:43
He's in the reream waiting room.
2:52:43
Let's bring him into the TVP Ultradome.
2:52:46
Austin, how are you doing? How's it going? >> Good. Thanks for having me on.
2:52:50
>> Thanks for hopping on.
2:52:50
Uh, kick us off with an introduction to you and the company and the news today.
2:52:54
Would love to get up to speed on Double0ero. >> Yeah, certainly.
2:52:58
So0ero is a blockchain protocol building an alternative internet for high performance distributed systems and blockchain.
2:53:05
And so at its core, this is a physical fiber infrastructure network run by multiple independent contributors to allow systems like blockchains to go faster than the public internet uh will allow them to go.
2:53:19
>> What's the status quo in the industry?
2:53:19
I feel like if I sign up for AWS, I'm going to get charged ingress and egress bills and they're kind of abstracting the like underlying infrastructure, the physical infrastructure.
2:53:32
At what level are you operating?
2:53:34
How does all of this fit together? >> Yeah.
2:53:38
So, this is what's called an OSI layer 1, two, and three network.
2:53:40
So, this is pretty much the lowest level down in uh the networking stack.
2:53:43
And so, this is the physical cables, the protocols that transport over them.
2:53:48
Uh but fundamentally this is exactly what you're talking about the egress and ingress problems in cloud data centers.
2:53:56
That's pretty much the only thing uh that you get build for uh in a data center when you really think about it especially the cloud data centers.
2:54:04
>> And so this is a way that they generate lock in.
2:54:06
It is easier to stay in the Amazon ecosystem because data moving within Amazon does not get charged whereas data leaving Amazon does.
2:54:12
And this is why you have such a vendor lock in in the cloud market today.
2:54:16
And so if you think about the ultimate mission of uh decentralizing the entire AWS stack or the GCP stack and basically rebuilding these on decentralized rails not just for the point of decentralization but so you have greater choice from a consumer standpoint.
2:54:34
>> The network and connectivity layer really is in my view the first piece of that that we need to tackle and it's in some ways the most broken piece of it today.
2:54:42
The internet is an incredible resource. It connects us all.
2:54:44
It serves data uh in far-flung regions of the world, but it was never built for high performance systems.
2:54:51
And so, Wall Street does not use the public internet.
2:54:55
Facebook does not really use the public internet.
2:54:57
Even all these big internet companies that we think of, they run private networks.
2:55:00
And they run private networks because the internet is too slow to show you Instagram reels.
2:55:04
And so, why are we building uh what is supposed to be the alternative replacement to traditional finance on those rails?
2:55:13
uh how do you who are the suppliers that are bringing internet connectivity to the network?
2:55:18
Do you need to do deals with companies like Akami?
2:55:19
I I don't really know who the major players are in the space that you would be pitching to bring uh resources to the system. >> Yeah.
2:55:28
So today we have 11 independent uh basically fiber contributors to the network.
2:55:34
And this ranges from the folks you'd expect, you know, large trading firms that own and operate their own private networks for highfrequency trading purposes like DRW Cumberland and, you know, Jump to digital asset native firms like Galaxy that are uh this is a new thing for them running private fiber, but it supports a lot of the other business lines that they already do uh like market making and trading.
2:55:58
And then you have data centers themselves, bare metal providers like Teras switch, latitude, servers.
2:56:02
com, cherry servers, uh getting involved on this as well as a way to basically make their regions more attractive for this new burgeoning blockchain and AI infrastructure.
2:56:14
>> Uh and then you sort of have the last category which uh are blockchain native groups themselves like JTO.
2:56:20
>> Uh how does the high frequency trading uh like example work?
2:56:22
I remember reading was it flash boys where they build the fiber lines between Chicago and and the New York Stock Exchange so they can trade commodities faster or something.
2:56:34
Do those just go dark outside of market hours and then you can resell the slack capacity is that the idea or like why would a high frequency trading firm if they built this pipe and they got it and they're making money off of it like in what world do they end up with extra capacity that they want to sell back to basically a grid that you're building? >> Yeah.
2:56:52
So there's two things that are going on there.
2:56:53
Uh, one is simply the technology changes very quickly and so if you are in the business of being a high frequency trading firm, you pay a pretty extreme premium to be on the absolute fastest path. >> Yep.
2:57:05
>> In blockchain, the second fastest path or the fourth fastest path is actually totally fine.
2:57:10
These are still orders of magnitude improvement over the public internet.
2:57:14
And so sometimes they have uh what you'd call sort of secondary inventory that they can contribute to something like double zero.
2:57:19
The other component of this too is there's not that many financial centers in the world.
2:57:25
When you think about it, it's a pretty tractable problem for a large trading firm with a large budget to go out and have a team that connects all these major financial centers.
2:57:33
You look at just Salana, it's run in over 300 data centers around the world. >> Yeah.
2:57:38
>> And when you include other blockchains, that number goes up dramatically.
2:57:40
It's too big of a problem for any one organization to solve and address.
2:57:44
And so what do we turn to?
2:57:46
We turn to a consortium model powered and aligned through tokens and that is what the token is the best unit at.
2:57:52
It can bring disperate efforts together again a common goal to benefit all and it really is an innovation on the traditional corporate structure which of course in itself changed the world in its own right. >> Sure.
2:58:06
>> Uh tokens just happen to be the best way to create this type of alignment. >> Yeah.
2:58:10
>> Your guys's token launched this morning sometime in the early hours.
2:58:14
uh it's now trading at uh somewhere around a $6 billion uh fully diluted valuation.
2:58:19
Uh what goes into like what what's like the what's the process to get to get to this point from a uh more from like a financial standpoint?
2:58:33
Obviously, you need traction with the network and usage and a bunch of different partners, but uh we talk to a lot of, you know, more traditional public company CEOs.
2:58:41
will often talk to them on their IPO day and we'll talk about the different processes around the road shows.
2:58:48
But on the digital asset side, launching a network like this, what are kind of the key steps leading up to this moment?
2:58:53
And congratulations by the way. >> Thank you. Yes.
2:58:57
Uh one of the things that is different about a blockchain protocol than you know any sort of traditional structure is we don't own and operate any of the fiber.
2:59:04
And now you can say sure Uber does not own and operate any of the cars.
2:59:08
Um but at the end of the day the role of the core contributors to the double0ero protocol is to build the software layer with other core contributors that then allow this contribution of fiber units to it.
2:59:20
And so a ton of the work in the early days on this project, apart from the software side of how do we actually build these things in decentralized fashions that allow for permissionless contribution of fiber, but how do we go out and convince these people who have access to fiber that they should take a quite valuable asset that has fairly high opex and fairly high capex and contribute that to the zero network.
2:59:41
And a lot of that comes back to uh you know our contributors for the most part are fundamental believers in the power of this technology.
2:59:49
We don't have today like an AT&T or a Verizon or a Lumen contributing capacity to the network because those are largely either publicly traded companies or companies that are owned by PE firms and they're expecting a very traditional uh dependable type of return from that. Crypto is not that.
3:00:07
Uh this is this is very much the wild west.
3:00:10
Uh the hope of course is that over time we you know uh basically prove that it's worth contributing even a small slice of a global fiber network owned by one of these subc cable firms or you know these traditional telos into the mix.
3:00:24
And we've had a number of outreaches just today uh from more traditional players asking how they can get involved.
3:00:30
Uh the real thing here is all of these companies, it's the same thing that the cable companies were going through in the United States maybe 10 or 15 years ago.
3:00:40
They're they're priced and they're valued and they sell effectively dumb pipes.
3:00:44
And I don't mean that in a derogatory way.
3:00:45
They're quite sophisticated advanced technologies, but they don't have the same ability to take a cut of what flows over them the way you see in other types of spaces, especially in software.
3:00:55
And so the promise of going from selling something for a fixed dollar amount to a token model where uh validators operating on it and traders operating on it pay a percentage of their earnings into the network is a potentially disruptive business model opportunity for these traditional players.
3:01:13
The same way that the cloud providers, yes, there was a technology revolution there, but it was as much a business revolution as anything else.
3:01:22
How has the new admin uh been helpful to you guys even just in terms of being generally pro- crypto?
3:01:29
I I don't know how long you've been uh building the the foundation in the network, but I imagine you know this wouldn't have been the kind of thing you wanted to launch two three years ago for a number of reasons.
3:01:45
>> Yeah, I mean I got involved in crypto in late 2017, early 2018.
3:01:47
Uh worked for a company called Republic.
3:01:50
uh worked for a company called Bison Trails which then got acquired by Coinbase and spent the previous four years working for the Salana Foundation.
3:01:58
>> So I've been through uh the changing regulatory landscape.
3:02:00
You know, we received a no action letter from the SEC for key components of the flow of tokens in the zero network.
3:02:10
>> This is really important.
3:02:12
>> Shout out That was your law firm, right?
3:02:13
>> Shout out >> Let's go. We love lawyers here.
3:02:14
They don't get >> you go. Uh I owe them a drink.
3:02:18
Um yes, >> I I I will say that the uh you know the whole thing that came together here was you know going to the SEC and asking for relief.
3:02:29
So this is a process where you can basically go to them and say we would like to do this.
3:02:32
What do you think of this?
3:02:35
Do you see any problems with this under current securities law?
3:02:36
And so we applied for relief on two specific areas.
3:02:41
The first area was around the reward distributions to contributors.
3:02:43
So basically the network programmatically paying out tokens proportional to the uh the utility that a contributor brings to the network.
3:02:53
This is not like proof of stake.
3:02:55
This is much more similar to proof of work where if you generate 10% of the hash power for Bitcoin, you get 10% of the rewards.
3:03:02
>> But this is really important to us because many of our potential contributors, especially as we sort of move lower down and closer to the direct fiber owners, they are not sophisticated crypto lawyers.
3:03:12
They may not even be sophisticated securities lawyers and and that sort of reassurance that we can get from the SEC of hey this programmatic payment based on this type of model uh they don't see any issue with that from a security standpoint.
3:03:26
>> Yeah, makes a ton of sense.
3:03:26
Um absolutely massive day.
3:03:29
Hope you uh I it's uh >> unlike you know uh public company CEOs, you know the the market closes then you get a little breather.
3:03:39
not you're not so fortunate but um massive massive milestone and uh >> as you start knocking down some of these massive partnerships uh come back on and tell us about it.
3:03:52
>> Thanks so much for hopping on the show. We'll talk to you soon. >> Cheers. >> Appreciate it.
3:03:56
>> He said that the internet is not a series of tubes.
3:03:58
Are you familiar with the famous line that was said in Congress?
3:04:03
I believe the internet is a series of tubes. There's a music video.
3:04:06
This is a video huge sign of respect.
3:04:09
They turned it into a meme. It's from 2006.
3:04:10
It's one of the one of the great sign of, you know, respect in my culture to show you a YouTube video that you might not have seen.
3:04:18
I want to show you this video from Internet Explorer 4. 0.
3:04:22
The internet being a series of tubes. Let's play this.
3:04:25
>> The internet is not The internet is not a big truck.
3:04:28
The internet is not something that you just dump something on.
3:04:31
The internet is not >> This is huge in 2006.
3:04:35
>> The internet of the internet.
3:04:39
>> This is a real clip from real sound. We can cut it.
3:04:42
Uh this is a real clip from a testimony that happened in Washington DC with a lawmaker who's trying to explain the internet to the like to Congress basically was saying it's not a truck.
3:04:56
It's not something you just dump something on. It's not fixed.
3:04:58
It's a it's it's a flow of tubes.
3:05:00
And he was lambasted and made fun of for this, but it's kind of true.
3:05:04
It's like it is the flow rate matters more than the fixed quantity, you know.
3:05:09
It's not it's not a it's not a data center.
3:05:11
It's a it's a series of tubes.
3:05:14
>> It's a series of tubes.
3:05:15
>> Anyway, how'd you sleep last night? >> Uh, brutally.
3:05:17
It's just been such a it's honestly been a brutal week of sleep. >> I got an 85. >> Got a 65.
3:05:23
>> Oh, give me the sound. Give me the sound. Let's go. Yes. >> There you go. >> Boom.
3:05:29
>> Yeah, they have this new feature that tracks your sleep debt. >> Oh, I saw this.
3:05:32
The sleep debt is really really negging me these days.
3:05:34
Like >> I've accumulated two and a half. >> Okay, I'm at zero.
3:05:38
I'm at perfect sleep balance. Not in debt. No surplus, no debt. I'm doing good.
3:05:44
>> It's great because I have this handicap of having a much longer drive in to work.
3:05:49
And so it kind of I really got to push myself, you know, >> equalizes for the twins >> earlier.
3:05:54
Yeah, it it equalizes for the twins. >> Yeah.
3:05:57
>> Um it was great to see we saw Will Robbins in the chat earlier.
3:06:00
He was thank uh pumping up dual entry.
3:06:03
We had dual entry and then double zero back to back a little bit confusing but uh apparently Will did uh the seed of dual entry so he's feeling pretty good with this announcement.
3:06:15
>> Uh in other news in uh inversion space announced ark the world's first space-based delivery vehicle.
3:06:20
This uh video looks remarkable.
3:06:23
ARC enables the ondemand delivery of cargo and effects to anywhere in on Earth in under an hour and offers unparalleled hypersonic testing. Sounds very sci-fi.
3:06:32
Uh they have a rendered video here.
3:06:34
Uh we'll have to dig in and understand the actual timeline for this.
3:06:38
It seems like a very ambitious project, but congrats to everyone over there at Inversion Space. >> Very cool.
3:06:45
And uh we got to get on with London. >> We do.
3:06:49
>> Tomorrow, unfortunately, we will not have a stream.
3:06:51
Uh, we encourage you to >> touch grass, >> watch our recap of the week.
3:06:59
>> Maybe we'll release the recap early. No show tomorrow.
3:07:02
>> Watch the first episode we ever dropped.
3:07:04
There's a lot of lore in there. >> Lot of lore. Enjoy that. >> A lot of fun.
3:07:07
>> We will be off tomorrow.
3:07:08
>> We will be back Monday full force. I cannot wait.
3:07:12
And uh we hope you all have a fantastic weekend, but don't start until tomorrow.
3:07:17
Maybe around five, maybe around 10, maybe burn the midnight oil.
3:07:23
But have a great weekend. We love you. We'll see you Monday. >> See you later. Bye.