0:09
[Music] Hey hey hey. [Music] [Music] Heat. Heat. N. [Music] [Music] Heat. Hey, Heat. [Music] Heat. Heat.
Search this video
1:23
[Music] [Music] [Music] Yeah. [Music] Power.
2:07
[Music] [Music] [Music] Team deathmatch.
2:39
Don't need [Music] [Music] [Music] We came to this world.
3:25
We came to this world to shape our future.
3:36
We came to this world [Music] to reach the stars to shape our future.
3:50
[Music] [Music] We came to this world to reach the stars.
4:24
We came to this world to shape our future.
4:35
We came to this world [Music] to reach the stars to shape our future.
4:49
[Music] Reaching feeling. Music.
5:15
[Music] [Music] Let me [Music] [Music] >> You're watching TBPN.
6:02
>> Today is October 16th. It's Thursday, 2025.
6:05
We are live from the TBPN Ultradome, the Temple of Technology, the fortress, >> the fortress of finance, >> the capital of capital. Um, big news.
6:10
We covered it a little bit yesterday, but uh, big news out of Google.
6:16
>> Sometimes when something just hits the timeline and we're live and we're doing the show, it's hard to process how significant something is. >> For sure.
6:24
And, uh, I mean, what a funny what a funny turn of events.
6:26
In the meantime, while OpenAI is fighting for their life in the timeline against allegations of moving into adult content, Google is saying, "Hey, we cured cancer. We've done it. We cured cancer." >> Yeah. Which way? Western Lab. >> Yeah. Yes. Exactly.
6:42
Uh they didn't actually cure cancer.
6:44
Uh but they made some progress and it's definitely updating some people on what AI can do in bio, what AI can do in in cancer research generally. Um it's very complex.
6:55
I'm not an expert in uh bio or or or any of this stuff really.
6:58
So, um I needed to use a Call of Duty metaphor.
7:03
And so, I read the piece.
7:03
Um it comes from Sundar Pachai.
7:06
He says, um, "Time is money. Save both.
7:09
Easy use corporate cards, bill pays, accounting, and a whole lot more all in one place. Go to ramp. com." Just kidding.
7:15
He said, "An exciting milestone for AI and science.
7:19
Our C2S scale 27B foundation model built with Yale and based on GMA generated a novel hypothesis about cancer cellular behavior which scientists experimentally validated in living cells.
7:32
Now that is not in people, it's not in mice, it's not in rats, it's just in cells.
7:37
And this is just one potential link between a drug and cancer cells.
7:42
But it's very exciting, very pro promising.
7:45
And so I was trying to figure out the the appropriate analogy and what I landed on was of course Call of Duty.
7:52
And so in Call of Duty there are a bunch of players on the map.
7:56
Some of them are on your team, some of them are on the other the opposing team. >> Force headed standby.
8:03
>> The opposing team, you can think of them as cancer cells.
8:05
When when cells develop cancer, when tumors exist, they are on the enemy team.
8:11
You got to hunt them down. You got to find them.
8:12
the people that are hunting them down, those are the killer tea cells.
8:15
That's your immune system.
8:17
Ideally, you want all the tumors, all the cancer cells to be really obvious to your immune system.
8:24
So, your immune system can go around and get a bunch of head shots, double kills, 360 nocopes, of course. Of course.
8:28
Um, but it's uh it's hard because a lot of these cancer cells, a lot of these tumors, they exist in what uh Google puts them as like it says they it says they're cold tumors.
8:43
Basically, they're invisible to the body's immune system.
8:46
You want to turn them hot.
8:46
You want them to light up on the mini map. How do you do that? You got to pop the UAV.
8:49
And that's exactly how this works.
8:52
So uh this particular C2S which is celltos sentence model uh ran a bunch of different uh correlations between uh different drugs and their effect on cells and found a drug that does just that.
9:11
It's a conditional amplifier meaning that it boosts the immune signal and it basically turns these cold tumors hot.
9:17
And now uh this is not new.
9:17
There's a lot of drugs that basically act as UAVs for the immune system that allow the immune system to target cancerous cells or tumors more effectively.
9:30
Uh but why this is special, why everyone's obsessed with this, why everyone's white pelling so hard is because they use AI to do this.
9:38
And of course this was in conjunction with scientists and then it was verified in a lab setting in a real in the real world setting.
9:45
But it's still very exciting.
9:46
Now just to put it in concept in in in in uh perspective uh Google found one drug that was correlated with this effect uh which is great.
9:58
There's something like 600 maybe FDA approved drug cancer indication pairs in the US alone.
10:03
Um and only about 5% of drugs that even get submitted to the FDA actually get approved.
10:10
And so the total number of like discoveries like this, if you just think about it as like a drug that helps fight cancer or helps identify cancer, uh if you think about the pool of those, humans have discovered probably tens of thousands, maybe hundreds of thousands of those, narrowed those down, sent those to the FDA, the FDA's, uh approved maybe 5% of those, and we get 600 that are on the market.
10:32
And then of course some of those are more effective than others.
10:35
Some of them have have uh you know less severe side effects and so they become more popular.
10:40
Some of them are just good at advertising probably and so and or good at you know the cost benefit ratio.
10:46
So they're very effective to a lot of people and it's really cheap that probably is a blockbuster cancer drug.
10:53
Something that's really really really effective but only in a very odd population.
10:57
Uh it's really expensive.
10:59
probably less of a blockbuster cancer drug, but what matters here is the slope more than the y intercept.
11:04
And so right now, if you just look at the scoreboard in your Call of Duty world, uh AI got one point on the board.
11:11
And humanity has like tens of thousands of these discoveries basically if you that's a very rough estimate, but it's like it's a lot.
11:18
We've discovered a lot of cases >> and this is meaningful for one very real reason which is that people have been desperately hoping yes that AI systems would be able to discover novel uh uh cures for cancer. Right?
11:32
This has been something that >> if you look at anybody at any of the big labs over the years, they probably have at least one sound bite where they're talking about this kind of potential.
11:40
talking about this kind of potential. Uh my question is does OpenAI really have time to even compete on this front right it's very easy for Google with hundreds of billions of dollars of revenue >> uh to uh dedicate resources to projects
11:59
like this whereas this doesn't feel core to >> open AAI's like even roadmap right so this is that but but what ends up being funny is that the the timing right within a basically a 24-hour hour period, you get the erotica announcement fast followed by >> uh this and um >> well played by Google. >> Yeah, I I don't think that Cinder had
12:20
>> Yeah, I I don't think that Cinder had this in his back pocket and was waiting to to drop it.
12:25
>> I'm not implying that either, but I'm I'm sure he knew exactly what he was what he and the team knew what they were doing when they hit fun.
12:30
They're like, "This is the perfect thing to launch right now.
12:34
The timeline is so primed to to love this." But uh there are Yeah.
12:39
Tyler, what's your take on all this?
12:39
Um I mean I think open definitely has the resources to do this kind of thing.
12:43
Like if you look at the the actual model Jim it's they used the 27 billion parameter model which is like >> I mean that's tiny compared to like what Frontier models are now right like GBT4 was >> over a trillion.
12:53
It was done pre-training in 2022. >> Yeah.
12:57
>> Um like they could throw a couple people at this.
12:59
They they just started that new like physics team that's just doing like physics models. >> Yeah. Open for science. >> Yeah.
13:06
So I I think this is like totally reasonable in it's in their capabilities. >> Yes. 100%.
13:10
>> Yes. 100%. Um I I think it just uh like clearly it's a race even if this is just like positive PR like hey AI is good for the world like every foundation lab should be competing in that because that's the thing that when you get on
13:23
Capitol Hill and somebody's saying like why are you using all the electricity you say well look at all my press releases about AI and science and cancer uh and it's good to point to like it's just good comms and so uh you want to be putting the points on the board first. This is yeah and this is why again
13:37
This is yeah and this is why again OpenAI has gotten so much push back on both Sora and the erotica announcement because only a month ago Sam was saying y >> uh if I don't get enough compute we'll have to choose between curing cancer >> and uh free education and I I don't want to have to make that call. >> Yeah.
13:55
Uh Tyler I I I think that this uh development updates me in two important ways.
14:02
One is that uh it's a vindication of Rune's concept of like text is the universal interface.
14:07
Like if you look at the actual model, it's a it's an LLM and basically as I understand it, the the data that went in is basically for every cell, they have just a bag of words that are just like every gene that's expressed ranked from most expressed to least expressed.
14:26
And so it's basically just like a text representation of a cell, which is interesting because I think when a lot of people thought about like creating a model of the human body, you you you just go to like a 3D model or something. I I don't know.
14:40
I I I don't certainly go just to text. >> Uh yeah.
14:44
I mean I I don't know if it's is it literally text or it's they just it's like tokens, right?
14:48
And then it's like okay, DNA is already in tokens.
14:51
It's just a string of numbers. >> Yeah.
14:53
>> It's not that different. >> Yeah, I guess. >> Yeah.
14:55
I mean, I don't think this like this is definitely a vindication.
14:57
I don't think this like was that big of like an update though, right?
14:59
This is anyone who's like sufficiently AGI is like, "Yeah, obviously scale things up.
15:05
We're going to get, you know, big >> well well that's number two.
15:06
So number one is that um you don't need some sort of like new data primitive.
15:11
You can actually just use tokens.
15:13
can use text to understand a cell and then find a connection between biological systems based on just a pure textual representation of the cell and the interaction.
15:26
Uh so I think that's that's an interesting vindication for like text is a universal interface not just in coding or text or knowledge retrieval or any of these other things.
15:35
It's also applying to bio.
15:37
You can take bio transform it into text and then do interesting things. What do you think? >> Yeah.
15:41
>> Yeah. I mean we so like if you look at AlphaFold Alphaold was like not that crazy of a new architecture right it was something we've already been doing uh on gaming stuff like this then you apply to bio so I think I'm not that surprised to see like >> architectures that we've already have been working well in other like areas
15:57
you apply them to bio and it works like yeah because it's not like >> those architectures are are like fairly general it's not like just for gaming or just for like text yeah right >> yeah yeah I want to talk about scale but first I want to talk about scaling your streams with one live stream 30 plus destinations reream, multiream, and reach your audience where they are. Uh, so yeah,
16:14
Uh, so yeah, the the second takeaway for me is that like scaling laws apply to useful AI applications in bio because that's if you actually go to the GitHub, you can see that they ran this uh this model on something like I think like four billion parameters or something and then they scaled this up to 27 billion.
16:30
It might have even been smaller model.
16:31
And so there's an interesting like they're they're scaling up and it's and it's, you know, qualitatively better system.
16:40
And so yeah, I don't know. >> Yeah.
16:42
I mean, when I see stuff like this, I I kind of I'm wondering what people who are like, oh, this is a top like we're going to crash any moment.
16:48
Like how do they respond to this, right?
16:50
Because this is exactly what everyone is like, oh, like, oh, it's just, you know, it predicts the next token.
16:54
Like, yeah, it's slightly better at coding.
16:56
It'll make someone 5% more effective, but like this is not going to be substantial change.
17:01
And like inevitably it's going to collapse, but then you see stuff like this where like obviously this can keep growing. >> Yep.
17:07
>> In 5 10 years this will be >> Yeah.
17:09
I think I think it's incredible but it but it's it's more tools for humans, right?
17:13
like this is human machine collaboration obviously like the key insight but but still we're at a point where this isn't like an AI 2027 scenario where you see the machine running you know has a novel insight runs away with it you know fully robotic uh >> lights out lab uh and then and then it just spits out a drug at the other end.
17:36
So I think like I think it's absolutely incredible but and uh and an an amazing proof point for the potential for the technology but >> based on this is exactly what you want to see based on all the promises we've been getting for the last decade. >> Yeah.
17:53
>> So I think I think like Google deserves a huge pat on the back >> but everyone the the industry should be breathing a sigh of relief being like okay we're actually doing what we've been saying we were going to do. >> Yeah I I I agree.
18:03
>> Yeah I I I agree. I'm just saying like if you listen to what super AGI people uh have been saying like we're just right on track with everything they say right >> like if you look at even if you look at like uh compute budgets like we're now
18:18
seeing there was the 2 gawatt data center being planned just yesterday coming on the show today that's I think that was ahead of like what a lot of people were saying >> even like Leopold or AGI 2027 like that's ahead of of what they were saying. Yeah. So like I don't see I mean Yeah.
18:29
So like I don't see I mean like everything it's not like we're just spending and then like we're not getting anything in return like this. >> Yeah.
18:38
I think the bare case is is for like a fast takeoff scenario I suppose is something like is this the chat GPT moment for AI and bio or is this the 2005 DARPA grand challenge that took 20 years until Whimos were on the street.
18:54
like there could be a rate limiting cycle time to actually testing these drugs in the wet lab going through the FDA approval process.
19:02
So even if the AI systems get better and better, how do you reinforcement learn on will the FDA approve this drug or not?
19:11
Like you have to actually test it.
19:15
>> Chad says got to give Martin Crowley a ring. We actually should. We should hit him up.
19:18
See if he wants to jump on >> right now. >> Yeah. Tyler figure it out. Let me I'll I'll DM him.
19:24
>> Um anyways, we can >> Is he live?
19:28
>> He's probably live, but he I'm sure he's down to to uh to co-stream uh going on uh there's a post here from G Fodor uh who says utterly totally destroyed worldview good riddance.
19:44
He's uh quoting a a post uh featuring a line from Yan Lun who says, "It's neither optimistic nor pessimistic.
19:54
It's just being real and stating what is the case.
19:56
It's a system with gigantic memory and retrieval ability, not a system that can invent solutions to new problems."
20:02
And then uh he shares Sundar's post.
20:05
So, um, anyways, uh, humans are somewhat of a of a a system with a gigantic memory and retrieval ability.
20:15
Uh, and yet we find ways to invent solutions to new problems.
20:21
So, uh, big moment and, uh, it's a it's a win win for the world. >> Yeah.
20:27
>> My question is like, okay, if can can Google whatever the process that led to this discovery Can Google produce thousands of these, right?
20:41
Can they >> It does seem like they're accelerating. Yeah. >> Yeah. Yeah.
20:44
Like I just imagine I just imagine at some point like a year from now is Sundar still sharing these breakthroughs or are they just kind of going >> taking them through FDA approvals? I don't know.
20:53
What do you think, Tyler?
20:55
>> Uh I mean I assume everything that we have seen in AI like says that like yeah they can keep scaling.
21:01
It's only 27 billion parameters.
21:02
billion parameters. scale it up to a trillion like >> you would think everything that that we've seen says that yeah this will give us more >> um yeah >> even if it's not like a novel insight if it's f like speeding things up way you know way faster than like what does it
21:16
really matter >> yeah just a super drug screener that allows you like a first pass of ideas like that's super valuable um I do wonder if there will be a data wall there's obviously no compute wall because you can just train a 27 billion model like that was not the gating factor here it. I I do wonder if they're
21:31
factor here it. I I do wonder if they're somewhat data limited on like you know actually indexing the every cell every cancer type all of that and what that cycle time looks like because we've seen with all the data brokers like the scale AI the merors like those companies have
21:49
been able to scale extremely quickly to do knowledge retrieval tasks and and put tons and tons of new you know answers to questions in the training set like what is the mercur scale AI of AI bio like >> yeah like I I don't know if there is one yet which just tells us that like this is still pretty early innings. Yeah, it
22:07
Yeah, it seems >> Yeah, it's interesting to see these people that that were hyper bearish transformer models like this this Brett Hall post.
22:14
He says once you understand the transformer >> as Lanni does and dare say I do and show on my website there you can literally see through what it does recombination and prediction. It's not creativity.
22:23
So it cannot possibly generate new explanations and we see that with all LLMs.
22:29
It's so funny because I like personally in my life when it comes to marketing and building brands and things like that, >> I think of creativity as recombination and prediction, which is like if you put these two ideas together, >> it could be very cool.
22:44
And I would give you the example of like this this uh we we just got this these rugby shirts in that we that we made that I'm very excited about.
22:51
And it's like there's not a single new concept here.
22:53
just kind of taking like a rugby shirt, which looks cool, making it kind of look like it's part of some Formula 1 team, and then putting our our our partners' logos on it >> and yet feels >> fresh.
23:06
This is the the Gwen take where it's like even if you can't just like prompt, give me a new uh idea for a joke or something, you can just, you know, brute force it by just giving it two random ideas and then saying connect them and then you can ask if it's interesting.
23:20
Then you just compound that and then eventually you'll get you're like brute forcing creativity. Yeah. Yeah. Yeah.
23:24
You're just kind of like indexing everything, ranking them against each other.
23:27
I mean, we know that that like this works.
23:28
This has worked in the Tik Tok algorithm.
23:30
Uh and yeah, just finding these correlations.
23:33
Um seems seems really valuable.
23:35
The question is just like yeah, how how how long will it be until uh the first FDA approved AI generated cancer drug hits the market?
23:47
like are there because the the the the the bare case to like the fast takeoff of the of the compute buildout is that at some at some point you tap all the debt, you tap all of the capital markets, you hit some sort of like you know fundamental law of physics around how fast you can move sand around the world and turn it into silicon or how fast you can spin up a new power plant or new uh new nuclear power plant.
24:09
Like these some of these things just take time and regulation like the the like there is the chance that this stuff gets regulated to the point where there's no fast takeoff.
24:24
Like we should have seen a fast takeoff in nuclear energy production once we figured out how great nuclear energy was and we just kind of regulated it out of existence.
24:32
The NRC stopped approving stuff and we just didn't see nuclear energy production.
24:38
Like you can see it's it's exponential and then it's sigmoidal because it just turns into an S-curve because we just said, "Yeah, we're actually good." >> Yeah.
24:46
I mean that that's kind of what Periodic Labs is working on, right?
24:47
It's like the the lab in the loop of the LM or like RL step basically.
24:53
>> Um >> but we're just like still so far from like actually tapping all the energy.
24:57
Like we're not close to that at all.
24:58
There's like >> no I'll keep saying it, but like% isn't like half >> a tiny amount.
25:03
We're not, >> you know, we could like literally today if if we wanted to, we could like double natural gas like very easily. It's just regulation.
25:10
Y >> and it's like there's no incentive to do that right now because we're still not that close.
25:13
It's like, oh, we're spending so much money, but like yeah, >> I mean, we could be spending way more.
25:16
I want to see way more spend. >> I No, I agree. I agree.
25:19
>> I No, I agree. I agree. I I really wonder like how quickly will Google pull the trigger on uh okay spend 10 times as much compute or 10 times as much energy on this because this uh cell to sentence scale 27b this feels like they could have just put
25:35
it on the ramp card like in terms of like the actual like compute budget how much weight went into that training run versus like the uh Gemini 3 that's launching like it feels like do you see the cut off date for Gemini 3 is like is like uh October of 25. And so it feels
25:48
And so it feels like that was a much bigger run.
25:54
Probably maybe a billion dollars of of compute went into it or billion dollars of spend went into it.
25:58
Like it's a very expensive project.
25:59
Uh but they have, you know, a bunch of applications with it.
26:03
They're selling it as an API.
26:03
Like they they can underwrite it very clearly on an ROI basis.
26:07
This is much harder to underwrite right now, but it's going to get less and less fuzzy as they as they actually see, okay, the things that we're generating, the ideas that we're generating, yeah, they're actually making it through FDA approval. So, um, exciting.
26:20
Very, yeah, very cool stuff.
26:23
And obviously just amazing for the Google brand.
26:24
Like, um, doesn't quite fit with the concept of, uh, organizing the world's information.
26:30
I guess you're organizing the cellular information.
26:31
Um, it seems like a little bit of a >> side new information.
26:37
>> Yeah, they organiz knowledge that existed that was discovered and now they're organizing it. >> Yeah.
26:43
>> Their mission is to organize the world's information and make it useful. >> Oh, make it useful.
26:46
Yeah, this is this is extremely on brand for them then made useful.
26:51
>> Well, let me tell you about privy wallet infrastructure for every bank.
26:53
Privy helps you makes it easy to build on crypto rail securely spin up white label wallets, sign transactions and integrate onchain infrastructure all through one simple API.
27:01
Um >> so the timeline is uh yeah still in turmoil about this a little bit but mostly everyone's just like this is awesome.
27:09
Uh, Greg comrade comrade.
27:13
Yeah, botching the pronunciation.
27:16
>> Comrade Greg >> says, "Someone needs to turn each TB TPN card into a physical collectible.
27:20
Slowly getting activation energy for this to start taking pre-orders. Limited to 10 prints.
27:26
First edition reserved for the subject.
27:28
All profits donated to TBPN cause of choice."
27:32
Uh, he says he's pulling the trigger on it. This is cool.
27:35
Uh, I think we were going to do this.
27:37
Uh, so maybe >> we were texting about this over the weekend and I wasn't sure about like how what the right way to handle like card printing.
27:44
They're very like news driven.
27:46
Uh, but I love that this just kind of organically happened.
27:48
>> I think we should do rookie cards for people that raise seed round. >> Yeah.
27:52
First time you get on the show. I don't know.
27:54
I'd say it's fun to print them all honestly.
27:56
Um, >> but I'm sure Greg will figure out like where the right where the demand is.
28:02
Like uh do the big people want their cards or is it more like the seedstage founder?
28:07
Is it more like the wedding announcements?
28:09
>> Do they want signed cards? >> Yeah.
28:11
>> Do they want uh like a piece of merch like you know uh on like some some uh trade, you know, actual trading cards, they'll put like a piece of the jersey fabric in them. >> Oh, really?
28:21
I've never heard I've never seen that.
28:22
I I feel like you'd want your own card if you did like a wedding announcement, but who would really want to collect a wedding announcement that's not for their own wedding?
28:29
Like that seems kind of like a niche thing.
28:31
That'd be kind of like what what are you doing?
28:35
>> Depends on the wedding.
28:36
>> Anyway, let me tell you about Cognition.
28:37
They're the makers of Devon, the AI software engineer.
28:39
Crush your backlog with your personal AI engineering team.
28:44
Um, and he said he he actually loaded the cards up into uh a a system.
28:46
He says, "Up polish is up next.
28:49
Going to mimic TVPN website style."
28:51
Thank you for your service, Greg.
28:54
You've been crushing this. This is very, very cool.
28:56
Uh, I would love to see how you go about actually getting these printed.
28:59
Um, you can go really crazy and actually print like real uh, trading cards.
29:03
Like there's there are systems and and companies that do that.
29:07
I was thinking that um it's possible to just go to Walmart and print like wallet size photos of things and they're like 30 cents I think for like a set of four and you just need to cut them out and people put them in the wallet photos and it's a photo quality printer and you could just take that and then find a sleeve that fits it and for like you know a couple cents be like making these pretty easily.
29:31
So there's a lot of different ways to uh to solve the problem but I'm excited to see where it goes.
29:34
Will Manitis says, "Internesting cultural indicator that Volvo quietly rolled out armored versions of their family cars direct to consumer over the summer. Probably nothing." >> Direct to consumer.
29:46
>> You can just buy an armored Volvo now.
29:49
>> It's pretty remarkable.
29:50
>> I almost bought an armored uh G Wagon back in the day that my friend Blake had.
29:55
It was the It was the former president of Kazakhstan's like actual like presidential like car that he'd driven around that that my buddy Blake had imported from Kazakhstan.
30:04
I really just wanted to own, you know, as a as a as a Borat enthusiast.
30:08
I really just I just really wanted to own that car.
30:13
>> I didn't pull the trigger because it's so it's got so much metal in it that it's like three times as heavy. >> Yeah.
30:18
The doors are massive >> and so it's just like not actually >> it must get like 2 miles to the gallon. Yeah. It must be so insane.
30:23
and you're just filling up like, "Oh, I'm driving from Malibu to LA. Got to stop for gas." Full tank. >> Full tanks. Full tank. Yeah. >> Rough.
30:31
>> Um, uh, >> let me tell you about Figma.
30:33
Think bigger, build faster.
30:34
Figma helps design development teams build great products together.
30:37
Uh, President Trump says that India's Prime Minister Modi has agreed to stop buying Russian oil. Wow. >> Trade deal. >> Trade deal. Yeah.
30:45
The the the geopolitical trade deals are all over the place right now.
30:48
There was also news that uh we're Poly Market said we're officially in a trade war with China.
30:52
Um I'm not exactly sure how that's defined.
30:56
>> There was a market on that >> I guess. Yeah.
30:58
>> Well, Trump came out and said yesterday we are in a trade war with China. >> Yeah.
31:02
So that that's the that's the end of that. >> that solves it.
31:06
>> But uh who knows these these trade wars can be very short and uh you know I guess uh India is playing ball hanging out with Daario from anthropic and hanging out with Trump and no longer hanging out with Putin and buying oil.
31:18
Um, the how did I get how did our family get so rich meme is going viral around the founders fund names.
31:24
Dad read the smearillion and bought a lot of domain names.
31:29
Uh, Sebastian uh cali Caleri says move quickly.
31:35
There aren't that many good Lord of the Rings names left to use.
31:39
Arthur McWater says he's camping the following.
31:41
Mordor, Lethoran, Merkwood, Numor.
31:44
I don't think >> Valenor Valenor has been used.
31:47
Valenor is an incubator I believe.
31:49
Amand >> I didn't know Amand >> Aman >> is that? No, there's no way.
31:55
>> Did Peter Teal back Aman Giri?
31:58
>> Um Joe Londale quoted the announcement yesterday uh that we shared around Arabore first getting approved.
32:03
He said finance done right enables massive job creation and wealth for a civilization and compliments builders like universities or media building a new with today's innovation frontier can upgrade how these areas work for all.
32:18
I'm just the founding board member and early investor in Arabore with HVC proud to support Palmer and his co-founders.
32:25
Yeah, I always liked uh Shadowfax.
32:25
If you were going to start a Founders Fund backed Helloax competitor or docuign competitor using Shadowfax.
32:33
Wait, have you not seen Lord of the Rings?
32:38
>> I have seen Lord of the Rings.
32:40
>> Do you know who Shadowfax is? >> No.
32:42
>> Oh, he's Gandalf's horse. >> Oh, okay.
32:44
Yeah, that's it's coming back.
32:46
>> And I think Shadowfax is a good name for like a docu competitor. Yeah. >> Yeah. A horse. A stallion.
32:51
>> A real stallion of a company.
32:51
but also something that could be related to the fax machine, which would be funny.
32:55
Uh anyway, congrats to the the the team over at Arabore.
33:00
Palmer went on Rogan as well.
33:02
So, there's a three-hour Palmer Lucky Joe Rogan episode uh waiting in your podcast player of choice if you want to go listen to that.
33:09
>> Uh Peg God says, "Welcome back, Command Economies."
33:12
And uh the classic template, "Get ready to learn communism, buddy."
33:17
which is that apparently the Trump admin is planning to set floor prices across a range of industries to combat market manipulation by China.
33:25
According to Scott Bessant, >> it's very odd that they're pulling that out as a tool instead of just focusing on tariffs. I don't know.
33:32
It seems like there's so many other ways to uh to deal with a trade war than just >> they like to >> they're like a they're like a DJ, you know, like sometimes a DJ's up there and like really turning all the different knobs.
33:45
Sometimes you just let the music play.
33:47
You know, >> that is a very funny metaphor for David Solomon.
33:50
>> Like the knobs are there, right? You can >> Yeah.
33:53
>> Some people just choose to use them more than that.
33:55
>> You should get They should get David Solomon in there. He's already a DJ.
33:58
He's already turning the knobs.
33:58
Why not have him turn the knobs of the global economy? >> Yeah.
34:01
I just I I do uh of course we never discuss politics on the show, but I do worry about >> uh the opposite end of the political spectrum being heavily inspired by these actions and then it just becomes the norm that you just have, >> you know, obscene levels of of government intervention. >> Yes. Yes. All time.
34:21
I I always go back to like trying to benchmark the level of government intervention in America versus other competitive countries.
34:28
So with the Intel thing, this happened the same thing as like what was it five or 10% stake that the US government took and 10 and a lot of people were like this is like communism, this is command and control.
34:41
Um but uh if you look at the uh the I believe TSMC was basically like 5050 with the Taiwanese government.
34:49
And so even now there's a lot of other players in the semiconductor supply chain internationally that have a higher proportion of government influence in their uh cap table effectively.
34:59
And so I'm not in total like, oh wow, we're like doing worse than what's going on in South Korea or >> I think Yeah.
35:07
And my my framework is that as much as Trump and Xi beef, >> I believe that Trump respects G. >> Yeah.
35:16
And I mean, America's bailed out companies successfully before.
35:18
America has come in been the lender of last resort for stronger companies like Intel and and turned the company around and then sold their position and gotten out entirely like uh government motors like the US does not you know fully the the the US government never like fully nationalized the the GM ecosystem uh even though they did bail them out at one point.
35:42
Um Joe Eisenthal says all businesses are banks except for banks.
35:47
Banks are media companies. Wow.
35:47
He said this back in 2018. That's crazy. From Earth.
35:53
Uh this is because Mi Mr.
35:53
Beast has filed a trademark to launch his own bank.
35:57
The organization will will be called Mr. Beast Financial.
35:58
We talked a little bit about this yesterday.
36:00
Um I'm I'm not sure what he will uh what he will wind up being, but we got to have him on the show.
36:09
>> What are you thinking?
36:10
>> I'm just laughing at the chat.
36:10
Sam Sam Schffer says is asking about uh real competitors to the USA at the country level.
36:17
And the chat says China, >> Russia, >> the EU, Guatemia, >> yeah, well, let me tell you about Vanta automate compliance manage risk, proof trust continuously.
36:31
Vantage trust management platform takes the manual work out of your security and compliance process and replaces with continuous automation whether you're framework or >> manage.
36:39
Did you see this consumer reports report uh Paris uh did an investigation? Yes.
36:46
uh did 60 plus lab tests on leading protein supplements uh and found that a quite a lot of them had uh high levels of lead.
36:58
Um this wasn't surprising to me. >> Okay.
37:04
>> And the reason for that is that a lot of different food has high levels of lead.
37:10
Even even food that uh would appear to be like quote natural, right?
37:13
So like dark chocolate is a good example of this.
37:18
A lot of dark chocolate brands have >> high levels of lead.
37:20
It's something that um >> that you should be looking out for.
37:24
Like not every not every brand is created equally.
37:28
>> Um and I think that people were uh particularly scared about Hule uh which the Hule Black uh had >> kind of an ominous name. >> Yeah. Kind of an ominous name.
37:37
Um and of course the timeline was having a lot of fun with it yesterday.
37:42
people being like, "Yeah, I could tell you've been using a lot of healed black over the last year.
37:49
>> You're just deranged from all the lead poisoning." >> Yeah. >> Yeah.
37:53
>> Um, >> we'll stick to the Optimum Nutrition, the Transparent Labs. Good.
37:55
Good to see Transparent Labs being exactly what their namesake implies, transparent.
37:59
It seems like they've they've done very well with the the lead exposure.
38:03
Um, is lead is lead more of a cause for concern than microlastics?
38:09
than microlastics? I feel like what was it last year Nat Freiedman with the plastic list like really shift the shifted the conversation to uh levels of microlastics and I think people kind of like stopped paying attention to lead but
38:22
>> it's all important you want to be uh you want to be aware of it but uh like the way that you need to think about it is there's basically healthy levels >> there's um certified levels >> and then there's like legal levels right and like healthy levels >> like ideally it's zero Right. Like lead
38:39
Like lead I'm actually I'm sure somebody some skits I will be like actually >> I like a little bit >> a little bit of lead is actually good for you >> because it makes you more aggressive. Right. >> Yeah. Yeah.
38:49
>> Doesn't make you a little crazy. >> Yeah.
38:50
If you want to have a strong Q4 >> Exactly.
38:53
>> You know up the lead and then detox in Q1. >> Um new year, new you.
38:56
Uh but yeah, you I just think about it as like healthy uh like certified as in like you can you can um go out and get various certifications like that generally show that your product is like healthy or clean.
39:10
Y >> and then there's like the legal levels which like none of these >> none of these brands were above the legal limit. >> Oh really?
39:17
None of them even even the hule was just uh was just not recommended.
39:21
>> So if you're a food company in America you can legally have quite a significant amount of lead.
39:26
Um, but anyways, I think you could do these for like almost every category of food and get some pretty shocking results. >> Uh, yeah.
39:33
I mean, I I think with plastics and lead, like the thing is, uh, we've kind of ran the AB test with lead where it's like in the kind of early '7s, I think that was kind of the peak of like leaded gas.
39:42
And then what you saw is like >> 15 20 years later, you basically saw the peak of like crime. >> Yeah.
39:49
>> Because it's like you grow up as a kid, you're inhaling all this lead and then you like it makes you way crazier.
39:52
But with plastic, we haven't really run the AB test cuz like there's kind of just always been like a buildup of plastic and >> there's like no one who like doesn't have a lot of plastic.
40:01
>> So, who knows what happens when we take plastic out of >> the potentially skitso alpha is that lead paint could actually be positive because it reduces the like of EMF in a room.
40:12
If you have like a bunch of lead paint on the walls, it's actually blocking some of that. >> Yeah.
40:16
All the plastic in your system is just bouncing the lead particles off, too.
40:20
So, you're just immune to lead poisoning because you have so much plastic in there.
40:23
You wrapped your entire body in plastic.
40:25
Uh, this went incredibly viral. 17,000 likes. Um, I'm surprised.
40:31
>> Yeah, because I think most people that look through this list are thinking, "Okay, I've had at least one of these in the last three months.
40:36
>> I have these every single day basically."
40:38
Um, yeah, pretty pretty wild.
40:41
And it's just I mean yeah it's it's uh a lot of the brands are going to push back because not every you know you're not going to see that if if they ran one lab test per >> product if you tested >> another product that has that was made in in another batch you could see wildly different levels.
41:01
>> Yeah the batches and also uh up the ingredient supply chain like a lot of times the the lead comes from two steps or three steps deeper in the supply chain.
41:10
This happened in Soilent.
41:10
We had some article that was like it was it was very funnily worded.
41:13
It was like Soilent's like so super heavy metal or something like that because they were tested it for heavy metals and they found something similar to like an early batch and uh it was like two steps up the chain.
41:24
Like one of the one of our suppliers, one of their suppliers had like changed their supplier or something like that, the lead company.
41:33
>> Uh Bobby wants to see a lead gong in the Ultra Dome >> for aggroaxing.
41:36
Yeah, let the let the lead.
41:40
>> I would hit a lead gong for Absurd, a new company out of Y Cominator >> that makes absurd AI launch videos turning raw.
41:47
This sounds >> uh this sounds kind of like it was written by Chad GBT.
41:53
>> Well, let's pull this video up, but first let me tell you about graphite.
41:54
dev code review for the age of AI.
41:56
Graphite helps teams on GitHub higher quality software faster.
42:01
You can get started for free.
42:02
And now, let's watch this launch.
42:05
>> Everyone wants to be different.
42:07
It starts with a familiar feeling to feel stuck.
42:13
>> The circular motif here was really cool.
42:15
And the uh and the piano like changing out what's around it.
42:18
I think that's a very interesting use of AI.
42:22
>> Can we go fuller screen?
42:23
>> Absurdity isn't designed in a studio. >> Fighting a robot. There we go.
42:27
>> Discovered in obsession born from truth so raw it feels like math to everyone else. This is cool.
42:37
>> A beautiful disregard for your own limits.
42:41
A truth your body understands before your mind.
42:46
>> This really sounds like Morgan Freeman.
42:47
A >> beautiful illogical.
42:49
>> Is this not stealing his likeness?
42:51
>> Your story only you >> I guess it's just like an homage.
42:56
>> Is Sam Shaffer in here?
42:56
We need Sam Sheffer's comments on this for sure.
43:01
>> He was in here before.
43:02
>> Truth >> is the most Just interesting when you >> Yeah.
43:08
>> When it's when it's difficult to make something >> Mhm. >> it's valuable. Yep.
43:13
>> And when anyone can have it instantly. >> Yep.
43:16
>> It's absolutely worthless. >> Yeah.
43:18
>> Uh so so again, I before I would judge this, I'd want to see >> I want to see 10 actual launch videos for 10 actual companies that you've worked with.
43:28
My guess is that um my guess is that uh these are might work as in they might get views but I don't think they will build your brand at all. >> Yeah.
43:42
I mean it I would assume it follows the same curve as like the first studio Gibli that I posted got like thousands of likes and it was like super basic.
43:50
I just took a screenshot from Oppenheimer of Einstein talking to Oenheimer from the movie, Giblied it and just posted that and it got thousands of likes because people were like, "Oh, this is like something I like and I know and I like Oppenheimer and I like Gibli."
44:03
So like, but if I posted that today, it would be a total flop because like it's been done so many times, right? >> Why are you laughing?
44:10
>> I just think it's funny because we should actually test it.
44:12
You should post like just a Gibbly image and be like >> just like Gibbly of that horse and be like, "Oh, wow."
44:18
you know, or whatever whatever memes going viral that day, you could you could gibly it.
44:22
And I mean, there were so many Giblies that went viral where it was just like every iconic image for the past few years.
44:27
It was like a picture of JD Vance, Gibbleied, okay, viral.
44:31
A picture of of Donald Trump, viral.
44:34
Like picture of Michael Jordan dunking, viral.
44:36
And but now it's like we have all seen it and it's been commoditized, so it doesn't it doesn't it like it ran its course.
44:41
And that's like the the the worry with this is that like the first AI launch video goes viral and then it like gets less and less.
44:50
But it does seem like they're building more of a SAS-like product on top of AI image gen.
44:55
And so you know the agents kind of plan, generate, edit every scene.
44:59
So you're kind of like you can still come with an idea.
45:01
Like the thing that stuck out to me in that in that uh video there were two things.
45:05
One was the the circle motif.
45:07
Hitting the circle again and again and again was cool.
45:12
And at the end, the very blurry images of the boxers, like that was an interesting artistic style. Yeah.
45:16
interesting artistic style. Yeah. And so right now, if I wanted to generate a video, if I come up with an idea of like, okay, let's, you know, let's do a horse themed launch video, >> it would be it would be sort of a hassle for me to take an image of the horse
45:33
like and come up with the concept of like the horse running and then each frame of the horse running where it's in a different, you know, uh, different segment of its walk cycle or its run cycle or it stride, >> each frame is transformed into a different art style or different world. It would be really hard to puppeteer
45:50
It would be really hard to puppeteer that.
45:52
I'd be copy and pasting the prompt again and again again.
45:54
And having an agent or having a system that kind of helps me pipeline that stuff where it's like, okay, I want to make you've seen those videos on Instagram that are like >> uh it's a car and it keeps and it just flips through like the the the headlights one after another and shows you like 50 cars, but it's always the image is always focused on the headlights.
46:13
So, let's Jack Cohen uh over at General Catalyst just shared uh a a video they did for a >> called Zavo. So, let's pull it up.
46:23
>> I'm very interested in that. >> Um it's in the chat. It's in the X chat.
46:28
We can try and put it in the timeline.
46:29
>> I think you're going to absolutely >> While we uh while we pull that up, let me tell you about Julius, the AI data analyst for everyone.
46:36
Chat with your data and get expert level insights in seconds.
46:40
Uh, do we have this video? >> Let's play it.
46:43
Oh, >> this isn't a restaurant. It's a battlefield.
46:48
Every ticket is a ticking clock.
46:51
>> Your best souls medium are falling.
46:55
>> Okay, meet Greg, your numbers wizard.
46:57
He'll tell you the sea bass special is secretly costing you a fortune.
46:59
But the truffle fries special as well.
47:04
That's pretty impressive.
47:06
>> He knew you'd run out of pino noir by 8:00 p. m.
47:07
on Friday and already reordered it yesterday.
47:09
He knows when your night will go from 0 to 60.
47:13
>> So, is this vertical sass for restaurants?
47:15
>> And then there's Maria. >> That's pretty good.
47:16
If so, >> she turns visitors into regulars by suggesting the perfect topshelf tequila for their favorite spicy margarita.
47:24
She'll find hidden patterns in your audience and understand exactly what they want.
47:28
>> That looks That looks like post that looks like after. Wow, I hate this.
47:36
>> See, this this looks like not AI generated.
47:38
This looks like they layered in their actual product, but maybe that's part of the workflow in the pipeline.
47:44
>> They need some drop shadow on that logo.
47:47
>> Uh it's a little white on white, a little low contrast, the boys now. But uh I don't know.
47:52
>> I think >> that is right now that's attention getting like right now if that pops into if you're a restaurant tour.
47:58
I I hope I'm getting this right.
48:00
That's that's it's a it's an ERP for for restaurants. >> Yeah.
48:06
So you you use that to run your restaurant.
48:08
That that's what I got from that. Is that is that correct?
48:09
Uh is >> the first agentic point of sale for restaurants and retail payments point of sale and AI agents in one platform to build the future of autonomous commerce >> over 400 businesses.
48:20
I already used Zava to accept payments and manage operations.
48:25
>> I mean it did deliver that message to me and it grabbed my attention a little bit.
48:28
So I don't know >> worth >> Yeah.
48:30
The opening scene was particularly bad but at the same time it was probably a good hook. Exactly.
48:36
>> But I think the hookiness of that will will dissipate.
48:38
I think you're correct on that.
48:40
Like like it will be like eventually you'll just scroll and be like, "Oh yeah, I've already seen like the AI slop.
48:44
I don't want to see that."
48:45
But like right now if you're the first like Toast does has is running an ad like that.
48:50
>> And so if you're the first company to be running that ad that grabs people in that way, like you will break through.
48:56
There's like there's like a window of alpha. I think I >> Yeah.
49:00
question is is uh if they had figured out a way to do this with the founder as like the key >> person and like some blend of AI plus the actual that's coming would be >> that's for sure coming.
49:14
I mean V3 launched yesterday and uh 3.
49:15
1 and uh and has like pretty phenomenal character consistency.
49:21
I was looking at some demos.
49:22
Uh, if you want to generate a bunch of video, go to fall generative media platform for developers.
49:27
The world's best generative image, video, and audio models all in one place.
49:32
Develop and fine-tune models with serverless GPUs and on demand clusters.
49:35
Um but uh I I agree with I agree with like all >> part of it part of it as an advertising enjoyer I like to like without having knowledge try to analy when I see a video when I see a campaign I'm trying to clock >> uh you know based on the stage of the company how much did they spend on this
49:53
for example like if Airbnb comes out with a launch video >> I feel pretty I like expect it to be incredible because it's Airbnb and they probably spent like half a million dollars like producing Apple ads when they get like, "Oh, Spike Jones is directing it." And you're like,
50:07
And you're like, "He does movies."
50:10
>> And then with with with seedstage companies, one thing I like is like being able to clock like, "Okay, how resourceful is this company?" >> Sure. Sure.
50:17
>> Uh I like being I like being able personally I enjoy being able to clock.
50:22
Okay, they clearly had to be scrappy here.
50:23
Like they spent like 10, 15, 20K maybe on this video.
50:29
>> But it's amazing because it's like really it's genuinely a really great idea. Mhm.
50:32
And so with absurd, I think like the the actual output generally looks great.
50:38
It doesn't but but I would say the the idea itself is not I don't know if it's like strong enough.
50:46
Basically, it's like >> white humanoid looking things in a restaurant.
50:51
But maybe maybe >> there were white humanoid looking things like robots.
50:55
>> I completely missed that.
50:55
Some it was very like it was very visually intense.
50:59
>> Um I Yeah, I agree with that. That's interesting.
51:02
Um, I think that there is a there uh there is a like a underrated side of using AI in a creative way that does show resourcefulness and it can show being on the cutting edge.
51:16
It's a way to tell your audience that you know how to puppeteer the models in unique and innovative ways that just dropped today.
51:27
today. We see this a lot with like justine Moore like a new model will drop and she'll be she'll have like converted her podcast into a cartoon with the correct lip syncing and that's not just a feature where you just click a button
51:40
and just like you have to actually wire up a few different systems and she usually will show underneath it's like I use this for the for the cartoon effect and then this for the for the video and then I use flow or something else. And
51:50
And so um >> there is a way to signal that you're sort of like tapped into all the different frontier capabilities because Sora is different has a different look than midjourney but you can start with a midjourney image and bring that into Sora and then you can use a different video or audio model or or mix the audio specially like I think that they added motion graphics on top that that the AI model didn't just oneshot those graphs.
52:20
Yeah, >> I think that that was after effects. >> Yeah.
52:22
So, the question is like with absurd, are they charging, >> let's say they're charging $10,000 for a launch video? Yep.
52:29
>> Are they spending energy, resources, timely, >> definitely not just like one shot, but but so today it's basically like a managed productized service business.
52:41
Um, and I've heard of another company, an an AI ad company who's effectively, I'm not gonna name them, but they're just doing like creative services work and they're selling it.
52:52
They're selling it like it's AI outputs. >> Yeah.
52:55
>> But it it ultimately is just them running an agency >> and >> and are they prompting a lot or are they actually filming stuff too or I mean you could do both.
53:04
>> They're doing everything.
53:04
It's basically it's it's it's an agency masquerading as like a as an AI company.
53:08
like a as an AI company. Yeah, there's still there's still something there where like >> is it I guess like zooming out like is it a better time to start a law firm from scratch where you can be AI native and you can just say from day one we
53:27
expect everyone at this firm to use AI as aggressively as possible or a creative agency and say hey we don't have we're not a thousand person organization that has workflows for you how we do things in in different, you know, digital products. And so we're not
53:42
And so we're not bought in or or we're not like stuck in our ways.
53:47
And so everyone we hire, we expect you to use every possible tool for the job. Yeah.
53:51
Like is it is it a better time than ever to start a creative agency because you can be AI native on day one as opposed to being like a legacy creative agency that then needs to go pull those tools? I don't know.
54:03
The challenge on the creative agency side is if you go and compete just on price, it's a really rough business because you get you get clients that are looking for cheap work and those clients tend to be really rough.
54:15
Like great companies understand the value of creative will invest in it.
54:20
They're not they're not just looking for the cheapest output.
54:21
They're looking for something great.
54:23
something great. And yeah uh on the legal side where you just there's there's certain amount of work let's say like nontransactional uh work that is you know purely corporate where if a new law firm emerged and they said >> we want to have your legal bill drop by
54:41
70% every month and we're going to do that with by properly leveraging these tools and if that's their initial edge I could see that working well because for a lot of just like day-to-day legal work that a company needs to do >> with external counsel. >> Um, you don't really care. All you care
54:57
>> Um, you don't really care.
54:57
All you care about is that it's done well.
54:58
It's not necessarily how it was done.
55:03
>> Let's stay on the legal example.
55:03
But first, let me tell you about Turbo Puffer.
55:05
Search every bite serverless vector in full text search build from first principles and object storage fast times cheaper and extremely scalable.
55:12
So, is there a good analogy between I'm graduating from law school, instead of going into the the big law firm world, I'm going to start a law firm that uses something like Harvey on day one and and really leans into the frontier of let's do as much as possible and set this and set this company up from day one to be AI native such that maybe our business model is different.
55:39
Maybe we're not so focused on billable hours or we have some sort of different different model that is enabled by AI.
55:45
We're still a bunch of lawyers but we're using AI very effectively.
55:49
Is that at all can we draw any analogies between that and the DTOC e-commerce era we're basically who had brands who said our secret how we're going up against Nestle how we're going up against Coca-Cola cutting out the middleman is we have Shopify >> and like SAS we have SAS and they don't.
56:06
Turns out Mark Mark Zuckerberg says, "Actually, I'm planning to be the middleman here.
56:10
I just >> It's a great take." >> Yeah.
56:13
I don't want to >> The money that you were going to spend on rent, well, I'm actually going to need you to spend twice as much with me.
56:21
>> So, do you think that Do you think that's how it plays out in the next generation of creative agencies?
56:25
I start a creative agency and I say, "I'm I'm cutting out the middleman of the photographer, the videographer.
56:35
I'm using AI tools from day one and I get a bunch of clients and I make a decent amount of money.
56:39
But over time, the platforms that actually act as the AI generators say, "We're the middleman actually and we're going to take the lion share of the profit."
56:48
What do you >> Yeah, it's a good question. It's >> tough, right? >> I don't know.
56:53
I think you always I in the creative world, I think you want to be charging based on value, not how much time goes into the work. Yeah. Right.
57:02
Because the best logo designers in the world, >> they might be if a startup comes to them and says, "Hey, we want a new logo." >> Yeah.
57:10
>> They might be able to >> do that in might it might take them weeks.
57:15
It might take them they might oneshot it, right?
57:16
They might sit down and just have this amazing creative spark.
57:19
And so if you're the best in the world, >> you you want to be paid for being the best in the world.
57:24
>> And so it doesn't necessarily matter how long it takes.
57:26
And so, uh, yeah, certain certainly on on, um, I wouldn't respond well to a pitch today from a agency that says, "Hey, you should work with us.
57:37
We're going to charge you >> 60% less than your current designer by using AI."
57:42
Like, that doesn't that doesn't that doesn't get me excited on the creative front. >> Yeah.
57:50
>> But maybe more so on on on legal. >> Yeah. I don't know.
57:55
It seems like it's like there's increasing value for true creativity.
58:01
Like when I think about the AI tools that are rolling out, I'm like I completely trust those in the hands of Gabe. Why at Mischief?
58:11
>> Like I think that like when someone who comes up with completely novel ideas just has an extra tool in the tool chest, there's going to be interesting stuff, but just acting as a wrapper around it is going to be a little bit tricky. >> Yeah. Yeah.
58:23
And the question so so going back to Absurd the YC company, >> there was a time when like the DTOC style of doing a website really hit.
58:33
Consumers would see it and you know these brands were like simple.
58:36
They had a distinct style, right?
58:37
Product on this sort of like colored background and consumers responded really well to it for a decent amount of time, right?
58:44
kind of the same period that we've been in like the cinematic launch video era. Yeah.
58:51
>> And then eventually people just stop responding that well to it because it it no longer says the sends the signal like this is a thoughtful person that's trying to stand out is just becomes you know everything becomes that way and then it became >> uh easier to get a response by going >> totally the other way and coming across like you were more almost vintage. Totally.
59:12
Yeah, it's very it's a very narrow it's a very narrow arbitrage.
59:17
>> Um >> but uh the timeline is in turmoil on threads because PayPal and Wise are taking shots at each other. Did you see this?
59:25
Uh PayPal said normalize sending money instead of memes and got 9,000 likes on threads.
59:31
A real ripper over there.
59:33
I didn't realize how that >> Conor Hayes Conor Hayes is cooking. >> He's cooking.
59:38
Brands are really going wild over there.
59:39
And Wise says, "Normalize sending money with transparent fees." Oh, dump session. It's very funny.
59:47
This is so like 2015 Twitter coded.
59:50
Like I feel like brands have not been >> I put this in.
59:53
It reminded me of when it was Coke and Pepsi saying like, "Wag me." >> Yeah. The metaverse.
59:58
Uh but >> hey friend, >> they're they're they're having fun for sure. >> Wow. This is great.
1:00:05
>> But first, let me tell you about Google AI Studio.
1:00:07
Fastest way from prompt to production with Gemini, chat with models, vibe code, monitor usage.
1:00:10
Uh, you can try Nano Banana.
1:00:14
You can talk to Gemini live and continue.
1:00:18
>> Speaking of ads, Ma Amar made potentially the product of the year.
1:00:22
We will be doing awards later this year.
1:00:26
She made a Chrome extension that hides all website content except ads. >> This is amazing. I love this so much. >> I need a link here.
1:00:34
I want to start running it. Yeah.
1:00:38
>> Um, >> it's remarkable the inverse ad blocker and some of these websites. Wow.
1:00:41
Some of these websites are so chocked full of ads.
1:00:45
I wonder which what website these are. >> Display ads.
1:00:50
>> Also, it looks like in German maybe or something. I don't know.
1:00:53
>> Um, did you uh did you hear that yesterday Paxos mistaken mistakenly minted 300 trillion of their stable coins?
1:01:05
I am not I did not follow that at all exactly what happened.
1:01:09
>> So uh stable coin uh issuers like Paxos and Circle and Tether they mint new stable coins. >> Okay.
1:01:17
>> Uh they had an internal error apparently that and they they b it sounds like it was a fat finger.
1:01:22
They they accidentally minted 300 trillion >> uh in their PYUSD uh their stable coin.
1:01:30
>> Did this like crash their market or something?
1:01:31
I wonder what happened here.
1:01:32
No, I think they were they they were uh Paxos immediately identified the error and burned the excess. Uh >> it's resolved.
1:01:39
Yeah, they said that there's uh no security breach. Customer funds are safe.
1:01:43
They've addressed the root cause.
1:01:44
But how fat of finger do you have to add?
1:01:46
Like I imagine like four extra zeros.
1:01:49
>> Maybe maybe the dev like fell.
1:01:51
>> It does feel like a cat on the keyboard. Yeah, it doesn't. It doesn't exactly.
1:01:56
And uh >> yeah, Joey says slight rounding error.
1:01:59
Don't ask about the dollar backing. Yeah.
1:02:00
So So people people are Like people are basically pushing back and be like, "Okay, like you were able to just create $300 trillion on chain." Yep.
1:02:09
>> Um what what was it backed by?
1:02:09
Is or or is it actually is there some sort of uh disconnect?
1:02:15
>> Maybe they bought 300 trillion.
1:02:17
>> Glad they glad they figured it out.
1:02:18
>> Maybe they bought 300 trillion of treasuries, you know.
1:02:22
>> Um well, before we move on, let me tell you about profound.
1:02:23
Get your brand mentioned in chat GBT.
1:02:24
Reach millions of consumers who are using AI to discover new products and brands.
1:02:28
Um, um, born in Texas in a Texas town of less than a thousand, writes the most iconic rock album of the past 50 years, quits music before its release, becomes electrical engineer for AMD, makes it to Sony's vice president of technical standards, helps create Blu-ray, doesn't elaborate.
1:02:48
>> What a wild uh what a wild life story.
1:02:48
A true like you can just do things.
1:02:51
It's never too late to like just completely switch gears.
1:02:55
This is James Williamson of the Stooges. I never knew this story. >> I I Me either.
1:03:00
I I don't I'm not even really familiar with who the Stooges are. Look at this guy. American guitarist.
1:03:05
What What's the biggest song from the Stooges?
1:03:09
>> We need to We need to uh convince the new the younger generation that this is actually the path that you want to go on in life.
1:03:15
Like go on a short but generational run as a musician and then go work in go work in tech.
1:03:21
>> Let's get 100 gs in open AI or something. That's the move. >> Uh breaking.
1:03:26
We got some massive news.
1:03:29
>> Hit get that gong ready.
1:03:29
Julius AI is now GDPR compliant.
1:03:33
Perfect timing with uh perfect timing with Europeans getting back from summer holidays. >> Oh, that's huge.
1:03:42
>> They can now get access to the AI data analyst that is Julius.
1:03:46
>> There's more Julius news in here some somewhere. I got to go deeper.
1:03:48
But uh >> Dean Ball says, "Could I like make a taxdeductible donation to the OpenAI nonprofit?"
1:03:55
you've been joking about. >> Curious. >> I think you can.
1:03:58
Uh, Alexander Burgerer says, "True story.
1:04:01
Openai once came up on the checkout screen of my Safeway as the local nonprofit to donate to."
1:04:09
>> I think that's that's got to be a joke, right?
1:04:11
>> Yeah, that's got to be a joke.
1:04:12
>> But maybe maybe it's just like a random they pick a random nonprofit or something. I don't know.
1:04:16
But, uh, yes, you should definitely donate to OpenAI nonprofit because they'll probably spin out another for-profit.
1:04:22
That's my that's my hottest take is that uh we're we're getting a >> well being a donor in the nonprofit does certainly does not guarantee you equity in spinouts.
1:04:34
>> No, but >> seen that with PT and Elon and others, >> but uh you it's probably a strong badge of honor.
1:04:40
You're on the Wikipedia page if you're a co-founder of the nonprofit, which is good.
1:04:45
>> What about if you're just a donor?
1:04:45
Yeah, you got to really uh you got a you got to post a screenshot of an email, screenshot of a receipt I donated before they invented AGI or something. >> Uh this is cool.
1:04:55
The mint is making some new coins showcasing innovation from each state.
1:04:59
There are four ones next year according to Shiel uh including Steve Jobs uh for California. They're doing Dr.
1:05:07
Norman Borlo Borlow >> the Cray one supercomputer for win.
1:05:14
>> Do you know who these guys are?
1:05:14
Steve Jobs of California and Minnesota with mobile refrigeration.
1:05:20
>> That's a huge breakthrough.
1:05:21
>> Let's give it up for mobile refrigeration.
1:05:22
>> I know the Cray supercomput and obviously Steve Jobs.
1:05:24
What a what a great coin.
1:05:27
>> How much do you I wonder I wonder Tyler, can you find out how many these of the Steve Jobs coins they're going to make because I feel like these things could instantly trade at moon >> like a hundred times. >> Yeah.
1:05:38
>> Uh yeah, I'll look that up.
1:05:38
And then also Norman Borla was an uh agronomist, agriculture guy, and he kind of did a lot of stuff that influenced the the green revolution.
1:05:47
>> He he he orders blog yesterday, which was uh very blackpilling.
1:06:04
Uh Run says, "Contra to all this, it's completely fine to go work in technology or anywhere for that matter and not really give a damn.
1:06:10
It is not from the benevolence of the SAS founder, the currency trader or GPT rapper that we expect our dinner, but from regard for their to their own self-interest."
1:06:21
And Reed says, "Wow, his website is very negative."
1:06:24
And Run says, "Feels like George Hotz is a zealot who has produced far less value than his clearly exceptional skills as an engineer should should suggest. But George H.
1:06:33
Hotz can do whatever he wants.
1:06:35
That's the beauty of of the free market.
1:06:37
He can go and and have his own thinker.
1:06:40
He can he can advocate for different things.
1:06:41
He can open source software.
1:06:41
Uh what uh Tyler, what how have you been reacting to the to the George HOTS news?
1:06:50
Do you think that you're uh you're you you expect that AGI will come from a corporation, therefore it's worth to it's worth it to be a corporate bagman?
1:06:58
>> I mean, I'm definitely proc corporation. Mhm.
1:07:01
>> Um but also I'm kind of a George Hart's truther, you know.
1:07:03
I I think he has a lot of good takes. >> He does.
1:07:06
He does have a lot of good takes.
1:07:07
>> Um he's pretty goated.
1:07:11
>> Yeah, I'm excited uh to see what he does next.
1:07:13
Um he's always working on a fun project.
1:07:15
Hopefully there's something new dropping soon.
1:07:17
>> But first, let me tell you about linear.
1:07:19
Linear is a purpose-built tool for planning and building products.
1:07:20
Meet the system for modern software development, >> streamline issues, projects, and product road maps.
1:07:26
Uh people were having a lot of conversations surrounding OpenAI's numbers from the Financial Times.
1:07:32
Of course, they have 800 million weekly active users.
1:07:36
5% of those are paying 40 million. Yes. >> 13 billion in AU ARR.
1:07:42
>> Uh which implies a $325 annual RPO or $27 a month per paying user.
1:07:49
>> Um and this post went uh pretty viral.
1:07:54
uh Google would turn 800 million users into 32 billion of revenue.
1:07:59
>> Is that just based on their current revenue and their current user base?
1:08:03
>> Because you have to think like Google's like so much more heavily monetized that the gap between 13 billion and 32 billion.
1:08:10
It's not as much as I feel like it should be based on how young Chad GBPT is.
1:08:16
Like they don't like you can't actually advertise on it yet.
1:08:18
Like they don't have an ads product.
1:08:20
And so, uh, it's pretty remarkable that they're monetizing at the rate that they are already.
1:08:26
That was my takeaway from this >> because Google's been grinding the the the ads product for 20 years now. >> Yeah. And >> Yeah.
1:08:35
And I think what what Malt was meaning to say is that he follows it up.
1:08:39
He says, "The tweet was meant to say, you can run this into a profitable business." Sure.
1:08:43
>> I expect value per user is going to be higher than Google. >> Oh, yeah.
1:08:47
because the the OpenAI is making 13 billion in revenue but burning like 20, right?
1:08:52
And so you add those together, you get about 33 billion in revenue to to fully offset the burn which is kind of like exactly Google's monetization rate.
1:09:01
And so if you monetize the time and the products that like similarly you get to break even pretty quickly. >> Yeah.
1:09:08
And and uh OpenAI is projecting to get to 100 billion of revenue.
1:09:14
>> The projections are crazy.
1:09:14
We we should go through those.
1:09:16
Uh Doomslide says, "Extremely polite thread by epoch.
1:09:20
OpenAI's projection implies it gobbling up about half of all software revenue by 2028."
1:09:25
We are going to be join joined by uh Mark Beni off in just a few minutes who of course uh captures a lot of that software revenue already.
1:09:31
Uh >> but gobbles up gobble is gobbles up the right way to do it or is it creating new software spend right?
1:09:39
If uh if openai can uh build consumer agents if they can monetize uh >> product discovery, right, which they're clearly planning to do.
1:09:51
They hired the person that built Facebook ads, right?
1:09:56
Like >> um I uh it's like they're they're creating new software revenue that re, you know, basically uh money that wasn't being spent on software that will now go to software.
1:10:07
They're going to be creating a massive >> ads business.
1:10:09
They're going to be capturing a ton.
1:10:11
They they have partnership with Walmart that they announced I think it was Monday.
1:10:15
>> They have I'm sure they'll announce a big partnership with Amazon.
1:10:17
They have a partnership with Shopify.
1:10:18
Like they will flip the switch and start to take a percentage of all the transactions that they're already driving.
1:10:24
So >> um I think that uh of course the 100red billion in a few years is uh is ludicrous, but uh I don't know that it's impossible. >> Yeah.
1:10:36
Epoch AI says one bubble oneway bubbles pop.
1:10:39
technology doesn't deliver value as quickly as investors bet it will.
1:10:43
In light of that, it is notable that OpenAI is projecting historically unprecedented revenue growth from 10 billion to 100 billion over the next 3 years.
1:10:52
It took Nvidia something like seven years to go from 10 billion to 100 billion.
1:10:58
uh Meta, uh Tesla, Amazon, Apple, Walmart, Google, they were all in the six to 10year camp to to actually ramp from 10 to 100 and Open AI is projecting doing it in three, which is very very aggressive.
1:11:12
Very very aggressive indeed. Um but yes, I I agree.
1:11:17
There's there's a there's a huge amount of uh software that it feels like it's like instantiated on the fly when you go into the chatbt app where uh I was talking to um uh a barber who was saying that chatbt is phenomenal at creating packing lists for travel and that's something that it could have been its own SAS product.
1:11:41
There might be a tool out there like packing list fortravel. com might exist.
1:11:46
Uh and now chatpt just like does that for you on the fly and it's something that Google used to like route you to that oneoff piece of software.
1:11:57
You can do a million of these like uh you know calculate interest rates, compounding interest uh you know you wind up on those like home mortgage interest calculator websites sometimes.
1:12:07
Calculate your BMI if you're into bodybuilding like diet plan.
1:12:10
Like there's so many different pieces of software out there that are just kind of like getting rolled up and we haven't really seen like the the SAS apocalypse come to those like small longtail pieces of software but that could be >> I think Google Google has also made a pretty big push into that obviously the calculator like you're not searching calculator on Google and then going to a website that's showing you display ads.
1:12:36
be so sad if you were like I had amazing lifestyle business >> just online calculator >> calculator.
1:12:41
>> calculator.com it was it was beautiful thing for 20 years >> they also they've also done that a uh they have a or or at least they did at one point like a mortgage uh mortgage interest calculator >> yeah it would just generate but that wasn't generated on the fly you know that was something that they handcoded
1:12:58
and then would like route you to if you if you searched certain terms um but open is able to do that uh in a lot of different ways we were talking about them expanding into the vice category uh with uh adult content and I uh was wondering if they could potentially um expand into gambling content. So I asked
1:13:13
So I asked Chachi BT to pretend you are a casino game of roulette. I put a 100 on black. Spin that wheel. Tell me if I won.
1:13:21
And Chachi says, "The ball spins and clatters around the wheel.
1:13:25
A tiny metal comet tracing the red black cosmos.
1:13:27
It bounces once, twice, then settles into its fate. Black 17. You win.
1:13:33
Your $100 bet doubles to $200.
1:13:37
The house size, the table cheers, and the wheel spins on.
1:13:43
Indifferent as ever to human triumph or ruin.
1:13:45
Care to let it ride or cash out and walk away with your small but honorable victory before probability demands its due.
1:13:54
And so, I mean, obviously that's a joke because they're not getting into gambling, but uh you know, they like JBD can just act as a casino for you.
1:14:00
And that's an example of like software that it can just instantiate.
1:14:04
You can just say like pretend you're generating a scratch ticket and it can just do it. >> Yeah.
1:14:09
In the same way you could ask a friend like present pretend to be a blackjack dealer and let's actually play with real money. >> Real money. Yeah.
1:14:15
I I don't think they're going to do that.
1:14:17
But >> uh there's probably an interesting eval where it's like uh how close are the actual odds >> like odds to to what it should be. >> Yeah.
1:14:24
Like you would think that it would be able to like store probabilities in memory and then use Python to generate like true random numbers.
1:14:29
Like I would imagine that it could get pretty close to reality if it like wrote some code in the background.
1:14:35
>> Yeah, I think it could do that definitely.
1:14:36
Um and also even if you just continue the chat and say like, "Okay, do it again."
1:14:40
But if you do a new one every time, I wonder how close it gets.
1:14:43
>> I I I did I did continue the chat.
1:14:43
I said, "Uh I said, "Let's double down."
1:14:49
>> And it said, "Bold move.
1:14:49
The dealer sweeps your chips into the felt circle. $200 on black." The wheel hums again.
1:14:53
A perfect machine of chance, physics, and misplaced confidence.
1:14:58
The ball launches, ricochets, clatters. Red 14.
1:15:01
The house reclaims its offering.
1:15:04
Your streak ends where it began.
1:15:06
Roulette in its quiet cruelty is a philosopher's game.
1:15:08
It teaches the fragility of of momentum.
1:15:11
Do you step away enlightened or spin again for revenge?
1:15:16
What do you think I should do?
1:15:18
Should I Should I double down again?
1:15:20
Put $200 I I should go to. I put $2 million down.
1:15:25
Um, well, whatever you're winning, make sure to pay your sales tax with numeral hq. com. Sales tax on autopilot.
1:15:29
Spend less than 5 minutes per month on sales tax compliance. Go to numeralhq. com. >> Uh, did you see that?
1:15:37
Uh, >> Kevin from the office was wearing a cubitus yesterday.
1:15:44
>> I didn't I didn't clock that on the stream.
1:15:46
We should have talked about that.
1:15:49
The denim the denim strap cubid.
1:15:51
>> Oh, it's a denim strap. >> Very cool. >> Uh, yeah. That's an $88,000.
1:15:55
>> I can't I honestly can't believe this is a real picture.
1:15:58
>> It's such a great pick for that character.
1:16:01
It's so funny in so many ways.
1:16:03
It's a very expensive watch, but at the same time, it's a watch that's not really loved by the watch lovers, and so it's it's sort of like cringey in certain ways, but that's exactly the character that he was trying to play.
1:16:14
It really is an incredible amount of detail that went into this performance.
1:16:18
Uh really uh really a royal flush by by the team behind this. So, congratulates.
1:16:24
>> Nick Carter says, "The government that I pay 40% of my income to has been shut for three weeks and nothing in my life has changed."
1:16:31
>> That's a rabbit staring in the mirror. >> Yeah, it is odd.
1:16:34
But, uh, I mean, the reason is that the government is not It is shut down, but most of the government agencies have like six to eight weeks of of cash, remember?
1:16:43
So, they can >> I'm glad they had runway.
1:16:46
>> They They have runway. They have runway.
1:16:48
Maybe Maybe they'll have to call Soft Bank and say, "Hey, we we need a bridge.
1:16:52
You can you can take 20% of the of the United States government.
1:16:54
Um well uh there's other news from Deutsche Bank.
1:16:59
European spending on Chachi PT has stalled since May. DB Data Insights.
1:17:01
Uh I think we all know what's going on there. >> Summer. >> Summer, baby. Summer each season.
1:17:10
>> You don't need to get anything done. >> You don't need chat.
1:17:12
>> You can use the You don't need reasoning models to tell you. >> Yeah.
1:17:16
Where to park your yacht.
1:17:16
You can you can you canot that with the with the with the free tier.
1:17:21
The free tier is more than enough when you're hanging out in Monaco.
1:17:25
>> Um yeah, I that that was my initial >> read on this is that you can't >> uh read too much into it, but it certainly is not good. >> Yeah. >> Right.
1:17:38
>> It would be really interesting.
1:17:38
>> It would be really interesting. Um I mean uh I would love to look at the ramp data on uh chat GPT spend across ramp customers whether that's still going up because I have this thesis that a lot of the paid subscriptions are essentially
1:17:54
like proumer work they're expens they're being expensed basically I think that uh for the general average person they are very much just using chacht on the free tier because $200 a month is a lot even $20 a month is still like well above Netflix and people see it as a productivity tool. So they either figure
1:18:12
So they either figure out a way to expense it or they or they and I wonder I wonder if that if that adoption rate is kind of uh plateauing in the in the you know broader uh you know ramp customer base.
1:18:23
That'd be interesting to look at.
1:18:25
Um what else is going on?
1:18:28
Starlink is live on United. Uh this is great news.
1:18:32
I wonder uh when this will actually be fully live.
1:18:35
this will actually be fully live. I imagine that it's going to be a slow roll out uh and they will work through uh you know one plane at a time but uh it's now live on board the first mainline aircraft and if you've ever
1:18:50
used Starlink on a on a plane it's remarkable jet X but a lot of private jets have them and you can FaceTime on them and stuff although I doubt you could FaceTime on a United flight it'd be somewhat disruptive to people but um whatever they do hopefully they use Finn.ai AI, the number one AI agent for
1:19:04
ai AI, the number one AI agent for customer service, number one in performance benchmarks, number one in competitive bake offs, number one ranking on G2.
1:19:11
>> What's that company that's competing with uh uh Starlink >> AS uh it has a retail army behind it and >> down 5% today.
1:19:21
>> Yeah, the stock's been all over the place.
1:19:23
They don't have revenue yet and so it's it it it's hard to exactly value, but they have a lot of contracts with cellular providers and cellular providers don't want to be completely locked.
1:19:33
they would prefer a duopoly and so I think they're willing to do contracts ahead of schedule, invest, do whatever it takes to get a second constellation up there.
1:19:42
Uh we did see Amazon yesterday talking about Project Koopier, which is another LEO satellite constellation that could potentially be a rival to Starling.
1:19:49
But I mean, man, the vertical integration at at SpaceX is hard to compete with when you think about them just putting up so many of them.
1:19:57
>> And Amazon has their play too, right? >> Amazon or Google? Google Google. >> Yeah, Koopier.
1:20:02
That's what we talked about yesterday.
1:20:04
>> Um >> well um our next guest is joining us in just a few minutes.
1:20:08
Um the timeline remains in turmoil over Nikita Beer.
1:20:14
Nikita said at this point I think creator payouts does more harm than good and we need to offramp to a different system.
1:20:21
And Elon said no the issue is that we're underpaying and not allocating payment accurately enough.
1:20:27
YouTube does a much better job.
1:20:27
And so Poly Market put up a uh a market for Nikita Beer out as head of product at X this year and Nikita said this is how I win.
1:20:37
So he is in the trenches. >> Now he can hedge. >> Now he can hedge.
1:20:41
Oh yeah, I guess he could.
1:20:42
He has the most inside information here for >> Yeah.
1:20:45
I mean I think going back to our conversation I think it was earlier this week on posts are so easy.
1:20:52
Every everybody that posts on X knows that often times their best post took the least amount of effort.
1:20:59
Whereas on the YouTube creator side for creators that are actually building a business. >> Yep.
1:21:05
>> Uh the average YouTube creator probably is inversely correlated in that >> some of my best videos have taken the most amount of effort.
1:21:12
Like my top five videos took the most amount of effort.
1:21:17
>> Uh some creative spark and then a ton of investment and everything from production to the editing process.
1:21:21
So >> again, I think YouTube creators uh certainly deserve to be paid well.
1:21:26
I have never felt as a you know relatively small creator on X that I should be paid. Right.
1:21:34
>> There's also the factor of allocating payment accurately. That is very difficult. >> It's so hard. Yeah.
1:21:40
What is the value of I is is a is a Jeremy Gon post that gets 30,000 views that's a lot more valuable to me than a meme that gets a million impressions. Right?
1:21:54
So, >> and so if a Germanon YouTube video were to go up, uh, and YouTube were to run a midroll video ad in that video, it's very clear that however much the advertisers paid for that specific ad in that specific video, it's very easy to allocate payment based on that.
1:22:13
But if you're scrolling a timeline and you see an ad here and then you see a post and then you see an ad here, it's very hard to say this ad was targeted at the reader of this particular post.
1:22:24
And so you run into these challenges and that's why uh Tik Toks embrace the the creator fund which is just which is just distributed based on impressions.
1:22:34
But impressions don't quite wait the quality of the audience, which is obviously incredibly important because advertisers want to get their message in front of customers at particular moments when they're primed to hear the message.
1:22:47
So, >> be capital Capital Bloke says the AI capex trade can't keep going up 2% every day on partnerships, projections, and information articles using 2030 estimates.
1:22:59
And of course, uh the AI capex trade actually can.
1:23:02
Uh Foxcon popped today, right?
1:23:05
Based on >> just the news that they had a conversation with OpenAI. >> Yes.
1:23:11
So the the news Foxcon jumped 8% based on this.
1:23:15
Uh they Foxcon shares rise after chairman says he met with OpenAI plans Nvidia next.
1:23:20
Uh so it's not actually a full deal yet.
1:23:23
Um, frenzied investing in AI has fueled talk of a bubble, but the chairman of Taiwan listed Foxcon, the world's largest contract electronics maker, says there's much further to go.
1:23:33
The application of AI is just the beginning.
1:23:35
Foxcon's uh, Young Leu told local reporters on the sidelines of an event on Wednesday, expressing confidence in the market's potential to keep going from strength to strength.
1:23:47
Uh he also said that Foxcon uh had met with OpenAI chief executive Sam Alman at the Taiwanese company's headquarters to discuss potential future collaborations.
1:23:57
And so the stock jumped 8% um on >> let's give it up for discussing future collaborations. >> Yeah.
1:24:04
And uh I mean AMD shares jumped 24% after the chip company announced a multi-billion dollar tie-up with OpenAI earlier this month.
1:24:10
Another multi-billion dollar they announced giving away with Broadcom similar. So I don't know.
1:24:17
I mean, Sam is vertically integrating like crazy.
1:24:19
We haven't seen this with a software company so fast.
1:24:22
I mean, Google eventually did it, but over a 20-year period, and Sam is just saying, I want, you know, multiple data center providers, multiple clouds, multiple chips, multiple, you know, fabs.
1:24:36
Like, he's really going deep into the supply chain and creating kind of a duopoly at every single phase.
1:24:40
Um, and that's really what I'm >> think about how much pressure there's going to be on the hardware launch.
1:24:45
like people are already going to the average person is going to be rooting for it to fail. >> Yeah.
1:24:52
>> And then uh actually delivering enough value to earn a place.
1:24:57
>> Are you talking about like consumer hardware? >> Yeah.
1:24:58
The Johnny >> Oh, because Yeah.
1:24:59
that Foxcon would be a logical partner.
1:25:02
>> That's what I assume the conversation is about. >> Yeah.
1:25:04
I was thinking about like what what a head fake it would be to be, you know, Samson saying like it's not a phone.
1:25:09
Uh and he's almost been saying like it's not a wearable, like it's not directly competitive with anything that's out there.
1:25:14
Like what if it is just like an Amazon Echo?
1:25:16
Like what if it's a direct competitor to like the Alexa, like that's actually kind of a great landing zone.
1:25:22
I think uh the more I was talking about Alexa and Echo, I was thinking that like yeah, being able to talk to an LLM and that can like it's a it's a much lower stakes uh environment to add add on.
1:25:35
Like I think a lot of people when they think about their their phone, they're like, I'm tied to iMessage and so I also have the Mac because I'm tied to the phone and then I also have the Apple TV because I'm tied to the Mac, but I don't feel fully tied to Apple such that I have to have Apple HomePods and the HomePod was not a massive success because people are like, well, I'll suffer through Sonos or I'll get a Google Nest thermostat.
1:26:01
people have been a lot more open to having an Amazon Alexa and then also or Google Home and then also an iPhone. >> Yeah.
1:26:09
Part of the thing with Alexexas I I bought an Alexa >> in college because it was a decent speaker that was cheap. Yeah.
1:26:17
>> And I maybe tried to use the feature order me some paper towels once. >> And so my Yeah.
1:26:22
Yeah, I guess with OpenAI it's like if it's just a puck that sits in your kitchen around your home and is listening all the time. Yep.
1:26:32
>> Is that really going to be valuable, more valuable than your phone?
1:26:34
Part of the draw with Alexa was, you know, you get a decent speaker out of it that at least can >> I mean, OpenAI could figure out how to produce a decent speaker at Foxcon, I'm sure. >> Yeah.
1:26:44
But then it's like, is it even really worth doing to go and just compete in the in the like low margin speaker market? >> Yeah.
1:26:53
I mean, Meta was trying with the uh the Meta Portal.
1:26:56
Do you remember that device?
1:26:58
It was a camera that would sit on top of TV.
1:27:02
>> Sonos is a $2 billion market cap company now.
1:27:05
Should OpenAI just buy it?
1:27:07
>> Equis Globos Global says Sonos plus OpenAI. I've heard rumors.
1:27:10
I don't know where you'd be hearing those rumors, but um I don't know.
1:27:14
Sonos >> I wouldn't be I mean I wouldn't be surprised.
1:27:17
It's like they have they did almost two billion of revenue.
1:27:21
OpenAI is great at software.
1:27:23
Sonos is great at hardware.
1:27:25
The the I mean the speakers look great at hardware.
1:27:30
>> Yeah, the the speakers look great.
1:27:30
And I don't know, there just is a question about like like if you're if you're entering these like hyper competitive markets, do you want to go for the most competitive phone?
1:27:38
Sam Alman certainly indicated that he might not pull.
1:27:41
Um >> hello in the chat says, "How much paper towels is Jordy?"
1:27:48
>> Yeah, Jord's obsessed with paper towels.
1:27:51
specifically paper towels.
1:27:52
>> I'll be honest, I have not bought paper towels in at least 5 years. >> Yeah.
1:27:58
>> But uh when I think about buying things in the home setting, that's an item >> you specifically want the lindiest paper towels possible.
1:28:05
Like paper towels from uh from a company started in the year, you know, 1000 AD or something. That's like your dream.
1:28:15
>> You're always on the hunt for a for an old school paper towel manufacturer. >> Yeah.
1:28:18
I want to know what tree it came from. >> Yes.
1:28:23
>> Um, continuing, uh, Vic says, uh, Derek Thompson said yesterday, OpenAI is an amazing company and these are impressive numbers.
1:28:32
Also, a company losing 20 billion a year with 13 billion of revenue, making business deals that project hundreds of billions in future spending with a private valu valuation of half a trillion is mental.
1:28:45
You would think, you would think, uh, John, wouldn't you think that at a certain point people would stop if a company It's one thing to be super bearish on a company that has massive losses and very minimal revenue.
1:28:57
It's much harder given the history of of companies that are in capital wars.
1:29:05
Like, you know, look back at Uber andyft.
1:29:06
The big critique of Uber was it's losing money.
1:29:08
It's never going to make money. >> Yep.
1:29:10
>> So, um, >> have you seen Uber's market cap recently?
1:29:13
I think it's like 200 billion >> TK. >> Yeah. 192 billion. Uh and a 15 PE ratio.
1:29:23
Like this business matured and is doing fantastically.
1:29:26
It's pretty pretty remarkable.
1:29:29
Like uh they very much like fought the capital war and won.
1:29:32
Like it's I mean it's like a testament to like they you know they they they went out.
1:29:40
I mean Lyft is an 8 billion market cap.
1:29:43
There was a time when those were then the spread between Uber and Lyft was like, you know, 2x, 3x, 4x, and now it's uh and now it's what 20x or something like that. >> Yeah. Yeah.
1:29:52
So, I think it's uh >> I just I just think that the company's losing money, hence it's crazy that it's valued at a lot is just not a good argument.
1:30:03
Vic says, "Stop looking at spreadsheets. Go try out codecs.
1:30:05
If anything, they're undervalued.
1:30:10
>> Stop looking at spreadsheets.
1:30:10
Go try out Adio Custom Relationship Magic.
1:30:12
Adio is the AI native CRM that builds, scales, and grows your company to the next level.
1:30:17
You can get started for free.
1:30:17
Um, also, it's the one-year anniversary of the first playable Chromatic going on display at the Portland Retro Gaming Expo.
1:30:29
This year, they're returning with the first playable M.
1:30:33
>> Portland has a retro gaming expo.
1:30:33
That is extremely on brand.
1:30:36
>> That's extremely Portland.
1:30:36
Got to give them credit for that.
1:30:39
This seems maybe we should dispatch Tyler to the Portland Retro Gaming Expo to go try the first playable M64. >> That would be cool.
1:30:46
I think we'll be able to get one of these pretty soon.
1:30:47
We can have Tyler can just spend a whole show >> beat Golden Eye >> playing the mod retro.
1:30:55
>> I wonder how fast you could speedrun Golden Eye.
1:30:56
Uh games back then were so short.
1:30:58
They just didn't make 100 hour games.
1:31:00
So, um people would people would speed through them.
1:31:03
Um, this is uh um what else is here?
1:31:06
How did you sleep last night?
1:31:09
I got a ton of sleep, but I slept poorly because I'm not feeling too well. I'm under the weather.
1:31:14
I got a 61% on quality, 79 overall. Go to eightsleep. com. Get a pod five.
1:31:21
Um, I slept >> my tracker got messed up because of >> hours and 40 minutes. This is a little one.
1:31:26
So, it's not I have >> We got to hit the gong again for Rahul and Julius because you can now connect Julius to your data bricks. Let's go. >> Hit that.
1:31:38
Uh this is a relevant post uh protesting for this.
1:31:42
>> Matt Slottnick posted on July 16th, 2025.
1:31:45
Bull case for nothing ever happens is that OpenAI runs its entire business on Salesforce and Slack.
1:31:52
Um, and this was in a blog post that an OpenAI employee posted.
1:31:57
They said, "An unusual part of OpenAI is that everything, and I mean everything, runs on Slack. There is no email.
1:32:02
I maybe received 10 emails in my entire time there.
1:32:05
If you aren't organized, you will find this incredibly distracting.
1:32:08
If you curate your channels and notifications, you can make it pretty workable."
1:32:11
Um, we got to ask Beni off.
1:32:14
Uh, does anything ever happen?
1:32:16
>> And which way which way does the deal go?
1:32:18
Because Salesforce has a deal with OpenAI.
1:32:21
Everyone assumes it's to buy a lot of tokens from OpenAI.
1:32:24
Uh but maybe it's a trade deal.
1:32:27
You get you get your Salesforce installation.
1:32:29
Uh the money it's one of those circular deals.
1:32:33
>> Sales works for tokens.
1:32:35
>> Zephr is sharing Broadcom's fifth customer isn't Apple or XAI. It's Anthropic.
1:32:39
They won't design a new chip.
1:32:41
They will be buying TPUs from Broadcom.
1:32:44
>> That's very interesting that they're not buying them directly.
1:32:46
>> Expect Anthropic to announce a funding round from Google soon. >> Yeah.
1:32:51
the anthropic thing has to heat up soon because Google like it's it feels like Google might be doing something.
1:32:58
That's what Zephr at least thinks.
1:32:58
But then you also have uh Amazon partnering with Enthropic for a long time and probably ramping that relationship up.
1:33:07
It seems like it would make a lot of sense for them to ramp that relationship up again.
1:33:10
And then um some folks on the timeline were talking about how uh how Apple might make a big acquisition, but maybe Anthropic's too big to make, but maybe even just a deal there would make a lot of sense.
1:33:23
>> So Google already invested in Anthropic.
1:33:25
>> Yes, they own a bunch. >> 14%. >> That's a lot. It's pretty remarkable.
1:33:29
>> How much does Amazon own? >> I don't know. Um look it up.
1:33:33
But uh while you do, let me tell you about public. com.
1:33:36
Investing for those that take it seriously.
1:33:38
They got multi-asset investing, industryleading yields, and they're trusted by millions.
1:33:43
>> Amazon owns an estimated 15 to 19% of Anthropic. >> Mhm.
1:33:48
>> Uh, and it's so funny to me based on the way on Anthropic's brand and approach >> and all these different things, the way that that you have OpenAI's cap table, which is owned by a nonprofit, the employees own a ton, and then VCs are like at the very bottom, right?
1:34:04
Just carving out a little bit.
1:34:06
Meanwhile, Anthropic is like Amazon, Google, like these big >> Yeah. basically.
1:34:14
>> Tyler, did you see this Eleazar Udicowski post?
1:34:16
He uh quote posted, "Barely AI, the Financial Times projected OpenAI's cap table following its for-profit transition.
1:34:23
Microsoft gets 30%, OpenAI employees get 30%.
1:34:26
Open AI nonprofit gets between 20 and 30%, SoftBank gets 10, and then the VCs get uh less than 10%.
1:34:35
And Eleazar said, "I wonder how the financial magnitude of this theft compares to the total amount of theft in the 21st century so far.
1:34:44
Wouldn't surprise me if it's a majority."
1:34:45
Is he talking about that?
1:34:45
He wants the VCs to have a larger ownership stake. >> I assume. Yeah. >> Yeah.
1:34:52
He must be just really because typically a company at this at this scale VCs could own 50 60%.
1:34:56
But they're sort of getting hosed and so less are standing up for the venture capital. >> Yeah.
1:35:03
He must be speaking up for for Josh.
1:35:05
Yeah, I would imagine just just bigger ownership stakes would have been more I I don't know if it's technically theft, but I mean maybe he's using that kind of in a loose in a loose way.
1:35:15
>> I think we need a some kind of Robin Hood figure to to to take from the rich and give to thrive. >> Yeah, exactly. Yeah. Yeah.
1:35:19
Take take the uh the nonprofit shares and just distribute them to the venture capitalist.
1:35:25
That that's probably what he wants.
1:35:26
That's what he's advocating for.
1:35:28
Um did you see uh Flock Alpha launch today?
1:35:32
We should pull this video up at some point. Um but uh they did it.
1:35:34
Flock launched a uh Americanmade public safety drone response system.
1:35:41
We were talking to the CEO of Flock Garrett uh about the role of drones, how you could potentially call a drone to escort you home, how you could dis uh how you could uh dispatch a drone to an emergency situation, start monitoring the situation.
1:35:57
And uh they did it and they're launching Flock Alpha today.
1:35:59
Uh Raul says, "Five years ago, I stood on the rooftop of my police station and imagined this moment.
1:36:05
There's no better feeling than seeing a vision come to life. I do not have kids yet."
1:36:09
Which is a good good point.
1:36:11
And so the the these can go on the top of all the police stations.
1:36:16
The police stations, as we talked to Garrett about, are already like geographically distributed across a city in a very equidistant way.
1:36:22
And so as long as the drones can get from one station to the other or halfway from one to the other, you can basically have full coverage of everything.
1:36:33
Uh and so you can read license plates.
1:36:37
>> Has there been a black mirror on >> drone, you know, drones for police that that go rogue?
1:36:46
>> I'm sure there's been some stuff, but >> this is uh yeah, I mean surveillance generally.
1:36:52
Dispatch >> Nikun says accidentally said sales instead of distribution and they kicked me out of SF. >> Got to use the lingo. >> Sales is evergreen. >> Yeah.
1:37:04
I I thought it was growth though.
1:37:06
People say distribution now.
1:37:06
I thought people would wait.
1:37:09
I thought sales was fully rebranded to growth.
1:37:11
People would be like, "Yeah, I'm on the growth team."
1:37:14
And it's like you're on the sales team.
1:37:15
>> I don't think it ever.
1:37:16
>> No, now it's under distribution. >> Well, oh well.
1:37:18
I gota we got to ask uh Beni off about sales specifically. Yeah.
1:37:24
Like what what reps are actually outperforming today?
1:37:27
Is it the reps that have adopted the most AI and have just increased volume or is it um >> is it is it reps doing things the oldfashioned way which is like actually getting getting to know existing and potential clients on a deep level and developing trust and and uh you know making sure they're getting uh great service etc. etc.
1:37:48
Um, Karina Nigan says, "After a formative time at OpenAI, I'm launching Mason AGI, a fashion house creating cultural artifacts for the AI era era.
1:38:00
Our first collection, Relic of Thought, is a collaboration between Ilia Sutskiver, featuring his original artworks alongside his signature hat modeled after his iconic."
1:38:10
>> They're actually doing this. Wow.
1:38:13
>> Do they have a picture yet?
1:38:14
>> Yeah, we we have to pull this video up at some point.
1:38:16
It is a study in conviction and the clarity of vision that gives thought its form.
1:38:19
We believe we're living through something extraordinary that deserves to be remembered.
1:38:24
Not in research papers or technical model cards, but in tangible objects we can hold, wear, and pass down.
1:38:30
Uh, working alongside some of the world's brightest researchers building Claude and ChachiT have come to realize they're among the most creative minds alive.
1:38:39
Research at its best is an act of imagination.
1:38:41
The ability to glimpse what doesn't yet exist and then build toward it.
1:38:45
Many of them remain unseen even as their work quietly reshapes our future.
1:38:50
Though AGI progress can feel incomprehensible from the outside, its story is deeply human, full of curiosity, conviction, and creation.
1:38:58
We're bridging that world with the creatives who give form to ideas.
1:39:00
This may be humanity's last time to create a handcrafted project before what we build surpasses us.
1:39:07
Each collection is also a message to super intelligence itself that we cared and that we tried to make beauty out of understanding.
1:39:16
>> Can we play this video?
1:39:16
I want to see what how they actually launched this.
1:39:19
Karina Win uh left OpenAI to do this.
1:39:22
That's a bold move, you know, been on a tear, you know, and now launching an entirely new >> bold to be post economic and pursue the arts. >> I guess you're right.
1:39:32
>> Become a patron of the arts.
1:39:32
It's kind of it's kind of a classic playbook. >> It is. It is. It is. It is.
1:39:36
Uh, we'll try and pull that up.
1:39:38
In the meantime, let me tell you about adquick. com.
1:39:40
Adof ofome advertising made easy and measurable.
1:39:43
Say goodbye to the headaches of out ofome advertising only.
1:39:45
Adquome expertise and data to enable efficient, seamless ad buying across the globe.
1:39:50
The chat has been on fire.
1:39:52
Mark Beni off is uh a little delayed.
1:39:55
We're going to remix the schedule a little bit.
1:39:56
He will be joining later the stream.
1:39:58
Uh Taylor Hodgej has a great take.
1:40:00
He says that he's going to use a trained falcon to take down the flock safety narc drone.
1:40:09
Uh I wouldn't be surprised if we see a little bit of that. Uh hopefully not.
1:40:12
I think that's wildly illegal.
1:40:14
You can't just uh you can't just sabotage police equipment.
1:40:18
You have to take it to the voting booth.
1:40:19
Taylor, >> what if it's flying over your prop private property?
1:40:23
>> Uh I mean, I'm sure that there are uh what is it?
1:40:26
The >> you train the Falcon.
1:40:26
You train the falcon to fly like low enough that you know there must be some law that like if it's within you know a hundred feet of the ground, it's fair game to take it out. >> Yeah.
1:40:37
The uh I think that uh you know the police can drive up and down your street.
1:40:42
>> Drones have to have search warrants to go fly around your property >> probably inside your house but not like on the street.
1:40:48
A police can a police officer can just drive down your street no problem.
1:40:51
Shine the light say oh is anybody breaking in?
1:40:53
Like you know that's that's helpful in many ways. You want that.
1:40:56
You want like a secure environment ideally.
1:40:58
Uh this is why we live in a democracy. We elect the >> Yeah.
1:41:03
Imagine if the the the UK right now, let's let's just be real, be very, you know, that I don't believe there's a lot of free speech left in the in the UK.
1:41:11
Imagine if they had this set up and that you could if you send the wrong post, a drone is actually at your at your door.
1:41:18
>> I think that might be overblown. I don't know.
1:41:20
I feel like who have you tal Have you talked to people in the UK that say it's bad >> or is it all just like American posters who are taking victory laps?
1:41:27
Because I love an American victory lap. Don't get me wrong.
1:41:30
I will jump on any opportunity to to wave the American flag.
1:41:35
>> Plenty of posts from people that are based in the UK. >> Yeah.
1:41:37
Say I got arrested for posting the Midwit meme about my enemy and I got I I depicted I depicted uh my enemy as this as the sad wjack and me as the Chad Wjack.
1:41:49
The trained Falcon taking out the the the drone is just too good. >> Be great.
1:41:55
>> The perfect form factor.
1:41:57
>> Do we have this uh >> we have Dante? Should we start early? >> Oh yeah. Yeah. Yeah. Let's bring in Dante.
1:42:02
Speaking of drones, underwater drones, we got Dante from Albaore coming in. >> Amazing name. >> The Yeah, great name.
1:42:08
Dante, >> welcome to joining us. >> Fantastic. >> Fantastic name.
1:42:15
I thought we were out of incredible names, but Albaor is strong.
1:42:20
>> Albaore is a great one. >> It's amazing.
1:42:21
I like >> Lord of the Rings stuff was taken, so we had to get creative.
1:42:26
>> Go to the tuna section.
1:42:27
>> Do you know the uh the the headline that uh is is in the the the group chat is YC backed Albaore swims into $6. 5 million seed round. I love it. >> Hit that gong. >> We've heard it all. We've heard it all.
1:42:41
We love making marathon funds. >> Congratulations. It's great stuff.
1:42:44
Uh >> yeah, take us >> who's in the round.
1:42:49
>> Yeah, so lots of folks.
1:42:49
Um it was led by uh Outlander.
1:42:52
Uh we also have our first institutional Thank you, Paige and AJ.
1:42:57
Um yeah, our first institutional backer was D3, which is a very drone focused fund.
1:43:02
Um they're sort of mix of US and and Ukraine um investments.
1:43:04
Um but they they've backed a lot of folks like Nearos.
1:43:09
I've got the >> they made two sequels to D1 Capital. We're on D3 now.
1:43:14
No, >> we're never going to run out of names there.
1:43:16
You just go to D4 and then D5 and eventually you go to D10, D11.
1:43:22
>> Yeah, A17Z is going to be a ripper for sure. >> Uh, >> who else?
1:43:27
Sorry, we got we interrupted. >> Oh, yeah. No, all good. All good. Um, yeah.
1:43:31
So, Brave Capital, uh, which is a bunch of exinkel people. Cool.
1:43:36
>> Um, uh, R squared, which is a national security focused fund.
1:43:38
uh managing partner there is a Navy Seal commander, Road Scholar who's on multiple national security councils. That's great.
1:43:46
>> Um uh a lot of different a lot of different investors.
1:43:49
You got Box Group in there, New York. Different stuff. >> Yeah.
1:43:53
Uh they've they've done a lot of defense.
1:43:55
I think they they also did uh Mock Industries. Oh, cool. >> Pretty recently. >> Uh yeah.
1:44:01
>> Um how did you get into this?
1:44:04
Uh yeah, so prior to this I actually started a a battery tech company and this was like while I was in college.
1:44:08
Um and really quickly identified we were working on like power systems that were high performance but um would make batteries for like really sort of uh complicated use cases um last longer.
1:44:18
Um and so uh that led us to the aerospace and defense market.
1:44:23
Did a lot of work with drones and started thinking about the power problems there. Right?
1:44:27
You know, the the issue with drones is you're really limited in terms of how far things can go.
1:44:32
Um and that problem just gets harder and harder um when you start thinking about the oceans, right?
1:44:39
Because uh oceans are really big.
1:44:39
Um you know, the people talk about in the Indoacific, the tyranny of distance.
1:44:42
And so I was at a Shabbat dinner, ran into this guy, John, who I'd known for a while through mutual friends, brilliant engineer.
1:44:50
engineer. can't share too much about what he was doing before this um but he's worked on some major unmanned systems uh production type projects um very knowledgeable about the sector and I was like John why why do underwater drones have just like such niche use cases right um they're very focused on um short range you know very human in
1:45:10
the loop type of operations there's nothing like an undersea loitering munition that we have right now um and uh we really need like longer range capabilities for anything that's happening in the Pacific so that's what we're working So is that like distinctly positioned against like the dive LD program from Anderole? Like how do you see like that
1:45:26
Like how do you see like that seems like a bold company to go up with against uh you know it's a great company.
1:45:32
Uh it's funny that you found >> right. Yeah.
1:45:35
I mean so actually the the founders of dive were were also investors in us really early on.
1:45:38
Sam Sam Rouso and Bill Lebo.
1:45:40
Um they're super knowledgeable about the space obviously.
1:45:45
U I was actually just with them earlier this week in Boston. Um, yeah.
1:45:46
So, so the dive platform is focused on going really deep and doing survey type work.
1:45:52
We're working on loadering munitions.
1:45:54
We're we're working on stuff that blows up.
1:45:55
So, you can fit a 250 pound explosive payload on our vehicle.
1:45:57
And that's enough to do really severe damage to a large surface vessel and if you're lucky enough to find a submarine to to sink that submarine. Wow.
1:46:06
>> Um, and >> what are the kind of um I don't know how much you can share, but like what what's the goal in terms of for like the timeline of a loitering munition?
1:46:14
you want to go out effectively, park it somewhere strategic, >> and then it's just kind of playing a waiting game >> before you would want to activate it and have use it for for a certain use case.
1:46:28
Like, how long do you want one of your drones to be deployed without any sort of like human or machine intervention? >> Yeah.
1:46:37
So, uh we're targeting 30 days uh for endurance and a thousand nautical miles.
1:46:42
And to kind of put that in perspective, the dive vehicle, which is, you know, a couple thousand pounds, um, that's less than a week and, um, a few hundred nautical miles.
1:46:51
And as you get to the the size of vehicles that that we're building, you know, our vehicle is like twoman portable.
1:46:56
And, uh, our our lead software engineer claims that he can bench one of our vehicles. So, >> yeah.
1:47:03
Well, he's he's really strong.
1:47:08
>> The best news I've heard all day. That's amazing.
1:47:10
Well, he actually he just got one of his fingers cut off and so not related to his work here, but um once his fingers are healed up, he says he's ready to do it on >> actually cut off. >> Yeah. Yeah.
1:47:20
He works on a tall ship and so uh yeah, log fell on his hand. It's a long story. >> Brutal.
1:47:26
>> He can still type faster than anyone in his in the office.
1:47:28
So >> that's you lose a finger though.
1:47:30
That's that's 10,000 aura points >> for sure.
1:47:35
That's like eye patch level.
1:47:37
>> Yeah, >> you're we docked his salary by 10%. Oh, stop. That's so rude. Stop it. That's very dark.
1:47:44
>> What uh what what's your path to program of record? How are you attacking? >> Yeah.
1:47:48
I mean, so uh contracting is is changing really really fast, right?
1:47:51
The current administration is doing things very differently.
1:47:54
And you can get pretty far without a program of record, but um I'm not sure if you're familiar with the defense autonomy working group that that just got released.
1:48:01
Um there's a lot of things that are that are moving through programs like that with with SOCOM.
1:48:04
Um, you know, Indopayom has various plush funds that are available and we're, you know, we're focused on on really connecting with the operators and making sure that, you know, from the start of, uh, founding this company, we were really talking with with end users, the people who are going to be at the tip of the sphere um, in the Indoacific who'd be using this.
1:48:22
Um, but you also have to go once you have an understanding of those requirements, you have to go to Congress, too.
1:48:27
And so, um, we we were actually lobbying Congress ourselves uh, within the first couple of months of starting the company.
1:48:33
We've since moved on to we work with a firm now.
1:48:34
Um but uh yeah, you have to basically approach it from all angles and um you know there's also a lot of you know FMS FMS interest as well.
1:48:43
So, um, can't get too much into specifics there, but there's a lot of companies that have been really successful, um, with a a route to commercialization that consists of a number of different buyers and, um, you know, largely speaking, DoD can be, um, really appreciative of of FMS as well, um, because that means you can reach larger scale and, um, you know, potentially even battlefield test your systems.
1:49:05
>> Are you guys based in the Gundo? >> We are not.
1:49:07
We are based in Philadelphia actually.
1:49:11
Is isn't there is there great robotics program out there? Is that is that why?
1:49:16
>> So, um there's a lot of great stuff about Philadelphia.
1:49:17
We're super close to Washington DC where a lot of the um you know customers that we'd be speaking with are based.
1:49:23
Uh so it's a very quick trip.
1:49:25
We can get up here on you know very short notice which is important.
1:49:26
Um there's also a lot of great engineering and industrial talent in Philadelphia.
1:49:32
Uh and so I know there's not like too many hard tech companies that are based here.
1:49:34
There's one they got pretty big called Ghost Robotics.
1:49:36
They make like robot dogs military. Yeah.
1:49:37
Um but we're we're based in the Penovation Center right now which is where Ghost Robotics started.
1:49:43
Um and we have lots of access to brackish water just um right next to our offices.
1:49:48
Um and yeah, I mean Philadelphia has a really rich ship building history and um you know recently the the uh naval yard it's it's being um re reindustrialized.
1:49:56
Um but it's it's actually not owned by a US company anymore.
1:50:01
Uh and so we want to build the sort of next generation of maritime capabilities in Philadelphia.
1:50:08
um and you know help sort of foster more of that industrial labor force here that's still you know really strong. >> Amazing.
1:50:14
>> Does your office have a pool?
1:50:17
>> So we're getting a giant fish tank and also a turtle.
1:50:19
Um but we don't currently have a pool.
1:50:21
We have limited limited space. >> Okay.
1:50:24
>> What kind of fish tank do you need to make sure that like a presumably metal drone doesn't doesn't smash through the glass?
1:50:32
>> It's going to be a pretty big fish tank. I don't know.
1:50:33
I'd have to look into the specifics more. >> Yeah.
1:50:36
You got to call Keith Rey.
1:50:36
He's He's got a big fish tank. >> Oh, yeah. He does.
1:50:39
He's famously fish uh fish tank build.
1:50:41
Uh well, thank you so much for coming. >> Great to meet you.
1:50:43
I'm sure you'll be back on soon.
1:50:45
Uh and uh congrats on on the milestone. >> Yeah. >> Yeah. Thank you guys. Been a pleasure. >> Talk to you later. Have a good one. >> Cheers. >> Bye.
1:50:53
>> Uh and if you're interested in some hardware for your wrist, head over to get getbzzel. com.
1:50:57
Your bezel concier is available now to source you any watch on the planet. Seriously, any watch.
1:51:01
Um, Andre Carpathy launched a new repo. Nano chat >> repo alert. >> New repo alert.
1:51:09
Uh, did you see this, Tyler? >> Yeah. Yeah, it's super cool.
1:51:13
>> This is among the most unhinged he's ever written.
1:51:15
Um, 8,000 lines of code to actually build basically chat from scratch.
1:51:22
Is that how I should understand it? >> Yeah.
1:51:24
So, um, he he's done he's done this whole like course lecture series on YouTube before. Yeah.
1:51:27
Um, where he goes through like uh different parts of how to train a model. Yeah.
1:51:31
Um, and then this is basically kind of all of them together plus some extra stuff.
1:51:34
So, it's training the full model.
1:51:36
It's doing the pre-train and it's also doing the RLHF to get it into like from predict internet text to be like a chat model >> and it's just using open source data sets basically for >> Yeah. Yeah.
1:51:47
I forget the exact name, but there's a a good like RLHF data set that you can use to to get it to become like a chat model.
1:51:53
>> It's pretty remarkable.
1:51:53
Is this so this is mostly for uh like educational resources?
1:51:57
Yeah, I I think mostly I mean um like presumably you'd have to run this on a much bigger scale to get like a really kind of capable model. Yeah. Um but yeah.
1:52:07
>> Uh what what was your interpretation of the hot take that somebody asked?
1:52:11
Curious how much of the code was written by hand?
1:52:13
And Andre Carpathy says, "Good question.
1:52:16
It's basically entirely handwritten with tab autocomplete.
1:52:18
I tried to use claude/codex agents a few times, but they just didn't work well enough at all and net unhelpful.
1:52:26
possibly the repo is too far off the data distribution.
1:52:30
>> Um, that's kind of interesting.
1:52:30
I think uh part of it is that like the end product of this is like the repo itself.
1:52:36
So you want very like nice code that is very readable.
1:52:38
It's in his style >> and he can like explain everything super easily which is like a little different from if you're building like actual product or or just like a vibe coding thing where the endpoint is just like the interface.
1:52:49
It's what you're actually like using where this is like the code itself is the product.
1:52:52
So you're it's like you need super nice, you know, formatting and all this like kind of custom stuff.
1:52:59
>> Do you feel that like the like the the whether or not the task is in the data distribution when you're firing off cloud code or codeex agents like are you noticing that it's like okay well clearly uh you know there's plenty of data around how to build like a mobile a modal or you know some sort of like rest interface and so it's really good at that.
1:53:21
But if you're trying to do something that's um you know probably a unique problem that there might not be like a direct open- source example of like you're going to try and code it from hand.
1:53:32
>> Uh yeah, that's probably true.
1:53:32
Um like I I find it's it's it's really good at writing React code and like Tailwind CSS components.
1:53:39
>> Um and then when I try to do some like >> custom, you know, AWS flow, it's it's a little harder.
1:53:44
Um and it just takes longer, but I still find it's it does it very well. >> Yeah.
1:53:48
Uh Jordan, did you see that uh Tyler Cowan put the timeline in turmoil taking shots at Europe?
1:53:56
>> So Tyler Cowan, we talked about this yesterday.
1:53:58
He said that uh the European Union is run by uh people who have incredible depth of knowledge in a bunch of different areas and if they ran the world, the world would be growing at negative 1% growth and uh and is pretty pretty hot take.
1:54:13
Uh Cassie Pritchard said, "Never understood why people liked this."
1:54:17
quote, "The EU does exist.
1:54:17
It is run by people like this and it does not have a negative 1% growth rate."
1:54:22
I'd say that's about the typical sharpness of analysis.
1:54:26
for Tyler Cowan though taking shots at Tyler friend of the show and young macro says difficult to overstate the extent to which you need to be underexposed to macroeconomic data to think that Tyler Cowan diagnosing Europe's growth as anemic is some sort of contentious statement or that his negative 1% hyperbole is anything but mild and contextually entirely suitable.
1:54:51
Young Macro continues and says >> I will defend Tyler Cowan.
1:54:54
>> I will defend Tyler Cowan with my life.
1:54:56
uh difficult to overstate the extent to which one's large audience must skew uneducated for this take to pass through with broad-based support.
1:55:02
Nothing to do with pedantry.
1:55:04
10 out of 10 pedantic domain experts would say Tyler Tyler's quote is quite fine.
1:55:09
Tyler Cowan, there's no way a regular guy would lose to a hobbit in a fight. A regular guy is what? Six feet. No chance.
1:55:16
what? Six feet. No chance. person who has come across height as a concept as insufficient number of times an insufficient number of times to be up to speed with baseline magnitudal inst intuitions ecstatic uh at the gotcha after googling the average male height
1:55:31
is 5'9 actually having a lot of fun on the timeline wow and yeah Cassie Pritchard's really going back and forth the young macro one of the greatest upand cominging posters um should we watch this clip from Rolof Ba from uh uncapped um unhinged >> the uh he is the steward of uh of Sequoia Rolapota. Here let's hear it. Here let's hear it.
1:55:54
>> One of the greatest stewards of capital in history.
1:55:58
>> It doesn't support the numbers.
1:55:58
So there was a lot of analysis back in the 1970s and 80s with you know the capital asset pricing model and people figured out that there's this asset class that supposedly has uncorrelated returns and a bunch of asset managers deem that they need to invest a certain percentage of their endowment or foundation or pension fund into this thing called venture capital.
1:56:17
If you look at the data they're basically 20 companies per year on average over the last 20 30 years that have ended up being worth in realized exits a billion dollars or more. just 20 companies.
1:56:28
Despite a lot more money plowing into into venture capital, we haven't seen a material change in the number of companies that are outcomes that are that large.
1:56:36
And I think part of that is that there's a lot more talent than really interesting ideas or interesting companies to be built.
1:56:42
And I think we're spreading a lot of that talent thin right now, similar to what happened in 1999, by the way. >> Yeah.
1:56:49
>> So when you look at the data, the amount of money going into venture capital right now in America is in the order of $250 billion a year.
1:56:53
And the numbers, you know, these are all estimates.
1:56:56
Let's just say it's $250 billion a year. >> Need a lot of exits.
1:57:01
>> Well, let's just do some very simple arithmetic for a second.
1:57:02
$250 billion going in every single year.
1:57:04
If you assume that the firms generate 12% irr net of fees and carry, which isn't that great, by the way.
1:57:12
I mean, over the last three or four years, the NASDAQ has compounded at 16 17%. Let's just say 12%. Not spectacular.
1:57:17
You basically just average performance. you'd need a 3.
1:57:19
7x roughly on 10 years >> on a seven-year exit horizon.
1:57:24
So, I'm being a little bit aggressive.
1:57:26
I mean, maybe it won't even be that good. So, 3. 7x on 250 billion.
1:57:28
That approximates to a trillion dollars a year >> coming out.
1:57:34
>> Coming out, by the way, that means and and that's what the investors own.
1:57:36
So, let's say that the investors own two-thirds of the company to make the arithmetic simple. That's 1.
1:57:40
5 trillion annually in company exit value. >> Yes.
1:57:46
>> Just think about that for a second.
1:57:47
>> Where's that coming from?
1:57:48
>> Well, Figma is worth what?
1:57:48
It's a lot but 40ish billion.
1:57:50
I mean that's worth you know. 03 trillion.
1:57:52
So you know if you start thinking in trillions and figma gets you 0. 03 trillion.
1:57:59
>> You need 30 40 50 Figas every year. >> Yeah.
1:58:04
>> To make that arithmetic week.
1:58:06
>> It just I don't see that many companies of that scale every year.
1:58:08
So the only thing breaks is the return assumption doesn't hold. >> Yeah.
1:58:13
>> And so venture is a return-free risk.
1:58:17
>> Not a risk-free return. It's terrible.
1:58:19
>> Yeah, you are better off investing in the index or holding t-bulls honestly.
1:58:25
And so I don't think venture is an asset class.
1:58:26
Asset classes scale if you add more money.
1:58:28
You can build more real estate.
1:58:29
There's a lot of equities.
1:58:29
There are, you know, trillions and trillions of bonds to be purchased.
1:58:33
Venture capital doesn't scale with more money.
1:58:38
>> I mean, the most important I mean takeaway here is Bobby's question in the chat. What's on the wrist? Risk. Risk check.
1:58:44
We'll work we'll work on that.
1:58:46
>> We will we will get to the bottom of that.
1:58:47
Venture is returnfree risk is a crazy line. >> Yeah. >> Um >> taking shots.
1:58:54
>> But uh there was some good followup from Hari. >> Yeah. >> Ragavan.
1:58:58
He said obviously Roloff's a legend and one of the smartest people in the valley, but I'd love to dissect this add a bit more nuance to this.
1:59:03
I completely agree with his directional point that most of VC everyone but top deskile sucks as an asset class.
1:59:09
But I think the same is true of hedge funds and the top cortiles less uh risky asset classes like real estate etc.
1:59:17
But the nuance is that the true venture capital segment formation through series A or modest series B is probably 20 billion a year maybe 25 billion because the remainder of the 250 billion is really private equity being deployed by a VC firm.
1:59:34
When you take it down in order of magnitude, the exit expectations make a lot more sense for true VC that is uh over let's say 8 to 10 year horizons 4 to 5x.
1:59:43
So 100 to 125 billion in investor ownership at 150 to 200 billion in exits for the PE growth equity class theoretically of 1.
1:59:54
7 to 2x clear similar hurdle return because the hold periods are say 3 to 5 years from series D E etc uh works out to about 350 billion.
2:00:04
So the total now comes out to 400 to 600 billion.
2:00:09
Nothing to sneeze at but not one and a half trillion.
2:00:12
And then you don't need 50 figs a year, but 15.
2:00:14
And more importantly, if a unicorn is a home run, a decacorn is a grand slam.
2:00:19
We're seeing more of the centacorn, which >> triple crown.
2:00:23
Let's go for the horse race analogy.
2:00:26
Switching out of baseball into horse racing.
2:00:30
>> Um, and he says because you could have 15 fig or literally one open AI. >> A centacorn.
2:00:34
If a if a unicorn's a home run, a decacorn is a grand slam, a centacorn should be like a world a world series clean sweep, something like that, like you won the whole se the whole season.
2:00:46
>> Um, he's saying I'm not saying we'll have one open eye per year. That'd be crazy.
2:00:50
But we could see per year one, stripe, data bricks, etc.
2:00:54
Two to three or sorry, three to five figma scale outcomes.
2:00:59
>> Uh, 20 times Wealthfront scale outcomes.
2:01:02
Wealthfront obviously was recently acquired by Gusto.
2:01:04
All we have to do is adjust that there are 20 unicorns a year for inflation and that's very plausible and that gets us to 400 billion in exit value.
2:01:12
Markets are so much bigger than they used to be 10 years ago.
2:01:14
We had zero trillion dollar companies.
2:01:16
Uh he says maybe I'm just too much of a techno optimist and Naval comes in with the reply and says nailed it.
2:01:26
>> Correction on Wealthfront. Wealthrun's independent.
2:01:27
You're thinking of a different company.
2:01:30
>> Uh >> Wealthfront's the robo adviser.
2:01:30
Oh, gusto guidelines, sorry, guideline >> somewhat different.
2:01:35
Um, yeah, the the interesting like nuance here is how weird open AI is.
2:01:40
It just breaks the venture model because yes, it is, you know, maybe going to put up a trillion dollars of value, but as we saw the venture ownership Roloff was saying it'll be 50% ventureowned.
2:01:52
Uh, and that's not the case for OpenAI.
2:01:55
And then also it just completely breaks the whole model of like who's founding these companies.
2:02:01
Sam Alman's like much more experienced, much older than, you know, some uh Gen Z needle in the haystack founder that you can pick out and get, you know, 20% of the company for for $10 million.
2:02:13
That just wasn't the nature.
2:02:15
Like the one of the first venture deals we talked about with Thrive was what $150 million at like a 12 billion valuation.
2:02:22
it was like, you know, building like a 1% ownership in the firm.
2:02:26
And so, uh, it's, uh, it's it's like even it's such it's just such an outlier.
2:02:33
But this, that's the game of adventure, right?
2:02:34
That's the game of venture.
2:02:35
Um, OpenAI just hired prizewinning black hole physicist Alexis Loop Loops Lucasa as the first member of its new opening eye for science team for Axios.
2:02:48
Uh, good timing here because they are >> uh, leaning more into science.
2:02:53
Uh, Loops Saska is a Vanderbilt professor known for his work on black hole photon rings and now he will help shape how GPD5 tackles advanced math and physics problems and guide open AI's push into scientific discovery.
2:03:07
VP of science Kevin Wheel said GPD5 can already perform limited novel scientific research, calling it the worst AI model you'll ever use again.
2:03:15
Um, I got to meet Alex earlier this year and I said, uh, OpenAI should hire you.
2:03:23
Uh, I I want to I'm hit him up after this and say, uh, I'm sure they he he he, um, he said something like that.
2:03:33
That sound that that sounds like it'd be fun.
2:03:34
I'm sure they were probably already talking at that point.
2:03:37
So, um, not the most original idea, but uh, awesome guy and super excited for him to be over at OpenAI. >> Yeah.
2:03:45
Um well, we should prep for our next interview uh with uh Iso Kant uh from uh Poolside.
2:03:52
Uh the news that we'll be digging into at some point during the show is uh a giant new AI data center is coming online to the epicenter of America's fracking boom.
2:04:02
We talked about this uh yesterday, but uh uh Cororeweave is partnering with them to build a uh a massive data center on more than 500 acres of land that sits on a sprawling ranch in West Texas.
2:04:16
The site is owned by the Mitchell family, which has run on oil and gas companies, which has run oil and gas companies for decades in the state and is located in the heart of the fracking boom.
2:04:27
>> All right, we've got our next guest coming in.
2:04:28
Okay, let's bring him in to the TV show.
2:04:33
Welcome to the shop and how are you doing?
2:04:36
>> Hey guys, good to see you. How are you doing?
2:04:38
>> We're pumped to have you.
2:04:38
If if for people that were watching yesterday, we were reading through the article about your guys' new project in West Texas.
2:04:49
And I was like, poolside?
2:04:49
Like, poolside's going this hard? >> Yes.
2:04:53
>> And when we realized that it was you, we were pretty excited. >> Yeah.
2:04:56
So, please tell us the the the story of the company.
2:04:58
Could you can you start kind of at the beginning uh little bit on your background uh then the first act of the company because it does feel like you're at a very unique moment in the company history. >> Yeah.
2:05:10
So my story into space actually started I was listening a couple of minutes before to your interview.
2:05:13
You were talking about Andre Karpathy and him writing code.
2:05:16
I've actually never said this.
2:05:18
My story into space started because Andre Karpathy wrote an article in 2015 called the unreasonable effectiveness of recurrent neural nets.
2:05:26
It's one of the early kind of blog posts about language models and it captured me so much that I ended up pivoting my entire company towards it called sourced >> and we spent the next four or five years building language models uh that were capable of writing code. >> Mhm. >> Sounds great today.
2:05:40
Back then 50 people in the world cared. >> Yeah.
2:05:43
>> Uh and uh but 2015 was an interesting moment in time because it was followed the year after by Alph Go coming up.
2:05:49
M >> and at that point I built probably an unreasonable conviction that the combination of language models and reinforcement learning should be able to generalize to frankly anything and everything that you can approximate including human intelligence.
2:06:03
And uh that was kind of my origin story.
2:06:05
Uh I met my co-founder in 2017.
2:06:07
He was the CTO of GitHub.
2:06:10
>> He made an acquisition offer for that company.
2:06:12
>> The guy >> wait you made an acquisition offer for what company?
2:06:15
So, GitHub made an acquisition offer for my company in 2017 because we had the world's first code completion models that were working back then. Makes >> sense.
2:06:23
Uh, turned down the offer, but nonetheless became really good friends.
2:06:27
>> Uh, and this company ultimately ended up not succeeding. Uh, we were too early.
2:06:32
And so, you can kind of figure out what happens when on November 2022, you see Chach come out. >> Yeah. >> Right.
2:06:38
It's everything you've been saying for years kind of into the void somebody else just did and did in an incredible manner.
2:06:44
And so it kind of gave this really deep realization in that time that everything was about to change.
2:06:50
It was kind of like a pre-post electricity moment like we're going to truly fully reach human level intelligence and go beyond.
2:06:55
But the narrative at the beginning of 23 was all we have to do is scale language modeling.
2:07:01
Let's just size them up, predict more tokens on the web.
2:07:03
And just fundamentally disagreed.
2:07:05
So you know the our our view and it's why we started Poolsite two and a half years ago was that reinforcement learning was going to become the most important scaling access for model capabilities.
2:07:16
>> Uh first 18 months of the life of our company that felt like one of the most contrarian opinions you could hold.
2:07:19
I think today it's it's it's clearly not anymore.
2:07:24
>> But that's kind of our origin story and why we decided to build a foundation model.
2:07:27
>> And what was the what was the go to market that you had in mind?
2:07:29
I mean we've seen the the market play out where there are uh kind of like synchronous IDE based code completion tab completion models.
2:07:39
There's a lot of custom models in that world almost like I call them like proumer use cases where I hit GPT5 and it winds up writing code and I didn't even ask it to.
2:07:48
And then there's the codecs, there's cloud code, there's agents, there's wind surf, there's cursor and there's so many different positions.
2:07:56
like how did you see the market map develop from your world?
2:08:01
>> So we bring it back to like what ultimately our mission is, right?
2:08:03
We're here to to reach human level intelligence and go beyond.
2:08:06
And in that world, intelligence in our view is a commodity. >> Mhm.
2:08:12
>> It's actually a commodity that gets created by only a small number of companies because of the sheer amount of resources and kind of compounding efforts that go into it.
2:08:17
But I think at the limit, we're not going to find very large differences between one foundation model and the other.
2:08:24
And so I think as a foundation model company, you're in two businesses.
2:08:27
You're on one hand in the what I often refer to internally as the barrels of oil business, >> right?
2:08:32
You're just selling your tokens behind an API and that's your commodity business.
2:08:35
And in a commodity business, you care about cost and scale.
2:08:37
And they'll come back to a little bit why the info project so important. >> Yeah.
2:08:42
>> The second though is what do you choose to do with that commodity? Right.
2:08:44
Once you have intelligence available to you, who do you want to be for?
2:08:48
And from day zero, we wanted to be for the world's frankly knowledge workforce.
2:08:53
We wanted to be for the enterprise.
2:08:55
We wanted to be for the world's most like high consequence environments.
2:08:57
So our models are now rapidly progressing in capabilities.
2:09:03
It's it's definitely been nonlinear.
2:09:04
And now we see a path to be at the frontier.
2:09:07
>> But when we weren't at the frontier, we kind of decided to cut our teeth in a go to market area which was really no one else was at which was defense and and government. >> Oh, interesting.
2:09:16
And so kind of our first customers were not just building the model but also building all the crazy enterprise stack to be able to deploy it anywhere like literally and workstations and Humvees all the way to like the larger models and like air gapped environments or guff clouds or places where you needed ATO's.
2:09:30
Uh and now we're kind of on track to start expanding out of that defense sector and going into the wider enterprise not just with coding agents by the way that's been our really our starting point.
2:09:40
It was our view that's where the market was first going to go adopt for frankly not a very intelligent insight just the fact that developers we've always been the first adopters of technology. >> Yeah.
2:09:50
>> So it was kind of clear that that was going to happen and >> and it's and it's really fun to build tools for yourself.
2:09:56
It's less fun to build tools for a role that you've never done. Right. >> Yeah.
2:10:02
Look, intelligence is we treat it as this one northstar like intelligence, but in reality, it's actually really multi-dimensional, right?
2:10:07
Like how good you are at writing poetry is very different than how good you are at writing code versus how you are, you know, being a researcher in biology.
2:10:14
And so the thing that was the unlock, I think, in our space is that, you know, first generation of models had only been trained on the output of humanity's work, right, like the web, but it wasn't trained on the thought process and actions that led to the creation of that work.
2:10:31
And it was going to be clear that coding was going to be the first domain where we could do that through reinforcement learning because we could simulate it, >> right?
2:10:37
So, so we spent the last two and a half years building what I believe is the largest RL environment in the world.
2:10:43
It's a million real world code bases >> where you know our agents can do hundreds and hundreds of billions of tasks.
2:10:48
And so that's kind of it was it was partially close to our heart and but it was also frankly very close to what we saw the path to where it's like more capabilities and models was going to run through.
2:11:01
get into verticalization. >> Yeah.
2:11:03
Talk about the core weave deal.
2:11:06
>> So the core weave deal we announced has two components to it.
2:11:08
Um we've been able to do in the last 2 and a half years what we did with 10,000 H200s.
2:11:15
>> Think of that as like annually 100, you know, 2050 million worth of compute. >> Yeah.
2:11:20
>> Uh but it's orders of magnitude less than what others had.
2:11:22
And so we built our efficiencies around orders of magnitude more efficient experimentation.
2:11:29
uh but your big model run is still your big model run. >> Yeah.
2:11:33
>> And so that like the size of a model you can train has uh and the duration at which you train it uh has clear correlation with the final intelligence and capabilities.
2:11:41
So we needed a lot of GPUs very fast uh because we saw now that hey we had gotten to a point where our models had gotten so good now that we knew if we'd scale them up we'd be on track to be at where the frontier was going to be.
2:11:54
And so that really started uh with conversations with with Nvidia and Cororeweave, right?
2:12:01
and Cororeweave, right? and and uh for kind of obvious reasons and we got really excited because we found a path to partner with coreweave that brought online more than 40,000 GB300s really quickly and I don't know how much you guys have discussed in the past
2:12:19
how much you guys in the in the past kind of discussed you know like the compute market but right now that scale of compute is impossible to get >> it is sold out for the entirety of 2026 and already like well into 2027 so >> we founded strategic partnership that that allowed us to do so. And uh so that
2:12:33
And uh so that gets computer online December, but that gets you to the frontier, but what then?
2:12:40
Like you can build the world's most capable model, but if you're not able to serve it, if you're not able to scale it up further, if you're not able to train the next generations, you're frankly you're not in this race, you're just cosplaying.
2:12:52
>> And so we had to take a step back and this was already a while ago.
2:12:54
This was already started looking at this, you know, a year ago.
2:12:57
And uh so well the true bottleneck in our industry is not chips and it's not energy.
2:13:01
There's a lot of like 400 kilovolt electricity that comes off the grid.
2:13:07
There's a lot of sources of energy in the United States.
2:13:10
And while at the limit it is the bottleneck, it's not the immediate bottleneck.
2:13:12
The actual bottleneck is bringing it all together and actually having powered shells like data centers online. >> Mhm.
2:13:19
Because while last year I could call someone for 50 megawws and I could kind of, you know, get it within six to nine months.
2:13:25
If I needed to call someone for 250 megawws, guys, there's no one you can call.
2:13:31
>> At least you can make like a multi-billion dollar payment or commitment, you know, today on a 15-year lease and then you can get it in 18 to 24 months. >> Yeah.
2:13:38
>> But I'm not meta, right?
2:13:38
And so we we understood that we had to own that vertical stack entirely if we were going to be able to secure our future.
2:13:45
to be able to secure our future. when how early did you guys make the call to focus you know clearly it sounds like you're focusing intensely on the go to market side as well right we've heard you know there's plenty there's labs out there that have like they they're taking the route of like you know if we build
2:14:02
it they will come type of thing and it very clearly even before you guys are at the frontier you're saying like no we're going to get customers we're going to get actual enterprise use cases we're going to be focused on delivering value and then hopefully those products just get better for the customers as the underlying intelligence improves. But uh
2:14:17
But uh do you feel like that uh uh yeah like was that just an easy obvious decision to make or did you guys like debate that a lot internally?
2:14:30
>> So it it was kind of the DNA from day zero who we wanted to be and and so we wrote ourselves and my co-founder like a day zero memo and we often go back to it to see you know like what's changed along on that and we kind of understood early on that there were three layers that were really going to matter for the next decade.
2:14:45
It was going to be energy, comput, and intelligence.
2:14:46
And in our view, a lot of things were going to become rounding errors compared to those three.
2:14:51
And but you also knew that if you if you see intelligence as a commodity, right, if you treat it like a barrel of oil, a barrel of tokens, >> someone else is going to deliver it, right? >> Exactly.
2:15:04
And and look, and so you care about your cost and your skill, hence infrastructure and vertically integrating.
2:15:09
But you don't want to be in a commodity business.
2:15:11
is you want to be in a business that either increases someone's revenue, right?
2:15:13
Or improves their cost basis, like it helps them grow their business.
2:15:17
And so from day zero, we said, well, we're not a consumer company.
2:15:20
It's probably hearing how I talk. It's not in our DNA.
2:15:22
I'm not a chat app in your pocket.
2:15:24
This is not I wouldn't know how to build that.
2:15:26
But we deeply cared about businesses and enterprises.
2:15:29
And so it was a day zero decision uh to do this.
2:15:31
And so from day zero, we built both parts of the org, our applied research org and what we call our production engineering like org.
2:15:39
org. this 2 megawatt facility that feels like leaprogging a lot of the one megawatt clusters that we've seen uh talked about whether it's Colossus 2 or what Meta's doing with Prometheus and what OpenAI and Enthropic are doing like
2:15:54
is that the intention or do you think that you're more just like catching up to the frontier and they will be coming online with similar capacity around the same time and you're differentiated on the type of model that you'll train. So
2:16:08
So there's big headlines of gigawatts of power and and we had one of those yesterday.
2:16:13
Uh and we have an incredible amount of power on that land.
2:16:15
We actually have six gawatts of gas that sits there that we can >> that that we bring turbines online for it to turn into electricity.
2:16:24
>> But what I think really matters in as a foundation model company is your lead time from when you need compute to scale to how long it takes to get online. >> Mhm.
2:16:32
>> Mhm. And this is where the partnership with coreweave is really interesting because with coreweave as the tenant in the data center we have the ability to determine ahead of that data center coming online how much of that compute goes to scaling pool side >> and how much of that compute we might have to put in the free market
2:16:50
>> and and they're of course world class like I mean I cannot speak highly enough of their ability to operate like large scale compute >> and so for us it was it was not about the big headlines but it was about building a company that could incrementally deliver data centers And the first 250 megawws are actually delivered in a quite unusual manner. >> There wasn't much about this in the
2:17:07
>> There wasn't much about this in the press cuz it's kind of a geeky topic, >> but I think here we like geeky topics of course.
2:17:13
>> So if you look traditionally at data centers, there are these big sticku buildings.
2:17:17
Everything comes on site is manufactured and put together on site.
2:17:22
Uh and that's been the vast majority of the data center industry.
2:17:24
But mobilizing large workforces and and dealing with that complexity means that your ability to scale is not really incremental.
2:17:31
I can't add an extra thousand GPUs, another thousand, another thousand.
2:17:36
>> And so we took a different approach here where we've got the big stick build that we're building, but it's a long corridor and we're bringing around essentially data halls of of 2 megawatts at a time.
2:17:48
>> Uh current generation GPUs would be be a thousand GPUs.
2:17:51
Next generation that's less because they're more dense in power.
2:17:53
But what we're doing is we're doing off-site manufacturing.
2:17:58
So a data center for GPU compute is effectively three layers.
2:18:01
It's an electrical skid, it's a cooling skid, and it's a comput skid. >> Yeah.
2:18:07
>> And you actually designed them that they fit on the the back of a flatbed truck.
2:18:12
>> And ah Mark, good to see you.
2:18:18
>> We have Mark Benny off.
2:18:18
Uh we can go way deeper with you, but thank you so much for hopping on the show.
2:18:21
We will talk to you later and we love you back very soon.
2:18:26
>> Thank you so much for joining. How you doing Mark? Good to see you.
2:18:30
Congratulations on all the fantastic news.
2:18:33
>> Thank you so much for joining the show. How are you today? >> We don't have audio.
2:18:40
>> Let's make sure that we have audio for Mark Benoff from Salesforce.
2:18:42
He is the CEO of Salesforce.
2:18:44
Let's bring him in to the show.
2:18:47
Thank you so much for joining.
2:18:49
Um, our team will sort this out in just a second. We can see you. We can't hear you. We are working on this.
2:18:54
Um, let's make sure that uh he is in the reream waiting room joining.
2:19:01
>> I wish I wish we we got to get ESO back on the show. >> I see him. I don't hear him yet. I would >> ESO.
2:19:07
>> Hey, how you doing, Mark? We got audio. Thank you so much. >> Got audio.
2:19:11
>> You have John Kugan and Jordy Hayes from TVP.
2:19:14
AGI is here, but we still have technical tech, but here we are. >> It's AGI. It's amazing the AGI.
2:19:19
I mean, it's unbelievable how far we've come.
2:19:25
>> I mean, all of the AI is, you know, is it isocond or isocond? I can never remember. >> I think it's >> Isoc.
2:19:32
Is it miso soup or miso soup?
2:19:35
>> I think it's miso soup. >> Okay. It's okay. So, isoc. Yeah.
2:19:36
I love by the way and I love poolside.
2:19:41
I'm so glad you guys know each other.
2:19:44
>> One thing you need to know.
2:19:44
One thing you need to know, >> I also love Poolside, the music, the Have you heard the Poolside FM?
2:19:51
>> And they do that great Neil Young cover, you know, on Harvest Moon. >> Yeah. Yeah.
2:19:57
>> If you haven't heard the Poolside cover on Harvest Moon. >> Yeah.
2:20:01
>> And Esokan and I were talking about that and I said, "Why why did you call it Poolside?"
2:20:05
He's like, "Oh, this and you know, he's building this huge data center now." >> Yeah. Yeah. Yeah.
2:20:09
>> And I said, "You should split the company, Eso.
2:20:10
you should have a software company building your model which is amaz have you seen his model is amazing. >> Oh yeah.
2:20:16
>> And then he's also building a d huge data center >> and I'm like you have one company which is the model poolside one d one company is the data center curbside >> so you have poolside curbside >> that'd be fantastic that'd be fantastic.
2:20:34
Uh anyway please give us give us the update how you feel. Congratulations. Break it down for us.
2:20:39
Dream Force feels like something that would never never get old. >> Super Bowl for sass.
2:20:43
>> Well, you guys aren't here, which I guess you're not Metallica fans, which is sad.
2:20:47
And I'm going to have to let Larriage know that you missed it.
2:20:48
And then also, you don't like Benson Boon either.
2:20:52
So, you're not pop or heavy metal. What are you?
2:20:54
>> Count us in for the next one. We'll be there. >> We'll be there.
2:20:56
We'll be there next time. Super Bowl of >> Why? Why? Why?
2:20:58
I mean, this is our 23rd Dreamforce.
2:21:00
You haven't been to any of them. I want an explanation. I want to know why.
2:21:05
>> No, you >> you guys are the number one podcast in the world.
2:21:07
I want to know why you're not at Dream Force and I want to know your music ch I want to know your music choice. >> I do love Metallica.
2:21:13
I grew I grew up on Metallica and I believe it.
2:21:15
What song do you What's your number one Metallica? Enter >> Sandman. No, >> no.
2:21:24
It's okay to like the >> Everybody knows that one. >> Let me ask.
2:21:27
It's It's It's uh >> What do What do you got? I'm forgiven.
2:21:35
>> I I do like Master of Puppets. That's a great one.
2:21:39
Uh, but I I was always in a little bit more of the slip knot, a little bit more of the of the corn and the tool, but you know, but I mean, Metallic is great.
2:21:48
It's definitely in the heavy rotation.
2:21:49
It's on my it's on my metal playlist a bit.
2:21:52
>> We not only have the greatest heavy metal band of the United States here, Metallica, really the Bay Area, Lars and Rob Trillilo >> and, you know, Jimmy Hedfield was here and um and Kirk Hammet, probably the greatest guitarist of all time if you like that.
2:22:05
But but >> we also have here Yoshiki. Oh yes.
2:22:10
>> The head of XJapan who's also one of the greatest drummers of all time and X Japan also.
2:22:15
He does classical piano and he's about to go on tour again.
2:22:19
Matt back to Madison Square Garden and Royal Fantastic Albert Hall. Amazing.
2:22:23
But anyway, hopefully you guys will come next year >> and works for and works for X the everything app. That's correct. Right. >> That's great.
2:22:31
Well >> X X the everything app.
2:22:32
This is the different X different.
2:22:34
This is extra pan, not related to rel. Okay, got it. Got it.
2:22:39
I I I saw I saw your post about being on X and I wasn't sure if it was a different thing. >> This is We Are we X? >> Yes.
2:22:46
It It's never been more >> Japanese.
2:22:49
Your Japanese viewers are going to go up if you just say we are X, which is Yoshiki official on Twitter. Yeah.
2:22:57
Fantex or whatever it is.
2:22:58
>> We're This This is so fun.
2:22:58
I uh we got to we got to do this more often.
2:23:01
Um, we were joking on the show earlier.
2:23:04
We were referencing a post, you know, the whole like nothing ever happens investment meme, right?
2:23:09
Which is like uh which somebody was highlighting that OpenAI runs on both Salesforce and Slack.
2:23:17
Wanted to ask you does not does nothing does nothing ever happen? >> Nothing else matters. >> Nothing else matters. >> There you go.
2:23:25
>> Um, >> you're still unforgiven by the way.
2:23:27
>> A super intelligence would use Salesforce and Slack for sure.
2:23:29
that that that would be the first thing that they pull off the shelf.
2:23:34
>> If you are a Slack user, you need to see the Slack bot that we introduced here which is built on Enthropic. It's also amazing.
2:23:40
So, we use OpenAI, we use Enthropic, we use Gemini.
2:23:42
I just did an interview >> uh with Sundar, you know, you can see it on YouTube. We XAI love Elon.
2:23:46
Yeah, we we love them all.
2:23:51
We love all our children very equally here. >> Yeah.
2:23:53
H how do you think about vertical integration?
2:23:55
I mean, it's a fascinating company.
2:23:56
You've had the opportunity to build your own cloud.
2:23:58
I I don't believe you've ever built your own operating system, but many companies now are saying we want to build our own ASIC and you've probably I don't know, have you ever thought about putting Salesforce on an ASIC or Slack on an ASIC?
2:24:10
Like how do you think about when to verticalize versus when to partner?
2:24:16
>> It's such a great question and you probably will enjoy my interview with Sundar because I asked him that because you know, you think about it, he's really got it all together.
2:24:22
He's got the data center, he's got the chip, right? Tensor, right? Yeah.
2:24:26
Tensor, right? Yeah. He's got he does a lot of the operating system type work including the model you know the Gemini model the applications he does as well and he even gets all the way to the robotic layer right so here he's got these Whimos running around the street
2:24:42
>> what other company is going from robot to chip you know to data center really only I mean and if the future is >> well what is the future is it billions of fast food restaurants on the planet you know all having their hamburgers made you know by robots And that's the world we're about to enter into. I mean,
2:24:58
I mean, I have no idea, but the one thing I know is uh if you look behind me, there's no robots walking around.
2:25:05
Well, there might be one, but not making the burgers here. >> Not yet. Not yet.
2:25:09
Uh, how do you think about the different foundation model brands?
2:25:12
Do your customers want to know that they can pick Claude?
2:25:15
Is model switching important or is there a world where you wind up wrapping that at an abstraction layer and you're just serving them your product and it's powered by the different foundation models but the the consumer doesn't really care.
2:25:32
>> This is a great question because number one you may know that Salesforce also has a huge research team.
2:25:36
We invented prompt engineering the first prompt right here at Salesforce research commercialized by others of course >> but also the first prompt Absolutely. >> Gone.
2:25:50
>> Thank you for your service. >> Thank you. Thank you.
2:25:53
>> And listen, >> yeah, >> look, some customers want to use our models. >> Yep.
2:25:57
>> Some customers want to build their own models. >> Yep.
2:26:00
>> Some customers like a certain brand of model.
2:26:02
By the way, some countries like a certain model. You just saw ISO Kant. Was it ISO or ISO? ISO, right? >> ISO. >> ISO K. He's in Portugal.
2:26:10
He is going to make a play.
2:26:13
He's going to have a data center in Texas, but he's also going to have a data center in Europe. Yeah.
2:26:18
There's Mistral in France.
2:26:18
There's Middle East models. There's Asian models.
2:26:22
You know, Quan from Alibaba.
2:26:22
You didn't mention that, >> right?
2:26:26
There's models all over the world >> and different customers and different geographies are going to all want different models based on where they are, the kind of customer they are.
2:26:38
They may want small models, they may want large models, they want micro models, they want foundation models.
2:26:43
Yeah, >> there's a lot of different models.
2:26:44
So, you have to give people choice.
2:26:46
But what we're going to provide is this incredible platform that we call the Aentic Enterprise. And we have Agent Force.
2:26:54
And Agent Force is powered by all of those.
2:26:57
And you can plug into whatever model you want wherever you are with the data residency, the governance, the compliance, everything you need. Boom. Right into Agent Force.
2:27:08
Because look at >> here at this conference.
2:27:10
you're not here unfortunately and I hope you do come.
2:27:14
We've had 23 of them by the way. But I know. >> But here's the thing.
2:27:17
Listen, >> Jord's barely 23 years old.
2:27:18
One of our one of our team members over there.
2:27:20
No, you're doing podcasts all day long.
2:27:23
You got your podcast all day long.
2:27:25
You're doing this podcast going. >> But listen, listen. I get it.
2:27:31
I know you're too busy for me, but now listen to this. >> Listen to this.
2:27:37
>> No, this is this interview is so fun. We're We'll fly.
2:27:38
We'll fly to Salesforce Salesforce Tower.
2:27:43
>> No, I know you're busy. I get it. I got it. >> You're too busy. >> No, no.
2:27:49
I understand you're not Metallica fans. You're too busy for us. Fine. >> But here's the thing.
2:27:55
I want to just say this to you.
2:27:58
>> Look, there's customers here from hundreds of countries who have flown in from all over the world. >> Yeah.
2:28:03
>> And they all have different needs, different industries. They're different size.
2:28:06
There's small businesses, companies like 0 to 200 employees.
2:28:10
There's 200 to 1,000 employees, 1,000 to 2,000 employees.
2:28:13
Very large enterprise, the Pepsis and the Dells. Michael Dell was here. You're not here. Michael Dell was here. Okay.
2:28:18
Laura Albert, the CEO of William Sonoma is here. You're not here. That's fine. But here's the thing. Hold on. But here's the thing.
2:28:25
Also, it's about governments.
2:28:28
It's about software companies.
2:28:30
You know, it's about all of these people are here because it's a highly diversified market software and these models don't just have to reach consumers.
2:28:39
They have to go through these businesses and go B to B to C. Yeah.
2:28:41
businesses and go B to B to C. Yeah. you know from business to business to consumer and that is actually complicated because you have different governance like I was saying different compliance >> all kinds of different rules based on country based on so it's it's not like you know we're look we're here we're Americans we're here in America you know
2:29:03
that we love our country and we have certain models that are built here >> great okay we love that but here's one more thing there's many people who are not Americans who are here and they have to comply to their laws in their countries in their languages and their governments and we have to adapt >> to those organizations and those countries as well. So you have to think
2:29:25
So you have to think about AI as a global phenomenon.
2:29:27
You have to think about AI as a highly diversified phenomenon and you have to think of a AI as the future. Yeah.
2:29:37
>> And we're just in we're in the current moment now.
2:29:38
We've been doing AI at Salesforce for 10 years.
2:29:40
We did Einstein 10 years ago.
2:29:44
>> You know, obviously AI has been around I think since the 50s, the ' 40s.
2:29:48
>> So, and you know, like the guy who wrote Minority Report and War Games and even Deep Impact. He's our futurist. He's 78 years old.
2:29:57
He worked at JPL on the Apollo mission.
2:29:59
He's one of the writers on all of those things.
2:30:01
>> And he's there, but we're not. He's there. >> Yes.
2:30:04
>> Everybody's here, but you're not.
2:30:06
Everyone is here, but you're not. But here's the thing.
2:30:08
>> Well, this is the largest tech conference in the world.
2:30:09
you know that 50,000 people are here.
2:30:11
The largest vendor led tech. >> Hit that gong again.
2:30:14
Uh Mark, one question for you.
2:30:16
Uh you Salesforce has been hugely acquisitive.
2:30:20
Uh what is your pitch to founders in 2025 that you want to acquire? >> That's great.
2:30:27
>> Like why why is it such an incredible opportunity?
2:30:29
>> We have bought more than 100 companies. You're 100% right.
2:30:31
Over 26 years we've bought more than 100 companies.
2:30:34
We've invested in hundreds of companies. We own 1% of anthropic.
2:30:38
You know, we just sold a company to Google, which is Whiz, the security company.
2:30:43
We, you know, we grew Snowflake and took it public.
2:30:46
We're investor in many companies.
2:30:51
>> Um, so we want to talk to you if you're an entrepreneur.
2:30:52
We want to invest in you, want to help you grow.
2:30:54
We want to bring you here, introduce you to all of our customers.
2:30:57
We have 50,000 customers here to introduce you to who want to buy your product now.
2:31:02
So come here to San Francisco or come to one of our world tours all over the world.
2:31:07
You know, from here we go on the road and we're going to be in every country in the world, every major city and we will come and meet you. We want to meet you. We want to know you.
2:31:17
Salesforce Ventures is our arm that acquires but also invests.
2:31:20
It's a huge part of our business. It's a $5 billion fund.
2:31:24
It's I think doing a 33% IRR.
2:31:28
I don't know exactly the number.
2:31:29
It's probably one of the highest performing funds in the world. >> That's great. Let us partner with you. >> Yeah. >> You know, come here.
2:31:34
You know, these podcasters don't have time. You do. Come here. Be with us.
2:31:38
Customers are here to meet you.
2:31:41
>> Uh >> you're good at sales.
2:31:44
You kind of got to be in the role.
2:31:44
But >> what's the current thinking on uh seatbased pricing versus consumption both at Salesforce or what you advise founders that might be >> Yeah, I think I think people have this obsession with with seatbased pricing right now and and this idea that that AI may may allow people to run.
2:31:59
may may allow people to run. uh we're not seeing it a lot yet but uh uh companies more efficiently but in my view I think companies are smart enough to to understand they buy so even if they're paying on a seat base they're paying for value at the end of the day
2:32:15
and if they're not getting value they'll they'll churn or they'll find another so but how do you think about this conversation around seatbased versus kind of valuebased pricing >> what a great question look at I just subscribed to chat GPT it's a seatbased model you know that you probably paid 20 or $200 I just absorbed. >> Is that what Did you use that by Did you
2:32:31
>> Is that what Did you use that by Did you use that by the way to to make that picture with with uh Sam?
2:32:35
It kind of that that picture you guys >> rock.
2:32:38
I said because we were wearing like conference necklaces and I said take it off.
2:32:43
And Chad GPT goes oh no no no no no copyright infringement.
2:32:45
And then and then Grock said great I'll do that.
2:32:51
And do you want me to put in the photo of the podcasters with you?
2:32:52
I said I'd like to do that also.
2:32:56
>> But you might have sepas pricing like those are that's a good example, right?
2:33:00
You might have all you can eat pricing.
2:33:02
So we have the agentic enterprise license agreement. All you can eat.
2:33:06
>> Don't even think about pricing.
2:33:06
We're going to give it all to you. Don't even worry.
2:33:09
Three, it might be per transaction or per per action pricing because you're more conservative.
2:33:14
You want to know exactly what you're paying per action per transaction or it might be some other model that some kind of flexible agreement.
2:33:22
right now what we're having to do and this is a great question so thank you for asking me the question we have to write pretty much a custom pricing agreement for every single customer and as you know this year we're going to do about $ 43.
2:33:33
1 billion in revenue and and we just gave guidance yesterday that we expect to be able to do $60 billion by fiscal year 30.
2:33:45
So that's, you know, it's still very fast growing, very exciting, >> fantastic.
2:33:50
>> But when you have, we have 150,000 maybe customers on our core.
2:33:54
There's a million customers on Slack.
2:33:56
Hopefully you guys use Slack. >> We do, of >> course.
2:33:59
>> When you have that many customers and that many countries and that many size companies, like I mentioned, it's across the board, it's every possible need.
2:34:06
And so we have to adapt more than ever.
2:34:11
Before when we started our company 26 years ago, we had one product, one price, $50 a user per month for our sales product.
2:34:18
We only had one product, one price, one type of customer.
2:34:20
That is not who Salesforce is today.
2:34:23
with Slack, with Tableau, with Mulesoft, with our sales cloud, our service cloud, our marketing cloud, all you know, our agent force 360 platform, all of the things that we offer, dozens of products, we ultimately have to adapt to the customer.
2:34:40
And what I can say to you is Salesforce's core values which are trust, customer success which is right what we're talking about right now, innovation, equality of every human being including pay equality and also sustainability which is our trillion tree initiative and all the work we're doing for the oceans and if you were here you could see what we're doing.
2:34:58
But listen, the number one thing number one you can say those are our core values.
2:35:06
At the end of the day it's all about customer success.
2:35:08
Are you happy using Slack?
2:35:10
Are you happy using our sales cloud?
2:35:12
Are you happy using our service cloud?
2:35:14
Is Agent Force fully satisfying you?
2:35:16
Are you getting, you know, you guys have a huge business now?
2:35:18
You guys are growing your business.
2:35:20
Are you go, you know, like, >> by the way, for example, you probably heard of the beast, Jimmy Donaldson. His team is here.
2:35:28
They're bu running on Salesforce.
2:35:30
They're a huge media company now. Everybody know >> Mr.
2:35:32
Beast runs on Salesforce. >> You know, the beast. >> The beast.
2:35:35
The beast running on Salesforce.
2:35:37
uses Slack very aggressively.
2:35:39
>> Of course, >> Jimmy is amazing. >> Okay.
2:35:42
He runs a great business.
2:35:42
It is growing like a weed beast media, all of these things.
2:35:48
>> We have to adapt to him. >> Yeah.
2:35:49
>> It's an example I'm giving you cuz I know you can relate to that in your own business that you guys are trying to grow and make money and have fun >> and look at this look at this great life.
2:35:58
You guys are you probably feel blessed every day that this is your career. >> We do. We do.
2:36:02
>> So, and I feel that myself.
2:36:02
I feel blessed every day that this is my career, that I'm able to do this for a living.
2:36:08
Last night, James Hetfield, probably one of the great singers of our time.
2:36:12
I'm going to introduce you to his music.
2:36:14
Listen, number one thing he said, he's on stage.
2:36:17
He said, "Metallica for over 40 years, gets up on stage and plays that song that you mentioned, Enter Sandman."
2:36:24
And you know what he said?
2:36:26
I feel blessed that I can do this every time I get on stage.
2:36:28
We feel blessed that we're here at Dreamforce.
2:36:32
We're blessed every time we get to meet with a customer, every time we have an interaction, every time we are able to write a piece of code, sign a deal, see customer success.
2:36:40
It's what gets us up in the morning.
2:36:42
It's why we do what we do.
2:36:45
And it's why we're here at Dreamforce.
2:36:46
And after 26 years, I still have the passion, the energy, and the excitement and the vision for Salesforce that I had 26 years ago because that's what makes me happy.
2:36:56
I love making customers successful.
2:36:59
I know that's what you like.
2:37:00
You love making your your viewers.
2:37:04
>> You're the final boss of Enterprise SAS. I absolutely love it.
2:37:06
What makes uh why why do you believe that sales is still a great career path for young people?
2:37:12
I personally believe it is.
2:37:14
I believe it is the most valuable skill in the world.
2:37:19
Um, and but I think a lot of people see all these AI agent for sales rep companies getting funded and they they might think I don't want to go down that path because I don't want to go down the sales path because I think that's going to get automated away.
2:37:31
Why do you what's your kind of worldview there?
2:37:35
>> Well, it's a great question.
2:37:35
I mean, you know, at one level, there's between 20 and 100 million people, we actually counted them that we have not been able to call back since we started Salesforce 26 years ago.
2:37:48
Between 20 and 100 million people, we didn't call them back.
2:37:50
Not because we didn't love them, not because we didn't like them. We didn't have a people.
2:37:56
>> And you know, those are that idea to be able to call everybody back.
2:37:59
You know, >> how do you do it?
2:38:01
Like you can't call every listener back who's already contacting you. You know that. Yeah.
2:38:06
>> So, one thing is, yeah, you're going to have a SDR, sales development representative, the ability to call everybody back, okay?
2:38:12
The ability to qualify to have a conversation.
2:38:17
>> But listen, face to face sales or facetoface communication like we're doing right here.
2:38:20
Like, by the way, I'm I don't I'm not an AI as far as I know. I'm here.
2:38:24
I'm not biological computer running an LLM. At least I hope I'm not.
2:38:29
>> I feel like an AI would grab a water bottle and hit it like that. >> Very suspicious.
2:38:32
What I want to say, I want to just say this to you.
2:38:37
>> I also just hired somewhere between three and 5,000 more salespeople. Wow. >> Okay.
2:38:41
I'm growing my sales force.
2:38:41
It'll be more than almost.
2:38:43
It's I think I'm going to try to get to 20,000 account executives this year.
2:38:47
That doesn't include systems engineers, managers, the infrastructurees.
2:38:51
We have >> we have 80,000 employees at Salesforce. 80,000.
2:38:55
And about a quarter of them, okay, are just people who are trained in our product, who are designed to help you.
2:39:02
We want to make sure that they can help you be your success.
2:39:08
That is why at the end of the day, I think sales, listening, asking good questions, empathizing with the customer, connecting deeply with the customer, having fun, which is super important for us and for you. I know.
2:39:22
I watch everything that you guys do.
2:39:24
You love enjoying and having a good time.
2:39:29
>> We want to do that, too.
2:39:29
Isn't that what sales is really all about?
2:39:33
>> And tonight, you know, look at the conference is happening, but across the street is the St. Reges Hotel.
2:39:37
And last night, I was there and I'm coming down. It's a long day. It's a long two days.
2:39:40
I haven't slept in two weeks.
2:39:42
Coming down the escalator looking the bar is filled. >> Yeah.
2:39:48
>> And the bar is filled.
2:39:48
And what is happening in the bar, do you think?
2:39:49
And it's not our people who are in the bar.
2:39:53
I looked it was all the customers talking to each other and connecting going more deeply ha you know having that human touch you know because look AI we love AI okay >> but AI it's not the same it's not AI doesn't have a soul it's not that human connectivity it's not you know at our depth AI is not born it's you know AI is made so you know we there's humanity that still needs to be nurtured and people want to buy from people and grown.
2:40:28
>> People want to buy from people. Thank you.
2:40:30
I mean that is what it's all about >> and that's why I think sales actually is and you can see right here my job right now is to sell you to come to Dreamforce but also to look at our products to core sell you our core values.
2:40:46
>> That's my job right now. Right. That is my job. Yeah.
2:40:49
>> And um >> how quickly how quickly can you clock if somebody's going to be great at sales?
2:40:55
Does it take you like 20 seconds, 30 seconds, a minute?
2:41:01
>> We've all seen We just had a great conversation about Mr. Beast. >> Yeah.
2:41:05
>> Would you say on a scale of 1 to 10, the 10 one of the best communicators, most aggressive sales people in the world, where would you put him? One or 10? >> 10. >> 10. >> 10. Yeah. >> And boom. That is important.
2:41:15
And I think that that is and you see like all of a sudden you're watching this thing.
2:41:20
One of the games is happening.
2:41:22
He's buried somebody underground for two years or something.
2:41:26
They're digging the person up. Are you still alive? Whatever.
2:41:27
And then just as they break, they go, "And one more thing.
2:41:32
I've got a chocolate bar to sell you.
2:41:33
And this is the best chocolate bar you have ever tasted.
2:41:36
Let me have the guys coming out of the ground now.
2:41:39
Hey, will you try this chocolate? What do you think?
2:41:41
Look, I've been underground for 2 years.
2:41:43
This is the best chocolate I have ever had.
2:41:45
I mean, he has to be one of the great communicators, but also one of the greatest salesmen I have ever seen.
2:41:50
Is that valuable in the age of AI? >> 100%. >> Absolutely. >> 100%. >> Absolutely.
2:42:00
Where do you get Last question.
2:42:01
Where do you Where do you get your energy? It's off the charts. It's incredible.
2:42:05
>> I'm getting it right now from you guys are obviously totally off charts and I'm just vibing.
2:42:09
I am so tired, so trashed.
2:42:14
Last week I've been on the road non-stop.
2:42:16
I haven't slept for an hour. I don't know where I am.
2:42:18
I don't know what conference this is.
2:42:20
I don't know if that was ESO or myo. >> You're doing great.
2:42:23
I just want to thank you.
2:42:25
>> I wish we I wish we didn't have to fade to black.
2:42:28
>> We'll talk to you soon.
2:42:29
>> You see what I You see what I did there? >> Right there. You googled it. You Googled.
2:42:32
You're still unforgiving.
2:42:32
I still >> I think I have I think I have a clip of me playing a Metallica song. I don't remember. We'll send it to you. >> I'll send it to you.
2:42:38
We'll work on >> Thank you so much, Mark. We'll talk soon. Have a good day. >> Super fun. Bye. >> Uh, >> fire me up.
2:42:46
>> Yeah, the mogging counter. Uh, relentless.
2:42:49
>> We got We got to add that. That's >> the chat.
2:42:51
What a >> having a lot of fun at our expense and I wouldn't have it any other way.
2:42:54
Everyone had a great time.
2:42:56
Thank you so much to Mark Beno for hopping on the show.
2:42:57
Uh, I am the show is in chaos at this point.
2:43:02
The timeline is in turmoil.
2:43:02
Our show schedule is in turmoil.
2:43:04
I believe we have more guests. I hope so. Alice coming in.
2:43:08
>> Bring in someone to talk to. Sorry. Hello. How are you doing? >> Welcome to the show.
2:43:15
>> Sorry for all the chaos.
2:43:15
Uh Mark Benoff really knows how to take control of an interview.
2:43:20
Uh it was a fantastic conversation, but but we are here to talk about you.
2:43:24
So, please introduce yourself and give us the news. >> Yeah, great.
2:43:27
Hi, I'm Alice Bentink and I'm the CEO and co-founder of Entrepreneurs First.
2:43:30
Um and I'm coming to you live from the EF office where behind me I have 40 of the most incredible founders that were pitching at our demo day yesterday.
2:43:38
Um so we had Jack Clark who's one of the founders of philanthropic who kicked off demo day um did an amazing opening talk that included the phrase I'm deeply afraid which I think is is always useful for grabbing lights in AI to uh to share. Yeah, totally.
2:43:56
>> We had 200 attendees, an amazing selection of our previous co-investors and the kind of leading lights of the the Bay Area uh fundraising ecosystem.
2:44:05
Uh partners from Kosla, Andre, Zeta, Susa, True Ventures, Baset.
2:44:08
Um and we had 20 companies uh 20 companies pitching the next generation of largely AI related products. Um and it was wild.
2:44:19
Standing room only, amazing vibes.
2:44:19
Um so yeah, it was a it was a great day.
2:44:22
What are the most common trends?
2:44:24
I mean, we've been to a couple YC demo days now.
2:44:28
We And being actually like in the room and talking to so many companies, you start to pick out uh really clear themes.
2:44:34
Uh what's get us to the frontier.
2:44:36
What's the October 2025 theme or different or even like anti-the >> I think the key theme for this one was really young founders early career founders who are who are obsessively solving some of the most um tricky problems in like old old industries.
2:44:55
So companies like Faction, which are the these two insanely ambitious young guys who are automating the um ordering process for industrial distributors.
2:45:06
Um and it's like where do these guys get these ideas from?
2:45:10
And that's actually part of the EF process is because we work with individuals um pre-team, pre- idea.
2:45:16
We help them go through that process of actually developing the idea from scratch.
2:45:19
Um, and Canal, the CEO, he had actually spent a summer spending eight hours a day hand processing these sort of purchase orders for these big, you know, old industrial distributors.
2:45:31
Um, and he's like, right, I'm going to become a founder and I'm going to make sure that no other intern spends their summer doing this.
2:45:35
Uh, and it's the most incredibly lucrative industry, an enormous industry.
2:45:39
But I would say that's the theme.
2:45:41
You know, another example would be got a great company called um, Cto and what they're doing is it's like VA for FMCG.
2:45:46
So, it's um automating the compliance process, but for um fragrance companies, for cosmetics companies, >> this is something that's holding back $400 billion dollars worth of new products every year because it takes months and months to go through.
2:46:05
>> Well, and and probably a year ago at this point, I had somebody pitch me this idea of something they wanted to do, but they they I believe I don't think they actually went and did it.
2:46:12
Uh but I was super uh bullish on the idea at the time.
2:46:16
What uh what kind of guidance were you giving the batch on uh kind of metrics around what would be compelling from for for the investors that were going to be in the audience around what like what what's the bar to come out of demo day uh with you know a seed round uh uh from your view?
2:46:36
>> So it depends on the company and depends on what they're trying to build.
2:46:37
Um half of this batch had more than $100,000 worth of traction and these are really really young companies.
2:46:44
So because we build all the the companies from scratch, we build co-founding teams from scratch.
2:46:50
We have this really unusual and unique process that uh creates companies.
2:46:54
Um the companies that were pitching at demo day were incorporated often less than 3 months ago.
2:46:58
So you're getting to really really impressive revenue traction within a short amount of time.
2:47:04
I think one of the things we do know that investors are looking for right now is the stickiness of that traction.
2:47:08
Um and particularly because a lot of our companies are selling to older uh industries, the revenue is super sticky.
2:47:15
Um so it's less the sort of the the JC curve revenue where you you know it's here today, gone tomorrow, you got to rebuild it.
2:47:21
Um often our companies getting you know year-long contracts, multi-yearong contracts with these big old companies.
2:47:26
Um so yeah, I think one of the things we're really focusing on right now is is the stickiness of the revenue that that they are creating.
2:47:34
>> Last question from my side.
2:47:34
uh how do you apply and do you have any particular uh application questions that you think uh make your application stand out?
2:47:44
>> So you can apply but honestly we like to find you. >> We'll call us. We'll call you. I like that.
2:47:53
>> Uh one of the ways to think about EF is a little bit like um you know CIA the talent agency.
2:47:57
We see ourselves in the same way.
2:47:59
We want to be the CIA for founders.
2:48:01
we want to be in their um in their corner as their talent agent.
2:48:04
And so we go out into communities across Europe, across um the US and across India.
2:48:10
And we build the relationships and the connections to actually work out who are the individuals that we should be going after.
2:48:16
And because we're going after them at the point where they're pre-comp, they're pre-teen, they're pre- idea, um our job is to work out from their behaviors, uh which are the ones that stand out.
2:48:25
So I suppose some of the behaviors that we um focus on are pace, productivity, their ability to build really fast, but you don't see that in an application.
2:48:35
You only see that from spending an extended period of time with these people.
2:48:38
So our selection process includes a hackathon.
2:48:40
Um it means that we're spending 2 days, you know, 48 hours pretty much straight working alongside them, seeing what they build, seeing how they interact with others.
2:48:50
And honestly, the interacting with others bit, they don't always play nicely with others.
2:48:53
We're not looking for uh for team players.
2:48:55
Um but it's really about the behaviors.
2:48:58
You can't really assess an individual their founder of potential by looking at a piece of paper. >> Yeah. Well, congratulations.
2:49:04
>> Hackathon is a very cool selection process.
2:49:07
>> Hit that gong for the whole batch.
2:49:10
>> Um it's great to meet you and uh hope to have you back on the show soon. We'll talk to you later. >> Amazing. Happy day. Thank you.
2:49:17
>> And we have Eric Suffort from Mobile Dev Memo in the reream waiting room.
2:49:19
We will bring him into the TBPN Ultra Dome.
2:49:21
We are getting back on track. Eric, how are you doing? >> What's up?
2:49:30
>> Sorry to keep you waiting. >> Sorry about that.
2:49:32
>> Absolutely chaotic day. >> I'm uh I'm great.
2:49:34
Uh I've I've enjoyed your guest appearances on podcast, your writing, everything, especially some of your t more recent takes.
2:49:40
Um, but maybe for those who uh aren't super familiar, would you mind giving us like a highlevel just introduction on uh your day-to-day how you describe yourself to folks these days?
2:49:53
>> Yeah, I call myself an independent analyst.
2:49:54
I run the website mobile dev memo.
2:49:56
It's got a blog uh podcast uh newsletter, Slack community.
2:49:59
Uh I wrote the book Fremium Economics and I spent kind of my operating career in the mobile space uh mostly in performance marketing roles and strategy roles.
2:50:08
Um, and now I just kind of blog, podcast, uh, full-time and and invest out of a fund called Heracles Capital. >> Oh, cool.
2:50:17
Do you, uh, uh, it it seems like one of the main, uh, like I don't know, takes you that you're just waiting to cash in on is this idea that OpenAI will be doing advertising soon?
2:50:26
Um can you walk me through like how you perceive the messaging from OpenAI the firm and then how like what led you to believe that it is inevitable that we will see advertising.
2:50:41
>> Yeah, that's my my motus Randi is strong uh strong opinions forcefully stated and uh I've said very forcefully open AI will monetize with ads.
2:50:48
I wrote a piece uh in May.
2:50:52
>> Uh the title was obviously open AI will monetize with advertising.
2:50:54
And and here's what here's where I think the um the sort of like most profound clue has been, right?
2:51:01
So uh Sam Alman in Ben's uh podcast uh last week, Ben Ben Thompson, he said uh he he he he pulled off like a rhetorical slight of hand, right?
2:51:11
He said like, "Hey, look, you know, yeah, we're not really considering ads, >> but I really love what Instagram does. Yeah, those aren't ads. We wouldn't do ads. Ads are attacks.
2:51:19
But I love what Instagram does.
2:51:21
Instagram has surfaced relevant products to me and I've bought products from those ads. Those aren't ads. Yep.
2:51:26
>> That's a separate thing.
2:51:27
>> Those are those are improving my life. That's discovery.
2:51:30
>> That's that improves my life.
2:51:30
Ads by definition uh you know deteriorate engagement.
2:51:34
But but these I I I extract value from right and that was a Spencer Noman moment.
2:51:38
That was the Spencer Noman moment.
2:51:40
It it so Spencer Norman uh Netflix CFO March 2022 Morgan Stanley TMT conference.
2:51:45
He said, "Look, it's not like we have religion against advertising."
2:51:48
It's not in our plans right now, but never say never, but it's it's not in our plan.
2:51:51
It's not like we have religion against advertising.
2:51:52
That was the Spencer Norman moment. That was March 2022.
2:51:55
When did Netflix launch ads? November 2022. Eight months later.
2:52:00
>> Obviously, they were already working on it. Yeah. Right.
2:52:02
And and by the way, I would I would I Yeah, I would dispute that statement that it's not like we we have religion against ads.
2:52:07
They very much exhibited religion against ads for years and years and years up to that point.
2:52:11
But that was a tipping point.
2:52:13
Yeah, that was a tipping point when they indicated to the market that ads are coming. They're inevitable.
2:52:16
I've said ads are inevitable with OpenAI with Chat EBT.
2:52:19
I said yesterday they're inelectable.
2:52:20
I got made fun of for using that word.
2:52:22
I've used it a lot, by the way. But they're inevitable. They're coming.
2:52:25
And I think they've been working on that for some time. >> Yeah.
2:52:28
>> Here's uh What do you think about ads where you pay for something and still get ads?
2:52:32
Because for me, I signed up for some HBO plan a while ago and I thought I was signing up to not have ads and now I I've been a little bit lazy.
2:52:43
I usually watch TV before I'm falling asleep.
2:52:47
>> And uh if when I get like a series of like when I get like four minutes of ads before HBO and I'm paying for the product, that's to me a really bad user experience.
2:52:59
>> Well, I mean, it's a choice.
2:52:59
I mean, there's trade-offs, right?
2:53:01
there's trade-offs, right? I think um you know the here so with any sort of with any with any digital product that has the capability of becoming like having a humanity spanning TAM uh you you you the the sort of uh the imperative is to reduce consumer surplus to as close to zero as you can right because that gives everyone an
2:53:19
accessible and to give everyone an accessible price point right some people the only accessible price point is zero right now for some people now now you have to weigh that against perception right and there's perception risk like Netflix I I I had sort of argued at one point that I want Netflix to go fast, a freeing ad uh freeing uh sorry, free ad supported tier. Yeah. Uh and and just Yeah.
2:53:37
Uh and and just drop the price point to zero and have ads.
2:53:40
I now I've changed my mind on that.
2:53:41
I don't think they should do that.
2:53:41
Um just because it would it would you know it would deteriorate the perception of the brand, right?
2:53:46
They have a they have a a perception of being like a high quality uh you know content studio, right?
2:53:50
And so I think that would be a mistake for them to do.
2:53:51
Um so it's just a trade-off, right?
2:53:53
Do you want to maximize TAM or do you want to sort of uh preserve some other asset?
2:53:56
In Netflix's case, it's the the the perception of being like a high quality service.
2:54:01
I don't think Chat GPT um has that risk, right?
2:54:04
So, I think Chat GPT can introduce.
2:54:06
So, first of all, when we talk about ads in CHPT, we're not talking about ads in the paid tier.
2:54:09
Just want to be clear about that.
2:54:10
We're talking about ads in the free tier, right?
2:54:13
And so, yeah, >> that's kind of what I was getting at.
2:54:16
It's like if I if I'm on the 200 a month plan and I'm getting slammed with ads, like that's going to be people are going to be really mad about that.
2:54:22
But for people that are on the free tier seems very fair.
2:54:28
>> Well, I mean that would create a lot of animosity on the $200 tier.
2:54:29
But I think look, so yes, yesterday uh the Financial Times uh posted some data about OpenAI.
2:54:36
So 800 million users, we knew that 5% are paying, right?
2:54:38
So you got 40 million paying users.
2:54:39
13 billion ARR, but only 70% of that is from consumers.
2:54:42
So when I saw a lot of the anal analysis about these numbers, a lot of them sort of anchored ARPO numbers to the overall 13 billion, but you've got to reduce that to the 70% that are that's consumer.
2:54:51
the rest of it is is enterprise API access.
2:54:55
So really that backs up to $228 annual not ARPO, right?
2:54:59
It's not ARPO, it's ARPU, it's AR ARPU, average revenue per paying user, right?
2:55:04
But if you so if you if you actually uh if you do the calculation for just ARPO, it's it's $11. 38 RPU, right?
2:55:09
So how does that compare?
2:55:11
That's that's less than Metas. Meta is 1365. This is global, right?
2:55:15
It's it's much higher than Snap. Snaps is 287. Pins is 164, right?
2:55:16
But I mean the thing is like they're pretty close to Meta, right? on an RPO basis.
2:55:21
to Meta, right? on an RPO basis. Now those 70% sorry the uh the 95% yeah of non-paying users are essentially freeloading right now the issue with uh openi for that and we saw that in the numbers as well is that inference does
2:55:37
isn't free inference has you know marginal cost of of uh you know of of providing that good and so they have to find some way to close that gap and I think introducing ads to the free tier will not uh meaningfully impact engagement um I think it could be additive to engagement. I think if you
2:55:53
I think if you look at a chatp chat, it kind of feels like a feed.
2:55:57
I think the real challenge with with chat and I've seen like a lot of conversations about well the difficulty is going to be like um the attribution or it's going to be the sort of middleware that uh connects things.
2:56:08
Um I don't think I think that's mostly a solved problem.
2:56:10
I don't think attribution is going to be any different.
2:56:12
Attribution meaning how do you credit oh chat for driving a sale or something.
2:56:15
I don't think that's going to be any different than a traditional social media ad um for e-commerce or like an app.
2:56:20
Um, I think the challenge is really going to be coming up with a format that is that feels native uh for the experience without calling into question the uh unbiasedness of the response.
2:56:31
And that's going to be the real issue with ads.
2:56:32
And I think it's also kind of going to be an issue with instant commerce. >> Yeah.
2:56:36
Walk me through uh when do we get to the line of uh as we're rolling out ad products at OpenAI hypothetically?
2:56:42
Uh when do we get to the okay, there's backlash, that's too far.
2:56:46
It feels like if I see an an just an ad in the Sora feed, that's feels very native to Instagram.
2:56:53
If I see an ad in my Pulse uh news feed, that feels very native.
2:56:56
That seems very reasonable.
2:56:58
But if I ask it for what are the best paper towels and it shows me an ad and it feels like the ads team is actually influencing the editorial, if you could call it that.
2:57:10
It feels like the the >> right now you're getting the average product recommendation from Reddit, right?
2:57:18
And then if they start to put their kind of uh >> they start to kind of tilt the scales. >> Yeah.
2:57:24
What are the landmines in terms of like just describing the nature of the firewall between the results when you ask CHP a a direct question and ads?
2:57:36
>> It's a bright red line.
2:57:36
You can't cross it because once you lose consumer trust, then it's gone forever, right?
2:57:40
If if a customer thinks that the answer is not objective, that's actually influenced by whoever is just paying the most to show the product to you, then you lose the consumer trust, right?
2:57:49
And and actually that's like that's a very specific problem for chat bots.
2:57:51
It doesn't exist for Google search, right?
2:57:53
Because Google uh wants you to click uh with any query, right?
2:57:58
You you you query for something, they want you to click.
2:58:00
Now uh when like sort of everyone's interests are all aligned the advertisers, Google's and the users if that first link is the most relevant well you can kind of measure relevance uh through the price the the sort of modified price that the person pays for the ad right because uh the the sort of the end price that an advertiser pays is actually modified by the relevance.
2:58:21
relevance. So um with with chat GPT though so if you just see a stream of links if the top link is an ad or if it's not an ad if it's the most relevant either way everyone wins right um but those interests are not aligning that way with the chatbot right so I think once you start inserting ads into answers um you're going to call into
2:58:38
question the unbiasedness of that answer and you're going to make users question whether they're actually getting the most relevant information they're just getting the information that was uh that was that was paid for that was paid to be shown but the the other thing is like I so So, a lot of people think that instant checkout uh is like kind of precursor for ads, right? Because well,
2:58:55
Because well, there are some partners now and and they kind of uh they all serve ecom, but like they might serve different categories, right?
2:59:02
So, there's there's probably not too much overlap in the type of products that could be served in any given instance, but there will be at some point.
2:59:08
And once there is a lot of overlap, well, then how do you mediate that overlap?
2:59:11
You run it through an auction.
2:59:12
Well, then it's ads, right?
2:59:12
Um I I actually don't think that's the case.
2:59:16
I don't think that's what what they're what they're trying to do here.
2:59:19
Um, you know, so the issue with that is what we know about instant checkout.
2:59:21
And there it's kind of light on details at this point, but it sounds like the way they monetize that is there's a flat percentage fee that gets applied to the product price, right?
2:59:30
And so that's actually economically suboptimal for OpenAI, right? Why?
2:59:33
Because well, if the uh lowest price item is the the the sort of most relevant and that's what gets shown, then they're going to make less money on that, right?
2:59:40
And so the the issue with that is like what you'd rather do is you'd rather combine the relevance with the bid, right?
2:59:46
And that takes into account the price of the product, right?
2:59:50
That's what the advertiser is willing to pay.
2:59:50
And so if you modify that into like an expected value and you you rank the potential ads on that basis, then that's economically optimal, right?
2:59:58
So open AI would be leaving money on the table by just sort of saying, okay, well, let's do an auction and but like or sorry, like just just with the flat fee and and and and then we're just going to like decide what to show there, right?
3:00:09
And the other problem with that is from like a merchant perspective, you don't have any control there, right?
3:00:14
So the thing is you can't change the price to sort of absorb that margin shock, right?
3:00:17
With ads, you price that.
3:00:19
You price that specifically.
3:00:20
You price the the bid that you're willing to pay that makes sense for you that that allows you to sort of like recover the the ad spend but also make money on on the checkout.
3:00:27
But you can't really do that with uh the instant checkout because you just sort of submit um like a JSON of your catalog, right, with the various price points.
3:00:36
So, I think I think instant checkout might and you know the the point I made in in in my blog post about this when it when it was announced is I think it's probably just a way to bootstrap conversion data for users to ultimately target ads against.
3:00:48
But I don't think that's going to be the ad surface area.
3:00:50
I think the ad surface area and so first of all I think it's it's a mistake to think that ads need to be anchored to the the the content of the query, right?
3:00:57
Like that's that's sort of like a contextual targeting way of thinking, but that's not what Facebook does.
3:01:01
That's not what Instagram does.
3:01:02
When you see an ad on Instagram, it's not because the the previous piece of content you saw was related to the things being advertised to you.
3:01:10
It's all based on what there is a history of of conversions uh for anchored to your account, right?
3:01:15
So, these things don't need to be related.
3:01:18
They don't need to be they don't they don't there's no there's there's no there doesn't need to be like a direct uh relevance connection to the actual content of the chat, right?
3:01:24
It could be totally unrelated except for there's some uh you know some logic that came up with the idea that this ad should be targeted to you because based on previous purchase data you seem likely to buy that thing. >> Yeah. Yeah.
3:01:36
Just it knows that you need paper towels.
3:01:38
We're going to show you an ad for paper towels.
3:01:40
Doesn't matter if you went and searched for you know the history of the Roman Empire. >> Yeah.
3:01:45
We haven't we haven't talked about uh pulse yet either which is another effectively feed that you can just slot >> in whatever you want in between the content and I don't think users would have any issue with that.
3:01:58
>> Well, no, but so so that I mean I think that could be a so so that's what I that's kind of what I meant when I talked about like the format's going to be the real challenge here.
3:02:04
Like I mean and they've got you know uh exceptional people to tackle these problems.
3:02:08
And I think one thing about, you know, OpenAI that people sort of like uh discount is is the number of people that they've hired from Meta, right?
3:02:16
So I mean like when people say, well, they don't have a team yet.
3:02:19
Of course they have a team.
3:02:20
They've been hiring people from Meta for years and years who worked on ads ranking because, you know, they worked on machine learning, therefore they were working on ads ranking. >> Yeah. >> Right.
3:02:27
So they they've they've they've tackled these problems exactly um in their in their career.
3:02:31
their in their career. where I think ads um will uh be very well suited for is or what I think ads will be very well suited for is uh like the app store that they are that they've announced that they're building right so what they said was okay we've got these app integrations you know Zillow that's
3:02:46
great it's actually pretty functional um but and you know a number of others uh Canva Figma but at some point we're going to have you know more app integrations than you can remember and so therefore therefore we'll have a directory well when you have a directory what do you uh what what sort of like natively follows >> ads, baby. >> I love it. Uh the the road all the roads >> I love it.
3:03:05
Uh the the road all the roads end in ads.
3:03:09
>> What about uh on the sorus side?
3:03:09
Do you think they can transition from a creative tool to a actual consumption?
3:03:16
>> It's still number one in the plat in the app store by the way. >> It's fascinating. It's very sticky. >> No.
3:03:21
So, I think I worked in I've I've worked in mobile uh for too long to be like very optimistic about the staying power of an app that's been number one for like a week or even >> Yeah, we had we had this conversation earlier where I was like I still put it at like a I I I would say like a 10% chance that it becomes a content consumption platform that people are spending >> on the Snapchat or Pinterest or Tik Tok. >> Yeah.
3:03:48
Well, I so first of all, I don't know if it was a great idea to spin that out into its own app.
3:03:53
It may have been, >> but the thing is, you know, OpenAI was not short of top downloaded apps, right?
3:04:01
They've got Chat GPT and that was a top downloaded app.
3:04:05
>> Um, the thing is like you look at the staying power of these uh generative apps and it's it's you know they're they're pretty short-lived, right?
3:04:12
they're pretty short-lived, right? I mean like uh you know uh >> Lensa >> what's that >> Lensa you remember that was like the original that's right >> but like you remember DeepS right like you know I mean >> the thing is like you know and then
3:04:28
whichever whichever company releases like the ne the sort of the latest nextG model that usually takes the number one spot but they sort of fall off after you know kind of a short amount of time a really novel idea >> do you think deepseek was was paying for downloads like it didn't seem organic at all. >> Um, yeah, that's I I looked into that at
3:04:44
>> Um, yeah, that's I I looked into that at the time.
3:04:47
I don't actually remember what the conclusion was.
3:04:48
I think they probably were.
3:04:51
>> I mean, they were giving away the first reasoning model for free like a week earlier than OpenAI.
3:04:54
So, there was just like it's the >> but the general public was not like I need to try a reasoning model.
3:05:00
I didn't see anybody outside of our bubble talking about deepseek and yet they were charting number one >> and it just felt totally manufactured to me. >> Yeah.
3:05:10
Well, one uh I don't know if you have a hard stop at two, but we will let you go.
3:05:13
But I'd love to know your thoughts on Tik Tok.
3:05:14
Uh there's a deal obviously in place.
3:05:17
Uh it the the valuation always felt low, but at the same time like competition in that category is pretty aggressive.
3:05:23
YouTube has a very serious competitor.
3:05:25
Instagram reel is a very serious competitor. Now you have Sora.
3:05:28
How are you thinking about Tik Tok in the uh in the ecosystem right now?
3:05:34
>> I mean Tik Tok not getting banned was a foregone conclusion.
3:05:37
I I wrote I write an annual predictions post and my post for 2025 was one of the predictions was that Tik Tok's not going anywhere. That was priced in.
3:05:45
There was no movement basically in Snap stock price um you know or metas.
3:05:50
Um I think uh so so Tik Tok is facing some challenges with it social shopping product. Yeah. Right.
3:05:57
And I think that >> well the challenge is that they're losing it a massive amount of money.
3:06:01
Like it's not really it's like a a retail platform that is massively lossmaking.
3:06:06
Yeah, and they they scaled they scaled back the team there.
3:06:08
I mean, just social commerce in general, I think, is um I don't want to say a dead end, but it it faces challenges in the west.
3:06:15
But I I I so I I don't think there was like any news with it, you know, with this deal being reached.
3:06:18
I I think when people get worried about the algorithm being retrained, keep in mind that's the content algorithm, not the ads algorithm.
3:06:26
So, I I really don't think we're going to see any meaningful change in the ecosystem as a result of this.
3:06:30
I think though because you mentioned, you know, YouTube being a competitor.
3:06:33
I think the big news this week was the deal that got struck between Spotify and Netflix.
3:06:37
That seems like a big deal, right?
3:06:39
Because they're sort of frenemies given the risk that you YouTube uh poses to both their businesses. Yeah.
3:06:45
And so I think it's really interesting that they're teaming up.
3:06:48
Spotify's bringing its video podcast onto the YouTube uh platform. Sorry. Sorry. >> The Netflix platform. Yeah. >> Not on YouTube.
3:06:54
Um >> I think Ben Thompson called it the anti-youtube coalition. Right. Right.
3:07:00
Well, but a similar coalition used to exist uh between you know amongst and at various times with different affiliations between uh Apple, Google and Facebook. Right.
3:07:08
So I think these battle lines are being drawn uh there.
3:07:11
But you know YouTube is a CTV behemoth right?
3:07:14
They you know uh they announced that TV is the largest consumption platform for for the for you know the service.
3:07:20
Um it's it's a behemoth and you're seeing Netflix trying to sort of poach YouTube talent onto its platform.
3:07:26
It brought over Miss Rachel, right?
3:07:26
So I I think there is kind of like a war brewing if not sort of raging at the moment across you know YouTube and Netflix and it's interesting to see this alliance form between uh Netflix and Spotify. >> Yeah.
3:07:38
>> Yeah. Did you have a reaction to Ben Thompson was kind of noodling on this take that uh essentially it's very hard for the second generation of YouTube creators who have built their entire business just on YouTube to then take a Netflix or Spotify deal and get off platform because they're so tied into the ecosystem versus someone who maybe
3:07:56
has subscribers on their own website and then yes they use YouTube as a distribution platform but that's not if they turn off YouTube it's not that big of a deal and so uh the the the the loose pitch as I was interpreting it was that Netflix and Spotify would kind of run out of talent to poach from YouTube because the next generation was so locked in. >> It could be, but it depends on the
3:08:15
>> It could be, but it depends on the structure of the deal, right?
3:08:17
So, when Miss Rachel went to Netflix, she wasn't forced to shut down her YouTube channel.
3:08:22
She could continue to monetize that content.
3:08:24
And actually, she didn't even make any new content for Netflix.
3:08:25
That's why it was such a win for Netflix.
3:08:26
I mean, we don't know what we don't know what they paid her, but she just took this content that was extant and brought it over to Netflix. They packaged it up. Yep. >> into a short season.
3:08:37
Um, and essentially just got paid for, you know, double dipping into existing content.
3:08:42
>> Yeah, that's pretty remarkable.
3:08:44
>> Last question and then we'd love to have you back on because this was an awesome conversation.
3:08:48
Uh, >> how uh how excited do you think the average app developer, business owner in general should be about uh OpenAI launching ads?
3:08:58
Because I know a number of businesses over the years that like we're basically be born because a new ads platform.
3:09:05
Like I have friends company with hundreds of millions of dollars of revenue.
3:09:08
They've told me like we would not be a big company if we didn't if we didn't get to hypers scale on on on Facebook in the early days.
3:09:15
And so I I assume OpenAI will want to make the ads pro product cheap and performant early on to just try you know drive this uh you know huge influx of of volume.
3:09:29
But I'm curious what you think the opportunity will be early.
3:09:33
>> Everyone should be excited about OpenAI launching an ads platform.
3:09:35
This is beneficial broadly for the economy.
3:09:36
Ads are the driver of the internet economy.
3:09:40
Look but for Facebook ads but for conversion optimized Facebook ads the DDC category wouldn't exist.
3:09:44
E-commerce would be much smaller than it was.
3:09:47
There would be far fewer small businesses in this country.
3:09:50
Ads is a growth engine not just for you know ecom or not just for like tech bros for everything.
3:09:55
It is the beating heart of the economy.
3:09:58
Everyone should be excited about this.
3:10:00
Look I remember there was Chimath was uh complaining at one point that like I don't remember the exact number but he's like you know I I I invest in these companies and 70% of the money goes to Facebook ads.
3:10:10
And I wrote a piece saying you should be thanking Facebook because if Facebook wasn't there to absorb this ad spend, those companies wouldn't exist for you to invest in. Right. This is this is >> Yeah.
3:10:20
Also, the companies wouldn't spend the money if it wasn't driving results.
3:10:25
It's not like you just you don't you don't if if a camp if you spend 20, you know, 20 grand on a campaign and the CAC is too just high, it doesn't make sense. You just turn it off.
3:10:33
Like that's the beautiful thing. >> Exact Exactly.
3:10:35
So I mean I think you know there there there's there's a reason to be very optimistic broadly about the uh new opportunity that will be engendered by um just just just by by uh LLM empowered uh content engagement right or AI empowered content engagement.
3:10:54
It doesn't have to be just LLMs.
3:10:55
But but I think like where people get mixed up is they think well you know these ads need to be different.
3:10:59
There's some different way to sort of integrate ads.
3:11:02
Well no the formats need to be unique and they need to feel native but ads are ads right?
3:11:05
If you drive performance and you get conversion uh feedback and you get the feedback loop going and you deliver more value than you were paid, right?
3:11:12
There's compounding there and it grows the economy. So I'm super excited.
3:11:14
I call this commerce at the limit because I think what all and and a lot of this technology is not being applied to the consumerf facing side of things.
3:11:22
It's getting applied to the back end, right?
3:11:25
That's what annoyed me about Facebook's last earnings or Meta's last earnings because people are like, "Oh, look at all all these big projects.
3:11:28
Gem, Andromeda, Lattis drove 3% improvement to clickthrough rate or 2%."
3:11:34
You know how massive that is at that scale, how meaningful, but the other thing people ignore is that compounds.
3:11:39
Every time I spend a dollar and I get more back, what do I invest the next time? Not a dollar, right?
3:11:45
I invest what I got back.
3:11:45
So, it's more and it compounds over time.
3:11:49
I'm going to make more money and it's going to grow over time.
3:11:50
So that two or 3% this quarter will will also exist the next time that money's recycled, right?
3:11:57
And also those products are not static.
3:11:59
They're not they're not frozen uh you know in time.
3:12:01
They will also continue to improve.
3:12:03
So the thing is like these these products and this is all in the back end.
3:12:06
This is not seen by consumers but this is where a lot of the investment is going.
3:12:09
And so people talk about the capex like the capex is delivering real returns.
3:12:13
What are you complaining about?
3:12:14
And and the thing is like that is that's driving real returns but it's driving more spend.
3:12:18
It's driving compounding spend.
3:12:19
And that's where a lot of the investment's going.
3:12:22
So I think there's there's reasons to be excited across a number of of of service areas here and it's not just the consumerf facing side. >> This is amazing. Thank you so much.
3:12:29
I feel like I just got >> let's let's get you back on the schedule next week.
3:12:35
>> We have to have you back. This is incredible.
3:12:36
Thank you so much for stopping by, taking the time out of your day. I learned a lot mostly. I just got fired up. This was incredible. Thank you, Eric. >> Thank you. >> Uh take care, guys.
3:12:44
And congrats on all the success. It's welld deserved.
3:12:46
The New York Times profile.
3:12:47
Mike Isaac, that's the big leagues. We survived. We survived. Thank you. Great to see you. We'll talk to you soon. Cheers. Have a good one.
3:12:56
>> That was one last general intuition.
3:13:02
We got to hop on on with we got to hop on with London in just a few minutes, but Pim is in the ream rating room.
3:13:05
And now he's in the TVP and Ultra Dump. Welcome to the show.
3:13:09
Sorry for keeping waiting. Give us the news.
3:13:14
Yeah, we've raised $133 million. 7 because we're gamers. >> 7. Yeah.
3:13:23
Uh to build um general agents for environments that require deep spatial reasoning. >> 1337 lead.
3:13:30
This is a gamer reference.
3:13:30
Uh tell me about your Didn't you make money as a gamer at like 18 or something like that?
3:13:38
So I uh so I grew up with Tourette's.
3:13:41
So I didn't really do uh much other than playing video games as a kid.
3:13:45
And um built the largest private server on Runescape when uh when I was a teenager.
3:13:49
Did about a million and a half in revenue by the time I were 18 years old.
3:13:53
>> Yeah, it was pretty funny.
3:13:53
And uh >> that's incredible.
3:13:57
I mean people come by with uh you know hund00 million fundraisers all the time but that is that is extremely >> we don't we don't see 133.
3:14:04
>> we don't we don't see 133.7 million seed that is that is special so thank you so much for stay taking the time um let's actually dive into the company tell me more about what you're building >> yeah so um we're building general agents for environments that require deep
3:14:20
spatial temporal reasoning so what this means is like look at drones for instance robotic arms they all ship with game controllers already so there's this general interface for um applications that's already in all the robotics where we don't need to reinvent the wheel. Um
3:14:31
Um uh you know um and so the bet that we're taking is that we can train these these foundation models on uh so much diversity represented in all the gaming data which increasingly looks more uh realistic right as physics engines get better and that we can transfer to novel environments with very very minimal new data.
3:14:49
Um and we're seeing you know we're seeing the scaling for this happening now.
3:14:53
Um it's very very clear we can do this one application that we're going after.
3:14:56
So uh after doing Runescape I worked at Dr. Borders.
3:14:59
Um I spent three years there and um so one of the applications that we're excited about is search and rescue drones which is basically you know decoupling the need for humans to even be observing um uh and you know can basically um uh cover a lot more space a lot more quickly uh if if if you have general agents that can that can navigate these types of environments.
3:15:21
So the future of search and rescue is in your vision like drone swarms that get let's say somebody's on a hike, they get lost, you you'd be able to launch a swarm and cover, you know, 100 square miles in uh an hour.
3:15:35
>> Yeah, I suspect there will still be like VLMs on the other side initially that sort of analyze footage and sort of process information.
3:15:39
But yeah, I think the the bottleneck is just make these things not stupidly run into things to be honest.
3:15:45
It's not uh that's uh so that's what we're focused on right now is just general navigation for novel environments in devices that have gaming inputs represented specifically.
3:15:54
And that's the key, right?
3:15:55
Because you have these inputs that already gamers use to control in all these like very diverse environments inside video games, including things like drones, right?
3:16:04
That then uh that then um the agent only has to adapt to an environment, not a new action space.
3:16:09
Um and so the bet is that that transfers.
3:16:12
What are some other specific categories that that are exciting to you or types of companies?
3:16:19
>> Uh yeah, so so so look at it this way.
3:16:21
Um the model sits in a general action space, meaning that it has an understanding of um how different actions relate uh to the world that it's that it's acting in.
3:16:29
Um you have to make sure that when you transfer it over to the physical world that you've solved for safety, right?
3:16:34
And so what we generally do is we um we deploy agents for instance in video games uh that have a larger action space that can do loads of things because you don't need to solve for safety as much there.
3:16:43
And then when you when when you can verifiably prove that you deeply understand a specific action space such as navigation, right?
3:16:49
Then you start transferring those actions over to um to the physical world.
3:16:53
And we have we have sort of an internal joke that we think maybe simulation might actually be larger than the physical world because there's only one physical reality and there's many simulated ones. I like that. >> Um and so yeah.
3:17:01
So, so our take is so we're already working with game developers to get these um to get these models deployed into their games which which means like much more fun playing video games.
3:17:10
Um and then uh it's also a great flywheel, right?
3:17:13
Because if we can verifiably prove that these models are performing great against game players, then um you know that's that's a great benchmark to also start transferring out to other environments.
3:17:23
uh walk through the usefulness of something like Unreal Engine versus uh I'm going to mispronounce it, Gaussian splatting and then a Genie3 style like generative word world model where you're generating the flame the the frames on the fly.
3:17:36
Uh we've heard about a lot of companies using any all three of those mixing them together using focusing on one or the other.
3:17:44
Do you have are you particularly bullish or bearish on individual one of those technologies?
3:17:49
>> I I thought about this question a lot.
3:17:49
I I went through the depths of trying to actually uh start building my own physics engines to like actually understand this from first principles.
3:17:57
So I'm going to try I might take a little bit.
3:17:59
>> Let's give it up for first principles. Yes. >> Love it.
3:18:03
>> Uh so so um the most computationally difficult things to simulate are high degree of freedom agents.
3:18:11
Uh and that grows exponentially as you have more agents in an environment.
3:18:17
an environment. meaning that at some point in simulation you just have you hit a point where you just have to bet on video transfer right and um and so and also there's then at the same time for instance there's loads of things that we currently do not have video footage of which means you cannot really bet on those things right and so like um
3:18:33
like cells and like you know smaller smaller things that like aren't well represented those type of things right so so you always have to use a combination of the two um you always going to have to use a combination of two the other thing is that when you sit inside um like generative role models like Genie for instance um you're not fully in verifiable domain, right? And so the problem with
3:18:50
And so the problem with to be clear, this is an incredible breakthrough, right?
3:18:56
But but um there's very much use for engines like Unreal.
3:19:01
And I think I don't think actually I think where we end up is it's actually really annoying not to have that like determinism in these world models, right?
3:19:08
Because you want to be in some form of verifiable viable domain.
3:19:09
So I hope that somebody manages to create some form of hybrid architecture like okay so for instance >> like GBT like like you can hallucinate all the text but then you can RL and then you have oh write some Python.
3:19:19
So if you need to do a complex math equation you draw Python right? >> Yeah.
3:19:24
Have you guys ever watched redstone CPU uh videos on Minecraft?
3:19:26
So there are videos on Minecraft where people actually um build redstone CPUs and there's also now they build the types of code.
3:19:33
So my point is, you know, we we might actually end up finding like really interesting um like my my dream is to be able to like simulate a CPU inside a world model where you sort of have some form of determinism and uh and not and again this is not at all possible or near postable today, but my my point being like maybe maybe a world model can Yeah.
3:19:52
>> Uh and so yeah, so so I I think you need both right now.
3:19:57
Uh you want to bet maximally on um uh world models and video transfer for uh hard for things that are hard to simulate and then maximally on simulation for the things that you want to stay in verifi verifiable domain on which is most things in my opinion.
3:20:12
>> That makes a ton of sense.
3:20:12
Uh thank you so much for coming on the show. Congratulations.
3:20:15
Uh just very fun conversation.
3:20:18
We'd love to love to have you back on.
3:20:20
Anything happens in the news that you're excited about, just ping us. We'll have you on. I'm sure.
3:20:26
>> Uh, we'll talk to you soon.
3:20:26
Have a good rest of your We'll talk to you later.
3:20:30
Uh, before we hop off, let me tell you about Wander.
3:20:33
>> Find your happy place. Find your happy place.
3:20:36
Sorry to to let to like to leave you hanging there for a second. >> Oh, you're good. >> Oh, no, I'm fine. I didn't even notice.
3:20:42
Book a wonder with inspiring views, hotel, great amenities, dreamy beds, top tier cleaning, 24/7 concier service.
3:20:46
It's a vacation home but better, folks.
3:20:48
And good luck to Regav in the chat.
3:20:48
He is going to a job interview and we wish him the best.
3:20:53
Leave us five stars on Apple Podcast and Spotify and we will see you tomorrow. Thank you for tuning in. >> Can't wait. >> Goodbye. >> Have a great evening.