CEOInterviews.AI
Start App
Gavin Baker
Managing Partner and Chief Investment Officer, Atreides Management, LP

Technology Is Back, Gavin Baker Live, Math Solved Again, Anduril Laptop, DoorDash + Instacart

📅 Oct 27, 2026 TBPN 130 MIN 9008 VIEWS 612 SEGMENTS · 6 SPEAKERS
01:47 Anduril Laptop 17:16 Apple Unveils New Macbook Pro 21:12 𝕏 Timeline Reactions 22:54 OpenAI Announces Three New Discoveries 36:51 Anthropic Founders Can Name Secret Successors 39:23 China's AI Bubble is Already Bursting 43:34 AI Startup Dating Show Goes Viral 47:57 DoorDash + Instacart 54:24 Harley-Davidson's Electric Motorcycle Stalls 59:57 Miles Arnone discusses his role as co-founder and CEO of Rebuild Manufacturing, a 2,000-person company restoring U.S. industrial capacity through engineering, domestic production, and new hard-tech ventures. He outlines Rebuild’s acquisition strategy,...

What Gavin Baker said

Written from the verified transcript and checked against it. Every figure links to the moment it was said.

Gavin Baker, managing partner and CIO of Atreides Management, said the Anthropic S-1 was an upside surprise, citing only $10 billion burned, profitability, cash generation, and unit economics. He noted half the Fortune 500 has adopted Anthropic, which represents only 0.1% of their expenses. Baker pushed back on the idea that AI spending is concentrated, pointing to Anthropic's 11,000-12,000 customers spending over $100,000 and only 250 accounts above $10 million. He highlighted that open-weight model companies like Fireworks, Base10, Together, Modal, and Cognition are growing faster than OpenAI or Anthropic. Baker said the compute spot market is tight, and the forward curve is in contango, making it expensive to secure clusters even into 2027. He praised Grok as the most impactful AI product for his firm. He also discussed AI's potential in biology, rare earth refining, and upcoming conference plans.

Key takeaways

  1. Anthropic's S-1 was an upside surprise with only $10 billion burned and strong unit economics.
  2. Half the Fortune 500 uses Anthropic, which is 0.1% of their expenses, showing low enterprise AI penetration.
  3. Open-weight model companies like Fireworks, Base10, Together, Modal, and Cognition are growing faster than OpenAI or Anthropic.
  4. Compute spot market is tight, and the forward curve is in contango, making 2027 clusters expensive.
  5. Grok is the most impactful AI product for Atreides, with Baker using 17 Groks daily.

Numbers and commitments

FigureWhat it refers toTypeAt
$10 billion Amount Anthropic has burned metric 1:21:54
half of the Fortune 500 Fortune 500 companies using Anthropic metric 1:21:54
0.1% Share of Fortune 500 expenses spent on Anthropic metric 1:21:54
250 Accounts spending more than $10 million at Anthropic metric 1:25:08
11 or 12,000 Customers spending over $100,000 at Anthropic metric 1:25:31
17 Number of Grok instances Baker uses metric 1:29:55
$10 or $15 a watt Cost per watt for long-term compute contracts price 1:41:08
three, four, or 5x premium Potential resale premium on compute contracts commitment 1:41:08
74% Share of Anthropic compute dedicated to training metric 1:48:29

Chapters

  1. 0:00Anthropic S-1 upside surprise
  2. 1:23:16Enterprise AI diffusion and adoption
  3. 1:26:55Open-weight models and inference clouds
  4. 1:29:55Grok and personal AI agents
  5. 1:33:55Compute market tightness and contango
  6. 1:43:21Biology and drug discovery potential
  7. 1:44:22Tomorrow X Summit and conference plans
  8. 1:53:11Energy, rare earths, and portfolio construction
  9. 2:00:01Public company liquidity and sentiment
  10. 2:01:06SpaceX share distributions and capitalism

Questions asked in this interview

12
  1. 1:21:15We're going to bring him in to the TVP Ultradome and catch up with him. How you doing, Gavin?
  2. 1:21:27But what has your interpretation been this week?
  3. 1:29:05What are you seeing on the Grok side that's actually showing acceleration in that product category?
  4. 1:30:27How have you been grappling with it?
  5. 1:37:06Like, do you think Elon studied that and said like we're not going to make that same mistake, and that's why they're adopting this more short-term contract?
  6. 1:40:48Is that dynamic going to look like they signed a multi-year offtake agreement with a Neocloud and then they want to just sell into the spot market?
  7. 1:46:17What are you hoping to get across?
  8. 1:47:51How do you deal with all the attention, positive and negative?
  9. 1:52:23How relevant or interesting are those markets to you in this world since they do seem adjacent?
  10. 1:54:11Like look at the diamond industry, right?
  11. 1:55:44Like, what does the next year look like from a perspective of like how we get the big dollars into the AI buildout?
  12. 2:00:01And it's not like totally binary where like a very, very small number of people have exposure, like, and it's direct exposure, right?
John Coogan 0:00 ↗
You're watching TVN. Today is Tuesday, October 27th, 2026. We are live. Angle, go back to that other angle. The temple of technology, the fortress of finance, the capital of capital.
We got some... We had a math breakthrough. We had a math breakthrough. We have been trying to calculate. It's really, really hard. The number of days till Christmas. So you take the date of Christmas, December 25th, 2026, and then you subtract the number of days, October 27th, 2026. And it's difficult because the different months have different numbers of days in them. Some of them have 30, some have 31. So it's a complex math equation, but we had a breakthrough. And we have a countdown to show you the results of this mathematical finding. They're saying Fields Medal for this. They're saying huge development. 58 days until business.
Jordi Hays 0:53 ↗
And Nick Asher is 58 years old.
John Coogan 0:57 ↗
Doesn't look a day over 40 personally, but I don't know. Anyway, let me tell you about ramp.com. Time is money, save both. Used corporate cards, bill pay, accounting, and a whole lot more all in one place. We have RAMP CTO coming on the show at 12:10. We also have Gavin Baker from Atreides Management coming on at 12:20. Baconator.
Jordi Hays 1:18 ↗
The Baconator. The Baker. I don't know.
John Coogan 1:21 ↗
And the big question on everyone's minds.
Jordi Hays 1:24 ↗
Yes.
John Coogan 1:25 ↗
As he breaks his silence. Yes.
Jordi Hays 1:27 ↗
Is he generally bullish?
John Coogan 1:28 ↗
Okay. We're gonna get to... We're gonna get to the bottom of it. We're going to get to the bottom. Going to find out.
No, it's great. So yeah, I mean obviously you can watch the show yesterday for the Anthropic stuff. That's what's on the cover of the Wall Street Journal and Financial Times. But there's a couple other stories that are interesting. So, we're going to, you know, continue covering what's going on in the market, but this new project from Palmer Luckey over at Anduril is very, very interesting and very fun. It's a new laptop. So, Anduril and Qualcomm just announced a new American-made laptop called Blue Chip. The first laptop they're saying is going to be available in 2027. That's right around the corner. They're going to build it for the government. And so, you're not going to be able to get one of these personally, unfortunately. But hopefully there's a consumer product in the works eventually because increasingly with AI and stuff it's pretty doable to use Windows, to go back and forth between Windows, Mac, even throw in some Linux. All these different OSs are very fungible, mutable. You can sort of change and do whatever you need to if you need a particular piece of software. Like I always had this where on PC, Windows has amazing window management, kind of in the name, but if you have two screens you can just hit the Windows key and then left and it'll put it perfectly on 50% of the screen to the left, 50% of the screen to the right. And for like years and years and years Apple didn't have that. I used like this third-party plugin called like Moom, move and zoom, and it would kind of do the window arrangement if you were trying to lay out multiple windows at the same time on Mac. Mac has since gotten the feature. But whatever feature is, you know, hot in Windows or hot in Mac, you're just going to be able to vibe code exactly what you want and be able to do it usually, unless it's like some crazy server-side stuff. But in general, I'm very bullish on new hardware, new mutable hardware, and I love the fact that this one is going to be made in America. What's interesting is, you know, Palmer has obviously been on the re-industrialization train for years. Anduril is a big piece of that. ModRetro is also pushing for that. It's part of the thesis, re-industrialization. He's spoken at the re-industrialized conference about this. Clearly cares about bringing manufacturing capacity back to America. And it certainly seems like it's a good time in the age of AI to have a lot of, you know, indigenous supply chains around manufacturing of both consumer and enterprise equipment. But he's quoted as saying it'll be the first time in over 20 years that a laptop was made in America. And I want to fact check that. And I think it's actually worse than what he said. I think he was underselling it. I think it is like way over 20 years. I think the last one, that American-made laptop, went into production like '93. But Jackson over there on our team, who's dressed down today, there's no Loro Piana going on, a little quiet luxury going on in the Ultra, but he... So, this is the Blue Chip. This is from Anduril and Qualcomm. It's built for government, defense, first responders, and public safety workers. So, maybe the way to get your hands on this: volunteer firefighter, you know, volunteer police officer. You do some ride-alongs, they give you the Blue Chip, maybe. I don't know. This is going to be a hot commodity, so any way you can get your hand on it is probably valuable.
Quick breaking news. It is Nollie's in the YouTube chat's birthday. Happy birthday.
Jordi Hays 4:57 ↗
It's great to have you here today. Hope you're having a wonderful... I hope it's the best birthday of your life.
John Coogan 5:04 ↗
Oh, is the birthday too close to Christmas? Is this a birthday that you're like, 'Oh, people are going to start gifting you like combined gifts because they're like, oh, your birthday's so close to Christmas, only 50 days apart.'
Jordi Hays 5:17 ↗
Say that to me.
John Coogan 5:17 ↗
No. To him. 50... He's 58 days away from...
Jordi Hays 5:20 ↗
But I'm... I know. I'm ours.
John Coogan 5:25 ↗
I never realized that about you. That's like, I mean, I obviously knew, but the implications and the ramifications, you know, it changed you as a person.
Jordi Hays 5:34 ↗
Yeah. This may as well be road to my birthday.
John Coogan 5:37 ↗
Yes. Yes. But it's sort of foundational lore, like you were born, you know, so close to Christmas. The spirit is embedded in you.
Jordi Hays 5:46 ↗
The farthest I could possibly be away from Christmas.
John Coogan 5:50 ↗
Oh yeah. So you had to grind.
Jordi Hays 5:51 ↗
So for me, my... 364 days... all I knew... All I knew was that initial road.
John Coogan 5:58 ↗
It's kind of a nightmare. Yeah. Okay. So...
Jordi Hays 6:00 ↗
But it formed me into the man I am today.
John Coogan 6:02 ↗
The first, when you were in the delivery room being born into this world, you had the hardest road to Christmas possible. 364 days till Christmas. That's brutal. I'm glad you overcame that. But it's impressive what you've done.
Jordi Hays 6:15 ↗
Anyways, um... Nollies, thanks for being here with us. Happy birthday.
John Coogan 6:20 ↗
Back to the show. Anyway, let's pull up the Blue Chip launch video because I want to see what Palmer Luckey has to say about this new American-made laptop from his own voice.
Palmer Luckey 6:31 ↗
America built the first electronic computer to win a war. During World War II, the US Army Ordnance Corps needed a faster way to accurately calculate artillery firing trajectories. So, American engineers built the world's first modern computer. And that breakthrough kicked off an entire industry that changed the world. Then, we gave it away. We outsourced the manufacturing of the technology that we invented. Now, we depend on other countries, including our greatest strategic adversaries, to make all of the hardware that runs our businesses, powers our economy, and protects our national security. That's insane. Any country that can't build its own computers can't secure them. If you don't control your technology, you can't control your future. 80 years ago, America pioneered modern computing to defend our nation. Now, we're bringing it home. This is Blue Chip. American designed, American built. Other companies might design their computers in the US and make them here. Yet, we're making parts out of metals that the US knows how to extract, knows how to refine, knows how to process, and knows how to mill at scale at a reasonable price. And that goes for the entire laptop. The same thing goes for the battery technology. We need to design things that survive very, very rough environments. And so what we've designed is a chassis that makes it very, very easy to replace the bottom shelf and even the top shelf, which you can easily integrate things like solar panels, other soft goods, other service insignias, even things.
Jordi Hays 7:59 ↗
Imagine lugging this with antennas.
John Coogan 8:02 ↗
Yeah. When I'm a volunteer firefighter, I'm going to be pulling this out.
Palmer Luckey 8:05 ↗
Blue Chip is America's comeback, and is making sure that the machines we depend on every day are machines we can actually manufacture ourselves.
Jordi Hays 8:11 ↗
Mind if I do? This frustrates Jordi incessantly.
John Coogan 8:16 ↗
This is America's birthright.
Jordi Hays 8:17 ↗
Because he doesn't have one and he wants one.
John Coogan 8:19 ↗
Time to bring it back to the valley. Okay. So, couple problems with this launch video. There was no video of caveman making fire.
Jordi Hays 8:29 ↗
There was no video of the Model T.
John Coogan 8:32 ↗
No brothers.
Jordi Hays 8:32 ↗
No Wright brothers. No Wright brothers and then no moon landing. How did they feel?
John Coogan 8:37 ↗
How am I supposed to know what to feel?
Jordi Hays 8:38 ↗
Yeah. And then no nuclear bomb detonating.
John Coogan 8:44 ↗
Just actual relevant footage to the product that they're launching. Yeah. And a direct history and clear message from the founder and the creator of the product.
Jordi Hays 8:51 ↗
Yeah. It's sort of odd. And really bucking the trend. I don't even know what to think about it.
John Coogan 8:57 ↗
It's contrarian. It might work, but it seems so risky to not include at least a scientist.
Jordi Hays 9:05 ↗
30 to 60 seconds. Yeah.
John Coogan 9:07 ↗
Yeah. Cut really fast, you know.
Jordi Hays 9:09 ↗
Yeah. Yeah. 3 seconds. 3 seconds. 3 seconds.
John Coogan 9:12 ↗
Yeah. Like, I mean, at this point, the rules have been established. You have to copy the Founders Fund vibe if you're launching anything. Doesn't matter what you're launching.
Jordi Hays 9:21 ↗
What was it? 2019.
John Coogan 9:23 ↗
It was 2019 or...
Jordi Hays 9:24 ↗
2020. Yeah. Annual meeting 2020, something like that. Yeah.
John Coogan 9:27 ↗
Mike over there, fantastic. Kind of invented the genre of vibe reel and, of course, mixed a lot of things together in an interesting way that was not just pure tech. There's a whole bunch of other culture stuff in there. But this is, you know, jokes aside, this is a great launch video because it's only giving you the B-roll for what's relevant. So like he's talking about the history of the American-made computer and they show you what the military was up to at that time, and then he talks about the computer that they made and they show you the computer. It's like this is the way you do a launch video. Very, very exciting.
Jordi Hays 10:07 ↗
Yeah. The vibe reel launch video is so dead, you can only bring it back in the year 2050.
John Coogan 10:15 ↗
So by 2050... Yeah. I think it will be kind of a cool nod... Yeah. To the 2020s.
Jordi Hays 10:23 ↗
Sure.
John Coogan 10:24 ↗
To use it. It'll be kind of like an if-you-know-you-know thing and it'll be fresh again. Yeah.
Jordi Hays 10:28 ↗
But for now, simply do not do it. Yeah. For the next...
John Coogan 10:32 ↗
I know why you're so annoyed by the odd years.
Jordi Hays 10:36 ↗
All the knockoff vibe reels.
John Coogan 10:37 ↗
It's because it's beating a dead horse and we love horses around here.
Jordi Hays 10:40 ↗
Yeah.
John Coogan 10:41 ↗
And so it's extremely rude to beat dead horses generally, and this is a dead horse that we definitely don't want to be beating. Let's...
Jordi Hays 10:47 ↗
Don't beat any horses.
John Coogan 10:48 ↗
Don't beat any horses, alive or dead, actually. Yeah. So, in that video, Palmer explains how American engineers built the world's first modern computer, and that breakthrough kicked off the computing industry. According to Palmer, any country that can't build its own computers can't secure them. That's a really good point, especially in the era of potential hardware flaws that could create cyber security risk. We got to secure this. I've heard, I don't know if it's actually like a confirmed story, but that at one point or maybe currently the NSA has their own fab to make chips because they don't want to make computers that run their code that are insecure, because what if a foreign country inserts a back door into the chips that they're racking to run their code? So they want to own the vertical stack, but it's very small scale and it's also not the cutting-edge fab process that is available internationally, but it's like good enough for what they were doing at the time. But I think that might just be a rumor. But if you don't control your technology, you can't control your future. Palmer says 80 years ago America pioneered modern computing to defend our nation. Now we're bringing it home. What do you think of the design?
Jordi Hays 12:02 ↗
I like it.
John Coogan 12:04 ↗
Anything else to say about it?
Jordi Hays 12:06 ↗
I think it's good. I think the strap is good.
John Coogan 12:10 ↗
Okay. The Tyler Cosmo review. Comment in hot. No, I like the fact that it has this like rough texture on the top, but then it has this little cutout for like a little bit more geometric detail and they cut the corner off. I don't know if you can see it, but if you zoom in, there's going...
Jordi Hays 12:25 ↗
And I also think they designed it well.
John Coogan 12:27 ↗
Okay, got it. Thank you. That's fantastic. So, let's go through a quick history lesson.
Jordi Hays 12:33 ↗
No, it would be like... the funny thing is from a retail standpoint, it would be really cool to have just an extremely rugged computer. Like part of what I love about an iPhone is you can really hammer on it and...
John Coogan 12:48 ↗
It's good, right? I mean, it certainly shows it's...
Jordi Hays 12:53 ↗
It definitely is, you know, not doing anything that doesn't happen day-to-day, falling on the ground and things like that.
John Coogan 13:01 ↗
Sure.
Jordi Hays 13:02 ↗
But having like a really rugged, robust personal computer would be cool. Like throwing it across the room, throwing it on the couch, throwing it at you.
John Coogan 13:10 ↗
Like if I have a funny video on my computer, if I could just throw it at you.
Jordi Hays 13:13 ↗
Okay. Yeah.
John Coogan 13:13 ↗
So let's go through the history of American-made computers. Give a listen. Give us a little bit more detail on the history here, how America built them and then lost the capability to build them domestically. So the Osborne 1, founded in California, built in 1981, first commercially successful portable computer. It had a 5-inch screen, smaller than an iPhone, two floppy drives, 64 KB of memory, kilobytes, not even a meg of memory. Now we're in gigs of memory. And it was priced at $1,795. That's pretty expensive. The company filed for bankruptcy in '93, just two years after launching it. But you can see there, you pull that out, people know you're doing business, don't you think? I think that makes a statement. What do you think?
Jordi Hays 14:01 ↗
Oh, I love it.
John Coogan 14:02 ↗
You want a Frisbee computer? What are you laughing at?
Jordi Hays 14:05 ↗
No, I love this. I mean, imagine pulling this out on like a commercial aircraft, you know, in the '80s and just click-clacking away. Great. Yeah, it'd be electric.
John Coogan 14:15 ↗
So, Grid Systems, which was founded in 1979, launched the Grid Compass in 1982. This is the first true laptop-looking device. That other one, it's kind of a box. This looks a little bit more like a laptop, built with a folding screen. It was priced around $8,000 and it was used by NASA aboard the space shuttle. That's a fun fact. Then in the early '80s and '90s, we saw the first mass-market laptops in the hands of consumers. Houston-based Compaq launched the LTE in... Jackson thinking the Texas victory lap straight out of the... in 1989. It delivered nearly desktop-equivalent IBM-compatible PC in a six-pound notebook-sized package with battery power and floppy disks. Then in 1991, Apple launched the PowerBook. And look at Apple's design language back then. This was 1991.
Jordi Hays 15:09 ↗
Wow.
John Coogan 15:09 ↗
'91. Insane stagnation.
Jordi Hays 15:12 ↗
You think this is better than what they got going on now?
John Coogan 15:15 ↗
No. No. I was just like... we were just like, you know, the...
Jordi Hays 15:19 ↗
Oh, getting century was just like, oh, you're watching everything getting so much better. And then for the last like 10 years, it's been like another MacBook.
John Coogan 15:25 ↗
I'm looking at my computer and I'm looking at that and it's like it's a gray notebook.
Jordi Hays 15:29 ↗
It's the same thing.
John Coogan 15:30 ↗
Keyboard, screen. I think that the '90s were like a big leg up in the capabilities and the function, and you get the touchpad, you get the screen that goes all the way to the edges. Like there's another...
Jordi Hays 15:44 ↗
I'm psyched on the '90s, but I agree that in the last 10 years, what happened since then?
John Coogan 15:48 ↗
Okay. Okay. So by the late 1990s and the early 2000s, companies like Dell, Apple, Compaq, IBM were increasingly relying on manufacturers in Japan, Taiwan, and China to actually build the laptops. Some laptops like Dell's notebooks and the IBM ThinkPad T30 was still being assembled in the United States well into the 2000s. Lenovo even reported reopening their North Carolina manufacturing line in 2012 to bring back laptop production, but it never became a large-scale revival of the American laptop industry. But American laptop manufacturing has never returned to the levels it reached in the 1980s and '90s. So Anduril and Qualcomm are trying to change that. More than 90% of the procurement cost for Blue Chip's prototype components already comes from the US or allied country suppliers. Anduril also plans to manufacture laptops in the United States, including circuit board assembly at Arsenal 1, the facility in Ohio. So fun, fun project. Very excited for this. But there are a ton of roadblocks and milestones that they need to hit because the supply chain is so complex that as you go, even if you assemble it here, you need to figure out how do you do the displays, how do you do the memory here, how do you do the chip here. There's a whole stack that's very, very complicated. But of course they are partnering with allied countries as well. And here's the stagnation mobile itself, the stagnation book as you call it. Apple has officially unveiled the 14-inch MacBook Pro with the powerful M6 chip. Pre-orders start October 28th, tomorrow. Has 12-core CPU, 30% faster graphics. It's thinner, lighter, faster, stronger. All the same thing. 24 hours of battery life.
Jordi Hays 17:34 ↗
Changes everything. Looks like a MacBook Pro.
John Coogan 17:37 ↗
Because we're sort of at the end of history there. Same thing with the iPhone. Like the new iPhone looks like the iPhone from 5 years ago. It's a little thinner, lighter, faster, stronger, all that. But basically the same device there. The rumor from G, the Germinator, is that they're doing a touchscreen one, which feels a little incremental to me. I don't know. The touchscreen has been around. Touchscreen laptop has been around for a while. It's mostly just been like a jump scare for BDRs who are rocking a Windows PC with a touchscreen and they can just go like, and you're like, what are you doing? You know, have you ever seen do that?
Jordi Hays 18:13 ↗
Yeah.
John Coogan 18:14 ↗
Yeah. What's, you know, people, some people think all laptops are touchscreens?
Jordi Hays 18:21 ↗
That's true.
John Coogan 18:22 ↗
And it's deeply offensive...
Jordi Hays 18:24 ↗
To my culture.
John Coogan 18:25 ↗
Well, it's going to get more confusing.
Jordi Hays 18:27 ↗
Well, yeah, it's going to get way more confusing because a lot of people are going to see you have the touchscreen.
John Coogan 18:31 ↗
Yeah. Nothing worse than somebody, you know, looking at something on your screen and putting their grubby fingers on your screen. The question is if they do a touchscreen MacBook Pro, will it be a lot easier to clean?
Jordi Hays 18:45 ↗
Yep.
John Coogan 18:45 ↗
Like you know the iPhone, you can see fingerprints, but at least you can wipe it off pretty easily.
Jordi Hays 18:50 ↗
Yeah.
John Coogan 18:51 ↗
And so that's my hope. Do you think a...
Jordi Hays 18:56 ↗
It is going to be... I'm interested to see like do they reinforce the hinge so you can really like properly press on it or is it going to be like your, you know, like...
John Coogan 19:05 ↗
You want like a boxing game where you can punch hard.
Jordi Hays 19:07 ↗
Look what's happening. Look what's happening.
John Coogan 19:09 ↗
Oh yeah, that's actually a good point. Yeah, they're going to have to reinforce it. Otherwise, it'll just be like again... like you're going to see like another wave of like, you know, 10,000 posts like Steve Jobs would be, you know, man, he set the bar too high.
Jordi Hays 19:26 ↗
So annoying.
John Coogan 19:27 ↗
That's the whole point.
Jordi Hays 19:27 ↗
Oh, yes, you're right. Yeah, that's the whole point.
John Coogan 19:30 ↗
Yeah, the bar was set so high that you had decades of the world's best product design and development.
Jordi Hays 19:36 ↗
Every other company, people in human history, oh, this is like a pretty good upgrade. And whenever it's Apple, it's like Jobs would never allow this, blah blah blah. There are some funky stuff in the old Jobs-era Apple portfolio. There are some odd devices, like let's be clear, it's not all... we only remember, we have a little bit of selective memory with the Jobs-era Apple products.
John Coogan 19:59 ↗
Let me tell you about MongoDB. What's the only faster than the AI market? Your business on MongoDB. Don't just build AI, own the data platform that powers it.
Jordi Hays 20:09 ↗
Do you think a hot dog is a sandwich?
John Coogan 20:13 ↗
Uh, no. It's a hot dog.
Jordi Hays 20:15 ↗
What about you?
John Coogan 20:16 ↗
No.
Jordi Hays 20:16 ↗
No. Do you know why people debate that?
John Coogan 20:20 ↗
Because it's like meat in between bread.
Jordi Hays 20:22 ↗
It kind of counts. What if you split the bun so it's two separate buns? You know, if the bun breaks into two, does that count as a sandwich then?
John Coogan 20:31 ↗
No, because no one would refer to that as a sandwich.
Jordi Hays 20:34 ↗
What if you slice the hot dog thinly and assemble it like it's a ham sandwich? At what point does the deconstruction of the hot dog become a sandwich?
John Coogan 20:44 ↗
I think if it's like... Yeah. If it's really thin and it's basically completely flat, there's no kind of...
Jordi Hays 20:49 ↗
White... cylindrical. Yeah.
John Coogan 20:50 ↗
Then you're just slicing hot dogs.
Jordi Hays 20:52 ↗
Um, yeah. Is there someone who had a really strong opinion on hot dog being a sandwich? What if you turn it like this? You know, instead of eating it like this, you eat it like this. No. Okay. Is the iPhone Duo a laptop? No. It folds like a laptop. Have you seen these videos?
John Coogan 21:09 ↗
There's no keyboard.
Jordi Hays 21:10 ↗
Pull up this video from Techie Boost Deals. Apple Duo running a full MacBook Pro style screen. This is not a laptop.
John Coogan 21:22 ↗
You make some good points.
Jordi Hays 21:23 ↗
I'm getting close, right? A couple others. Look at the... There's another, the next video. Person doing full CAD on it. Look at this. Look at this. You're telling me this isn't a laptop? I think it kind of counts as a laptop.
John Coogan 21:38 ↗
Can someone on the team... Yeah... just take one for the team and get it? Get the next one.
Jordi Hays 21:45 ↗
Is it still... So, my question is, is it still really hard to get?
John Coogan 21:49 ↗
I don't know.
Jordi Hays 21:50 ↗
Because obviously, you know, it's been dunked on a million billion times over the last week. And so I wonder if that's causing people to...
John Coogan 21:57 ↗
That's cool.
Jordi Hays 21:59 ↗
That's a laptop, right? Fun. Anyway, I don't know.
John Coogan 22:03 ↗
You love a good circus.
Jordi Hays 22:05 ↗
Yeah, it's a gimmick.
John Coogan 22:09 ↗
Okay. Well, question from the chat: When is Gavin Baker on? He is on at 12...
Jordi Hays 22:14 ↗
In 1 hour.
John Coogan 22:15 ↗
20.
Jordi Hays 22:15 ↗
Yes.
John Coogan 22:16 ↗
PM Pacific time.
So, let me tell you about CrowdStrike. Your business is AI, their business is securing it. CrowdStrike secures AI and stops breaches.
Jordi Hays 22:27 ↗
Hear that? Backing up the truck for CrowdStrike.
John Coogan 22:30 ↗
Yes. Wait, why?
Jordi Hays 22:35 ↗
Their services are in demand.
John Coogan 22:36 ↗
Oh, okay. Okay. Everyone's backing up the truck for CrowdStrike. Yeah. I want a sound effect of CrowdStriking the gong or something like that every time.
Jordi Hays 22:44 ↗
Well, we do that ourselves.
John Coogan 22:44 ↗
I guess we do. I guess we do. Anyway, OpenAI has some additional announcements in the world of math. Dan Roberts says, 'We are announcing three new discoveries in mathematics, computer science, and physics about harmonic analysis, complexity theory, and quantum field theory.' And this was a post that went out and it had to make it through the ringer to actually have the impact that I think it eventually did. Andrew Curran contextualized a little bit more. OpenAI announced three new discoveries in mathematics and computer science: local smoothing and the Kayaka conjecture, the exponential time hypothesis is false. You asked, they answered.
Jordi Hays 23:29 ↗
Exposed for not being...
John Coogan 23:30 ↗
I do not pretend to know this. I'm not an advanced mathematician by any stretch of the imagination.
Jordi Hays 23:38 ↗
Oh. Oh, okay. Break it down for us. Get out the whiteboard. Explain the exponential time hypothesis. Explain the cocky conjecture everyone was asking.
John Coogan 23:48 ↗
But people were excited about this. There's a big long post from Rashab Carnowat who says, 'Many of you don't understand how serious this is.' That's me. So let me break it down for you. In computer science there's a famous kind of puzzle called 3SAT. You are given a long list of yes-or-no switches and a set of rules about how they must be arranged. And you have to find a combination that satisfies every rule. The obvious way to solve it is to try every combination, but each new switch doubles the number of possibilities. So a puzzle with a few hundred switches would take longer than the age of the universe to brute force. So it becomes exponentially complex. For decades, the smartest people in the field found clever tricks that made this a little faster, but every trick still hit the same wall. The time required kept exploding as the puzzle grew. About 25 years ago, two computer science scientists, Russell Impagliazzo and Ramamohan Maturi, working with Francis Zayn, turned that wall into a formal idea. They proposed the exponential time hypothesis, which says this explosion is no longer a failure of human cleverness, but a law of nature: no algorithm, however brilliant, will ever solve 3SAT without the time blowing up exponentially. Nobody could prove it, but almost everyone believed it, and it became one of the foundations of modern computer science. Researchers used it as a measuring stick. If you could show that solving some other problem quickly would also crack 3SAT quickly, you had shown that the problem is hopelessly hard. Hundreds of papers across scheduling, graphs, logic, and much more rest on this one assumption. Now an internal model at OpenAI claims to have broken it. It says it found an algorithm that solves these puzzles meaningfully faster than the hypothesis allows, and it has checked the proof in a computer verification system called Lean. And here's the remarkable part: the very people who proposed the hypothesis, along with MIT's Ryan Williams, are now leading the public review of whether a machine has disproved their own idea. So what happens next? If the disproof holds, a large part of theoretical computer science has to be revisited. Every result that said this problem cannot be solved faster assuming ETH loses its foundation, and researchers will have to work out which of those walls were real and which were simply stopped looking at it. It does not mean everything becomes easy. The new algorithm is still slow in absolute terms, and this is not the famous P versus NP problem being solved, but it means our map of what's possible was wrong in a place where we were very confident about it, and that AI redrew the... If the disproof falls, the hypothesis survives and the field breathes out, but we still have watched an AI produce serious, checkable work on one of the deepest questions in the discipline closely enough that the founders of the idea had to take it seriously. Either way, something has shifted. So there you had a take on this of like, okay, a lot of people are just like, okay, cool math result, right? But like, what's the practical application? And there are some knock-on effects that are interesting. Do you have anything to lay out there or still trending?
Jordi Hays 27:00 ↗
Yeah, my continued take is that anytime you see a breakthrough like this that is like against a sort of an open problem that's just generally expanding humanity's knowledge that isn't exactly immediately commercial, you have to assume there's many other highly commercial discoveries happening in the background that people, both the labs and people individually, are not sharing broadly because they made a discovery and there's an opportunity to monetize it. And so I think that anytime there's like a big breakthrough, my head is just going to everything that's not being shared. Yeah. And that's what I'm actually like.
John Coogan 27:40 ↗
So stuff advances in AI research, advances in bio that need to go through actual testing.
Jordi Hays 27:46 ↗
Yeah. Material science.
John Coogan 27:47 ↗
Material science that might need to be patented at the very least. Advances in cyber security that need to be rolled out incrementally and can't just be sort of thrown over the fence. A lot of these math results, it's totally fine to just get them out into the math community, have everyone engage with them immediately. But every other industry is sort of more unique. Yeah. We were talking to Doug Ollin about like what's the next leg up? White collar work, this type of stuff, like obviously the coding agent boom has fully arrived. But what's next? And I was wondering like how close do you think the models are to being able to like answer that Tyler Cowen, Dworkin debate about like when will you walk into Walmart and see just like a fully AI run business? And then would that live within the labs? Like you could have a big lab say like go and build a business and it could just do that and it could do the market analysis and then it could like create a new product and model it in CAD and send it to a factory and send it to a distribution center and pitch it to Walmart and get it on the shelves and like make the money and run that flywheel. Like it could do that, and if you added up a bunch of those, you might have something economically interesting.
Jordi Hays 29:05 ↗
Yeah. I don't think CPG would be there first.
John Coogan 29:08 ↗
Yeah. Yeah. Maybe Dworkin is correct in the sense that it's just like not the best use of time.
Jordi Hays 29:15 ↗
I can, I mean, in that exact context it would be interesting to think about a retailer running an AI that looks at all the store-wide data and figures out white space.
John Coogan 29:26 ↗
Yeah. Figures out white space and then autonomously introduces products.
Jordi Hays 29:31 ↗
Yeah. But again, it's like how much is that, how relevant is that when Amazon and every retailer sort of always done this naturally? Can you speed it up faster? Can it be more niche? Like who knows?
John Coogan 29:44 ↗
And you have like, you know, effectively the thing always, people are jumping on those opportunities and using the tokens for that like independently. The mathematicians are not like, yeah, let me go risk $10 million to like buy a bunch of compute to like...
Jordi Hays 29:59 ↗
Yeah, churn on this question.
John Coogan 30:01 ↗
So, but yeah, this reminds me of like the vanity metric of like the one person $1 billion company, right? Like...
Jordi Hays 30:09 ↗
It's possible that will happen in some shape or form.
John Coogan 30:12 ↗
It's possible it's already happened. But if you have a billion dollar opportunity, do you care more about like magical fake internet points of being like I'm the one person $1 billion company or do you just want to win?
Jordi Hays 30:23 ↗
Yep. So most people are just going to be like, 'Yeah, well, we can bring this person on and they would be helpful with this or that and we're more likely to win, so let's do that.'
John Coogan 30:31 ↗
Yeah.
Jordi Hays 30:31 ↗
And I think the same thing with the fully AI run company where there's like a strong incentive if something is starting to work to like add more human capital to the equation.
John Coogan 30:43 ↗
Yeah. All sorts of people.
Jordi Hays 30:45 ↗
Again, that's from where we're sitting at today. And there's a possibility that you get such a increase in capabilities that adding humans actually does just slow it down and derail the potential of a certain project.
John Coogan 31:02 ↗
That's true. Well, let me tell you about public.com. Invest across all asset classes. Automate anything with AI agents on trading strategies, risk management, money movements, and monitor markets with AI agents for investing on public. You want to pull up this GPT video?
Jordi Hays 31:18 ↗
Yeah, let's pull it up.
John Coogan 31:20 ↗
He's really hitting a stride. There's a new level to the content.
Jordi Hays 31:25 ↗
Excited for this bright AI future. Let's keep these fun AI vibes going.
John Coogan 31:30 ↗
Absolutely. We are full on embracing these unstoppable AI vibes, knowing that this bright AI future is just getting brighter, and we'll keep soaring forward with full AI.
Jordi Hays 31:39 ↗
Exactly. AI is here to stay. So, we'll keep riding this fun AI wave so we don't get left behind.
John Coogan 31:46 ↗
Absolutely. AI is here for the long haul, and we'll ride this exciting AI wave with full optimism, ensuring we thrive in every step of this amazing journey. We love every part of AI. The most in all possible... AI companionship and of course AI freedom. What are you most excited about?
Jordi Hays 32:04 ↗
I'm beyond excited about AI building better AI because more AI means even more AI progress. And I'm all in for that AI future.
John Coogan 32:11 ↗
More, more, more.
Jordi Hays 32:13 ↗
Yes. Yes. Yes. More AI, more breakthroughs, more wonders. The more AI, the better.
John Coogan 32:19 ↗
History. The memory feature must be so deranged at this point. It's just like this. He's like, I need a newer... I need a new like... and talks like me. It's hilarious. Anyway, back in the AI world, Sam Parr is sharing the journey of Jack Clark over at Anthropic from reporter to Anthropic founder. Mike Morates is another journalist to tech billionaire guy. What a crazy run. Yeah, it's fun because Joe Weisenthal like knew Jack Clark. I think they worked together maybe at Bloomberg or at least overlapped in the journalism world and I think Jack went back on Odd Lots to talk about the journey.
Jordi Hays 33:05 ↗
You think Joe Weisenthal is going to get in the arena?
John Coogan 33:08 ↗
Maybe. I'd like to see.
Jordi Hays 33:10 ↗
And he shipped.
John Coogan 33:11 ↗
I'd like to see. Yeah, he's got to start like a neocloud. I think that's the move for Joe. Or maybe like straight to oil and gas extraction.
Jordi Hays 33:21 ↗
I mean, his band, you know, Light Sweet Crude. Yeah.
John Coogan 33:23 ↗
Yeah, Light Sweet Crude. He should get in the game. Start mining some stuff.
Jordi Hays 33:27 ↗
It's time.
John Coogan 33:29 ↗
What's... to talk about the billboard? Okay.
Jordi Hays 33:31 ↗
Did you figure this out?
John Coogan 33:33 ↗
I had a theory.
Jordi Hays 33:35 ↗
Okay, so there's a billboard in San Francisco and it just shows a dock from, it looks like a MacBook dock of some sort with Claude, what's that? Perplexity, Notion, Grokbot, and ChatGPT. And the post says, 'I can't even tell who this billboard is for.' And I agree. It could be any of these. It could be Notion because...
John Coogan 34:02 ↗
Yeah. So, my theory was Notion because they want to be a collaborative workspace where you have a bunch of agents coming in. They put themselves in the center. They're positioning themselves as an AI company. They also have the most contrast I feel to the other logos. Like the other logos all have like some circular design or a bunch of lines intersecting like the flower and the Claude logo all sort of take the same design whereas the Notion is like a cube, a block, you see the end, it sort of draws you to it. So is that true? Is it Notion billboard?
Jordi Hays 34:37 ↗
I don't know.
John Coogan 34:38 ↗
We don't know.
Jordi Hays 34:39 ↗
We don't know.
John Coogan 34:40 ↗
Well, but it got the people talking.
Jordi Hays 34:42 ↗
Okay. What'd they say? That's us.
John Coogan 34:49 ↗
Well, actually the co-founder says thank you to whoever paid for this. So...
Jordi Hays 34:53 ↗
Yeah, he could be thanking somebody on his team.
John Coogan 34:56 ↗
Okay, maybe.
Jordi Hays 34:58 ↗
But yeah, there's literally, you can narrow it down as like would Anthropic do this? No. Would Perplexity do this? No. Would Grok, SpaceX do this? Absolutely not. Anthropic and OpenAI would definitely never put each other on a billboard together.
John Coogan 35:13 ↗
So you can narrow it down to one. There's one company that would do this.
Jordi Hays 35:15 ↗
Okay. Gemini.
John Coogan 35:17 ↗
Apple would do it.
Jordi Hays 35:18 ↗
Apple.
John Coogan 35:20 ↗
No, like I believe it's Notion, but my second guess would be...
Jordi Hays 35:24 ↗
Is the... No, it definitely wouldn't be. It wouldn't be Apple either.
John Coogan 35:27 ↗
I'm curious if this is just like somebody making... We caught him. Image. You think it's Notion? We caught him.
Jordi Hays 35:31 ↗
Here we go. Okay. Well, that's the end of that mystery, I suppose.
John Coogan 35:36 ↗
Could it be the VC that invested in all those? Oo... is I think...
Jordi Hays 35:40 ↗
Did anyone invest in all of them?
John Coogan 35:43 ↗
Let me... or something, I don't know, like who is... the Ontario Teachers or something. Yeah, I...
Jordi Hays 35:51 ↗
I think Sequoia has invested in all five of those.
John Coogan 35:53 ↗
Sequoia's in all five. Wow. The royal flush. Well, hopefully they should add Figma to that list. Let me tell you about Figma. Agents need a canvas. Your AI agents can now create and modify your Figma files with design system context. Why'd you cut me off from the invest like the best soundtrack? I love that song.
Jordi Hays 36:12 ↗
Just joking around.
John Coogan 36:13 ↗
There's more interesting reporting from the Anthropic S1. Anthropic founders apparently can name secret successors. We were discussing this yesterday. We were like, okay, so there's seven founders that all have the class F shares or something. What was it?
Jordi Hays 36:29 ↗
There's one. Isn't it one class F share?
John Coogan 36:31 ↗
It's a single class F share. Yeah, that is held by Founder LLC.
Jordi Hays 36:36 ↗
Founder LLC and Forbes is reporting that the Anthropic founders can name secret successors. Anthropic's IPO includes a private line of succession, confidential lists of would-be heirs to its founders' voting power chosen by the founders and approved by one another. So, you actually have to go to the rest of the group and say, 'I'm giving this to my son or I'm giving this to my brother or my shrimp spouse, my shrimp, I don't know, whatever. Whatever you want to do, my dog, my domesticated bear, I don't know.' But yeah, I mean, clearly they have thought several steps ahead of what reporting is.
John Coogan 37:16 ↗
This rocks. There's no one... there's no company doing it like this. Hasn't Elon said he's going to give his companies to these kids? I thought that was like a whole thing. Yeah, he's doing hereditary.
Jordi Hays 37:28 ↗
Yeah, but a lot of talk.
John Coogan 37:30 ↗
A lot of talk.
Jordi Hays 37:31 ↗
And let's see if it actually... This is like they're setting... it's much harder for Elon. I think it's going to be a bit harder. Right. These are...
John Coogan 37:44 ↗
Seven live players that decided on a structure.
Jordi Hays 37:48 ↗
Yeah. And of course things could change over time, but this feels like quite different and it's cool to be thinking... it's cool to see a company thinking in centuries.
John Coogan 38:00 ↗
Yeah, it's true. Yeah. Beyond decades. Yeah. What other American dynasties are still... Yeah. The Caner says they're going to give it to a future model.
Jordi Hays 38:10 ↗
Oh. ASI run board. Yes. One share for the final model. They could have some sort of contract between them that in Founder LLC the decisions are actually made by the model, which would be interesting. Who knows? I have a feeling that's not the way it goes though. It doesn't feel like that's the plan. But I don't know. I'm still very interested. Like is Ben Bernanke happy about this? Like he had, you know, he's on the super board. I forget what it's called. Like the trustee board or something like above the board. Like Yasmin is on the actual board and then Bernanke and there's two other like super board members who...
John Coogan 38:52 ↗
And the Founder LLC is kind of like the cool kids table.
Jordi Hays 38:55 ↗
Yeah. So how does the Bernanke board feel about it? I don't know. We'll see. It's a lot of different governance. A lot of different governance. A lot of stuff to untangle for the media. But good luck out there if you're reporting on Anthropic. Fireside Alpha Catrini says the preclinical bottleneck is monkeys, which take four years to breed and are harder to scale than memory supply. There's a huge shortage of monkeys and you can't really... that's the problem. That's probably the most important thing for pre-clinical testing. And I know a lot of people are going to be morally opposed to this, but I'll just say one thing, which is for, you know, the medicine is closer than ever before to a world where it really is... I say closer than ever before, this is a really rough transcript, Fireside Alpha, you should clean this up a little bit, but I guess it's accurate to what was said in the podcast. I know that sounds counterintuitive because it's... well, okay, well are you investing in this company that does animal testing? Well the reason for that, as you get closer to that world, what you need is more data, right? Like you need more experimental data from animal testing and monkeys are irreplaceable. So a bull market in monkeys.
John Coogan 40:08 ↗
We got to get into the monkey business.
Jordi Hays 40:10 ↗
Yeah. You got to get into monkey business. Yeah. Serious. There's serious money to be made in the monkey bottleneck, I suppose. I mean, we did hear reports about the Chinese labs, Chinese monkey labs, I suppose, like going up in price, right? So you remember that. But you were, Tyler, you were saying that the Chinese AI bubble has like popped. What does that actually mean?
John Coogan 40:31 ↗
Yeah. So, Reuters just posted this yesterday, but I think since June, like AI companies broadly are down like 40%. Interesting.
Jordi Hays 40:38 ↗
A lot of the labs that went public like Zai has been...
John Coogan 40:41 ↗
Hurting. Unitree since they IPOed like a few months ago is down like 50%.
Jordi Hays 40:45 ↗
I wonder why. Is it fears of a ban or just the economics aren't looking as good because of the open source competition? I don't know. Also, I mean the market cap across all the Chinese labs has to be a small fraction of the market cap in America. I mean we're looking at almost four trillion across the lab efforts in America, five trillion, and I believe it's like 100 billion equivalent.
John Coogan 41:12 ↗
Yeah, I've heard some people say like in China broadly there's less incentive to kind of help out these companies because like the average person in China just doesn't hold that much, like they don't hold that many equities compared to the US. It's like their company's largely based on real estate pump.
Jordi Hays 41:27 ↗
Yeah. So there's like less incentive to pump.
John Coogan 41:28 ↗
Got to go over to Korea, South Korea where the real pumps happen. Everyone's on leverage. Also, I mean, I wonder if their diffusion story is different in China because they have less white collar work and their economy is more manufacturing driven, which has been harder to diffuse AI into. I'm sure there's a lot of machine learning that has already diffused with like image recognition and automation, but just in terms of like we have a language model that can talk and do spreadsheets and word documents, how useful is that to like grow your GDP? Well, it's going to be way more useful if you do a lot of legal work and a lot of finance work and that's the bread and butter of America. What do you think? Bullish China, bearish China?
Jordi Hays 42:10 ↗
You buy.
John Coogan 42:11 ↗
I just think, doesn't it all come down to like the way fewer people owning stocks... maybe... relative like wealth in China is tied to real estate.
Jordi Hays 42:22 ↗
Yeah. And it's also the government doesn't, the government has no incentive to like sort of really support or pump the market in the same way as the US where so much of like...
John Coogan 42:32 ↗
US wealth is tied to the market.
Jordi Hays 42:34 ↗
Yeah. Yeah. Yeah. Also a lot of international funds were deploying capital in China and have pulled back on that because of capital control. So there's a lot of... it's hard to put together a hundred billion dollar equity raise in China. Surely harder than America where you can be pulling from Europe...
John Coogan 42:53 ↗
dollars, you can be pulling from Canadian pension funds, you can be pulling from the Middle East, you can be pulling from American hedge funds and all sorts of different, and then retail obviously. Yeah. Um, well, where do you want to go next, Jordi? Uh, pull something up and I will tell everyone about Shopify. Shopify is a commerce platform that grows with your business, sell in seconds online, in store, on mobile, on social, on marketplaces, and now with AI agency.
Jordi Hays 43:17 ↗
Thank you guys. Uh, so wives, yes, put the timeline in turmoil. Spec said it's like TVPNYC and Caller Daddy had a baby. Nuke.
John Coogan 43:30 ↗
Um, anyways, uh, so, so here's the thing.
Jordi Hays 43:33 ↗
Yes.
John Coogan 43:36 ↗
Sarah and Vivian, I believe are rage baiting. Okay, I believe they're sort of intentionally rage baiting.
Jordi Hays 43:42 ↗
They talk about this in the episode. They said they wanted to go further with the rage bait and they wanted to call the show, or one of them wanted to call the show Trophy Wives as a joke, but it would have been even more of a rage bait. I think that's back. But I believe they like their strategy's working. It's working.
John Coogan 44:01 ↗
It's working. Uh, maybe a little too well.
Jordi Hays 44:05 ↗
Um, this is so, the stats on this post are so funny. A million views, 450 likes, 160 comments. They somehow got the made with AI tag on it, which I have no idea how you do that when it's a podcast clip. Like it's clearly not AI. The lighting is amazing and the production value is really...
John Coogan 44:24 ↗
We don't know that. Okay, maybe this is some new model that's fantastic. Maybe it's Gemini or the new video model or world model.
Jordi Hays 44:34 ↗
But yeah, I've never seen so many people want a Nuke SF.
John Coogan 44:37 ↗
The Nuke SF meme is really strong. I've seen a lot of these images of the Nuke and every variation of it, too. You get the animated one, the picture, and then you get the full 3D animated explainer video of what would happen if a nuke hit San Francisco.
Jordi Hays 44:58 ↗
Yeah, people are not reacting well outside of the tech bubble. I think inside people will be into it, just like I'm hanging out with some people who are tech insiders and just hearing them talk about it. It's probably like a very relatable conversation for people that live in SF.
John Coogan 45:16 ↗
Yes, that's totally reasonable.
Jordi Hays 45:17 ↗
But then the video is being seen by people all over the world.
John Coogan 45:21 ↗
Well, this is kind of what The Economist said about us. Remember, the headline was, The Economist, uh, TVP. Let me see it. It was, superstar podcasts are Silicon Valley's way of talking to itself. But to much of the world, the sound is jarring. So that's what The Economist said when they talked about us all in and Dwar Cash. And the critique was like a lot of what we're talking about just makes no sense unless you're following tech pretty closely, and this is another example of this. I think that there is an interesting trend in there are a lot of news shows, there are a lot of interview shows. Where is the actual white space? And there's something interesting about taking people that are actually tech insiders because there's a lot of content that's basically take actual tech insiders and then have them talk about something else. So a lot of tech people would rather get their sports news from tech people than from sports people in a week.
Jordi Hays 46:20 ↗
Yeah. So you have The Harry Stebbings Show, you have The Ben Thompson Show.
John Coogan 46:24 ↗
Yeah, yeah, yeah. On Dithering, uh, John Gruber and Ben Thompson always kick it off with like three minutes of discussion about sports. And even though I don't follow sports, I'm like, I'm really into this. This is how I get information about sports. And, uh, you could even how a lot of people get information about Christmas and cars and houses, you know, like this is a real thing. Um, and uh, and there's something there. The interesting thing about the first episode, it was like 30 minutes long and it feels like they went through a lot of biography in a good way. They definitely introduce themselves. You have a full context on their careers. Um, it'll be interesting to see with the second, third, fourth episodes. Do they do more story time from their history? Do they run out of those stories and they do more news stuff? Do they wind up being more techy or do they do more lifestyle content, more family style content? Um, but either way, uh, people are definitely aware of it and that's, uh, and that's 90% of the challenge when you launch a new media product. What do you think?
Jordi Hays 47:24 ↗
Yeah, Halo in our chat says, 'Need that Christmas countdown, then I dip.'
John Coogan 47:30 ↗
There you go.
Jordi Hays 47:33 ↗
Just coming in, right, just for the open.
John Coogan 47:35 ↗
Yeah, yeah, needed that. Thank you.
Jordi Hays 47:37 ↗
Yeah, no, I'm good. Get that anywhere else. You can't Google it. There aren't websites for it. You can't calculate it yourself. There's no apps that would do a countdown on your phone.
John Coogan 47:45 ↗
You can't even ask the models.
Jordi Hays 47:46 ↗
You can't even ask the models. They don't know how.
John Coogan 47:49 ↗
Um, anyway, let me tell you about Cisco. Critical infrastructure for the AI era. Unlock seamless real-time experiences and new value with Cisco. Um, DoorDash is in advanced talks to acquire delivery rival Instacart. The deal could be...
Jordi Hays 48:03 ↗
Let's give it up for advanced talks.
John Coogan 48:04 ↗
Advanced talks. The deal could be completed as soon as this week and value Instacart around $17 billion. It's a pretty great... I don't know where Instacart's been trading. I don't know what its all-time high has been, but uh, I know as a YC company, uh, it's from an era where $17 billion is a grand slam. I know we're in the AI era and we're looking at $60 billion acquisitions and crazy, crazy numbers that get even higher, but uh, really a testament to what the Instacart team built uh, with a delivery network that was highly competitive. Uber Eats, DoorDash, you're going up against a lot of different companies and they built this business to a place where uh, it can, you know, trade at 17 billion. I'm just impressed with that as someone who saw them when they were in YC. Uh, but...
Jordi Hays 48:55 ↗
And you invested, of course.
John Coogan 48:56 ↗
Of course, of course. I definitely backed up the truck. Play the sound effect for me when I had $5,000 in credit card debt and negative net worth. Not investing in Coinbase, not investing in Instacart. Instacart was in my YC bad batch. Like I got the demo. Like...
Jordi Hays 49:11 ↗
What was...
John Coogan 49:12 ↗
Instacart was in my YC bad batch. Like they were at the YC demo day with us and we were at parties with them. I remember just hanging out on a roof in San Francisco with them and uh, didn't write a single diamond.
Jordi Hays 49:24 ↗
Okay, you know a lot about the history.
John Coogan 49:26 ↗
Anyway.
Jordi Hays 49:26 ↗
Pop quiz. What is Instacart's uh, entity called?
John Coogan 49:33 ↗
Oh, that's a good question. I don't know. I can't... It's gonna...
Jordi Hays 49:39 ↗
It's animal themed.
John Coogan 49:41 ↗
Animal themed.
Jordi Hays 49:42 ↗
And Canada themed.
John Coogan 49:43 ↗
Canada themed. Moose.
Jordi Hays 49:46 ↗
Maple Bear Inc. Maple Bear Inc. Okay.
John Coogan 49:48 ↗
And they haven't updated it yet.
Jordi Hays 49:50 ↗
Interesting.
John Coogan 49:50 ↗
So when you Google Instacart Inc...
Jordi Hays 49:53 ↗
It will link you to Maple Bear Inc. doing business as Instacart.
John Coogan 49:57 ↗
Interesting. Which is crazy, crazy. Like there's so many options. Company and didn't just redo that. That's wild. Um, but yeah, this happens all the time with YC companies because a lot of people get into YC with one idea, they incorporate. Uh, back then it was done on a product called Clerky. This was pre-Stripe Atlas. And uh, yeah, they still have Maple Bear Inc. DBA Instacart like in their footer. Wow. Yeah. So, they probably got in as like we're doing some tech company or we're doing some other thing or we're doing software or whatever. And then during YC they might have said like, oh well, you know, we should... it's kind of a hassle going out for groceries. Let's figure that out. And then they built that and got it off the ground. Uh, yeah, it was fascinatingly manual at the start and then of course scaled and scaled and scaled. Uh, so Instacart went public in 2023, had a market value around 13 billion. The shares rose 8% Monday following the report. Uh, DoorDash has a value of around 73 billion. Its shares dropped 6% Monday. Buying Instacart would add a major grocery delivery partner to its growing roster of services. Instacart's core business involves online grocery shopping. I think this makes a lot of sense. Like they are competitors, but they fill slightly different segments of the market between grocery delivery and food and dinner delivery, like actual prepared meals. Um, and they could work together really, really well. Um, and it would allow them to differentiate against Uber. What do you think? Do you think this deal makes sense?
Jordi Hays 51:31 ↗
I have no idea.
John Coogan 51:32 ↗
You have no idea. Okay. I have no idea. Do you think it would get approved by the FTC? Hopefully.
Jordi Hays 51:39 ↗
I think it, but you know, it is closeish to creating more concentration in the market, which I think is what they need right now.
John Coogan 51:49 ↗
Yeah.
Jordi Hays 51:49 ↗
Because there's a lot of competition and there will be lots of other products in this category. Uh, but yeah, it does feel like something where a previous FTC would look at it and say, I don't know. Yeah, let's think about that.
John Coogan 52:01 ↗
Yeah. Anyway, yeah, it just feels like one of those things if it doesn't happen now, there will be value destruction, right?
Jordi Hays 52:08 ↗
I just think that uh, DoorDash is generally sort of an unstoppable force.
John Coogan 52:14 ↗
Yeah, great company.

299 more exchanges in this transcript

Sign in free to read the rest of this interview. No card required.

Sign in to read the full transcript

Cite this transcript

APA, MLA, BibTeX
APA

Baker, G. (2026, October 27). Technology Is Back, Gavin Baker Live, Math Solved Again, Anduril Laptop, DoorDash + Instacart [Interview transcript]. TBPN. CEOInterviews.AI. https://ceointerviews.ai/interview/3149614/

MLA

Gavin Baker. "Technology Is Back, Gavin Baker Live, Math Solved Again, Anduril Laptop, DoorDash + Instacart." TBPN, 27 Oct. 2026. Transcript, CEOInterviews.AI, https://ceointerviews.ai/interview/3149614/.

BibTeX
@misc{baker2026_3149614,
  author       = {Gavin Baker},
  title        = {Technology Is Back, Gavin Baker Live, Math Solved Again, Anduril Laptop, DoorDash + Instacart},
  howpublished = {Interview transcript, TBPN. CEOInterviews.AI},
  year         = {2026},
  month        = {oct},
  url          = {https://ceointerviews.ai/interview/3149614/},
  note         = {Speaker-attributed transcript with timestamps}
}