← This Week in Startups
This Week in Startups
Amazon’s “Age of Efficiency,” LLM distribution, AI wearable worries, and more with Elad Gil | E2197
01:23:34
Loading summary
Transcript235 lines
- [00:00]
A
I think in general, my sense is that overall companies that have been considering going public have now sort of opted to do that sort of route.
- [00:08]
B
Okay, I was asking because there's a great Note in their S1, Jason, you'll love this. They were talking about how they're using their AI powered virtual agent chatbot and they say that our ability to control customer support costs over time, even as volume has gone up, has contributed to an increase in gross margin from 60% in fiscal 24 to 68 in fiscal 25 and even more so essentially they're automating things, Jason. Static team size, cost savings and it's helping the company look more profitable. So kind of an AI story and a fintech story a lot.
- [00:35]
C
The age of efficiency is upon us. Every unit in every company, particularly in startups who are always resource constrained. They are the ones to first use these tools because they're resource constrained. They can save a, they can turn a nickel into a dollar. Startups going to do it now? Big companies like, well, we're spending $3 to get a dollar in value and we can afford to do it because we're sitting on a bunch of cash. There's no imperative, there's no existential dread about money at Apple when you're sitting on hundreds of billions of dollars in cash. So yeah, we'll get to it when we get to it. But at a startup it's quite the opposite. You're seeing that as well, I assume. That is startups doing really fascinating stuff with this technology already.
- [01:17]
A
Yeah, super exciting. And the ramp on revenue for some of these companies is out of this world. Right. I mean, zero to a few hundred million dollars in revenue in two, three years is something that I haven't seen in a very long time. And there's multiple companies doing that. So I think this is the big sea change. And you see it in enterprise adoption. You see it in all sorts of things. The company that I started with, Jared and Eric Wu and others that we mentioned earlier is basically working on this for big enterprise. How can you help big enterprises really make that adoption of AI and what applications should you provide for them to be able to really thrive in this post AI world? So I think there's an enormous amount of stuff to be done there.
- [01:54]
C
Yeah, and it's going to be really interesting. The cynical take on this is a lot of people sampling AI solutions and trialing them and they'll be big churn issues and of course competition issues. And I guess the optimist view of it, which I'm Primarily in that camp is these things are providing massive value at very low price. Therefore, why wouldn't you try it? It's like, you know, come get three hamburgers that are grass fed beef for 20 bucks. It's like, how is that possible? Okay, I don't need to know. I'm just gonna order the burgers. Like, if you want to charge me 20 bucks for something that saves me $2,000 a month. Okay. And it's just wild to see how mispriced some of these products are for the value they're creating.
- [02:41]
B
This week in Startups is brought to you by Squarespace. Turn your idea into a beautiful website. Go to squarespace.com twist for a free trial. When you're ready to launch, use offer code Twist to save 10% off your first purchase of a website or domain. Sentry. Your team should be focused on shipping features, not chasing down bugs. New users get three months free of the business plan, which covers 150,000 errors. Go to Sentry IO Twist and use the code Twist and Pilot. Focus on your product. Let Pilot handle your bookkeeping. Pilot provides the most reliable accounting, CFO and tax services for startups and small businesses. Head to pilot.com twist and get $1,200 off your first year.
- [03:34]
C
All right, everybody. Welcome back to this week in startups. I'm your host, Jason Calacanis. With me, my co host Alex Wilhelm and special guest today. Introduce our guest, Alex.
- [03:44]
B
Our special guest today is the former founder of Mixer Labs and Color Health. He helped incubate Brain and Co. But you probably know Alad Gil as an investor par excellence. He has backed many, many well known names in the technology world. And Jason, I believe, commands the largest single AUM of any solo GP in the world. Elad. We're happy to have you.
- [04:02]
A
Al, thanks for having me on board.
- [04:04]
C
Yeah. What are you working on these days? Before we get into all this crazy news, what's tickling your fancy? What are you excited?
- [04:11]
A
It's kind of a mix of stuff. Yeah, it's a mix of stuff, I guess.
- [04:14]
C
Three things.
- [04:14]
A
One is recently launched a with Jared Kushner, Eric Wu and a few other folks to bring AI to the world's biggest enterprises. So we're really excited about that. Second area, just investing. So a lot of the companies I got involved with early, Perplexity, Harvey Decagon, Abridge, et cetera on the AI side are obviously all seeing real growth, acceleration, et cetera. And then third, there's kind of some side projects I've been working on. Around building large scale monuments, translating books, using AI, a variety of different things. And so there's also more of the societal or philanthropic stuff that I think has been really fun. Amazing kind of those three areas. Yeah.
- [04:50]
C
Well, tell me about this. Translating books with AI.
- [04:54]
A
Yeah, we're calling it Alexandria. The basic idea is that if you look at every society we've benefited since the days of Alexandria, the Egyptian Library, from large scale access to human knowledge and books. And so we're taking a thousand of the most important books that are off of copyright and we're translating them, working in concert with OpenAI and Anthropic and 11 labs and a few other folks on creating a version of the book in every language, audio books in every language. So anybody around the world can stream it, they can download it, they can interact with it. We've actually built a book reader. We'll be doing a launch for that in the next few months. But that's been incredibly rewarding. And we've actually done machine versus human eval where people check do they like more what we've done with translation or what human translators have done. And people prefer the machine translated works by far. And so that's been a really interesting and an unexpected.
- [05:49]
C
What do you attribute to. What do you attribute it to?
- [05:52]
A
I think it is often more modern language, although you can obviously tune that based on your prompts. And then I think often it's more concise and often it's more consistent. You have better qa, especially if you're thinking about a long work or corpus. You can enforce certain things that a human translator may or may not be able to do.
- [06:11]
C
Fascinating.
- [06:12]
B
I want to know how long until you add a zero to that and then another zero? When is it 10,000, when is it 100? And what's stopping you from going to a million? There's a lot of books out of copyright. A lot.
- [06:21]
A
Yeah. I think we decided to start with the things that we thought were books that often should exist in other languages that just have never been translated. Because if you look at it, there's maybe two dozen languages that cover 80% of humanity for even some of the greatest works ever produced. Or at least the consensus from, you know, academics is these are the greatest books ever produced. There just aren't copies in all sorts of languages. You know, 20, 30, 40% of humanity can't access them. So we just thought, let's start with this stuff that's kind of obvious. I think to your point, you could do we have the infrastructure built now so you could Run an arbitrary book through, and if you really wanted to, you could turn this into a service where anybody should be able to upload any book and then basically have this output in terms of every language, every audiobook. We've also been looking at building AI guides to the book so you can start interacting with it, asking questions, et cetera. So I think it's almost like this modern Library of Alexandria. We want to bring humanity's knowledge to the world in an accessible way. So it's been fun.
- [07:21]
C
So have you also gone down like the bit Tensor and distributed computing as related to AI and LLMs Rabbit hole?
- [07:29]
A
Yeah, I've looked at it a bit. What's your conclusion on it?
- [07:33]
C
I don't have one, but I made a small investment in a hedge fund that's looking at all the nodes of Bittensor every day, trying to figure out the different projects, buying the tokens, et cetera. And it feels to me like now that we're in the sort of SAX era of crypto regulation, I feel more comfortable spending a little time there. And I've always felt like distributed computing would be a viable coexistor to cloud computing for certain jobs. And we'll move from that SETI at home kind of, you know, I don't want to say silliness, but hobbyist toy to something more meaningful. And perhaps at some point, even if we're building out too much infrastructure, it could be like Bitcoin and that people will start putting up infrastructure and know they'll make a high return on it and you could get some, you know, I don't know, scale that was not previously considered realistic. In the same way Bitcoin and the number of, you know, people participating in the number of nodes is. Would be hard to comprehend 10 or 15 years ago, right?
- [08:49]
A
Yeah. It's an interesting question because I think in the case of anything that has strong economic viability, compute tends to collapse into centralization. And so, for example, to your point on bitcoining, I think it's like the six or seven largest mining pools are over 50% of Bitcoin volume. And so that centralization in terms of buying of hardware, in terms of energy and negotiation for that. And so anytime you have something with a huge economic output, it seems to almost always centralize because of those economies of scale. And then to your point, there's all sorts of circumstances where you have more distribution, either because it's people who are just generating an extra bitcoin every once in a while because they have some idle capacity, but on a volume basis, that's pretty low.
- [09:30]
C
If you want your business to succeed online, it has to stand out. And that's why you need a beautiful, world class website. The good news is you don't have to hire an expensive designer anymore, nor do you need a bunch of developers. No, Squarespace has all the tools you need to claim your domain and start building your company. Hey, maybe you got a product to sell. Maybe you want to show off some samples of your work. Maybe you've got a new service you're providing. Well, Squarespace is the all in one tool you need for your business to grow and flourish. And they offer beautiful templates. Maybe you want to start an online course or you're scheduling appointments for people, maybe you're generating client invoices. All of that is built into the product. Everything you create with Squarespace is pre optimized to show up in search engines as well. So you don't need need to hire some expensive SEO person. Nope. They're building your meta descriptions, they're setting you up with an auto generated sitemap, and the SEO is done before you even get started. Plus there's a new AI powered feature, it's called Blueprint, which makes it easier than ever before to customize your website and make it really pop. So check out squarespace.com twist for a free trial and when you're ready to Launch, go to squarespace.com twist to get 10% off your first website or domain purchase. That's squarespace.com twist.com.
- [10:49]
A
Or it could be things like SETI at home, which is more philanthropic or very early bitcoin mining or crypto mining, where, you know, there isn't that same liquidity in terms of mining of the token. So it's a really interesting question of when do you transition from decentralized to centralized. And.
- [11:05]
C
It'S an open consolidation that occurs because it's not like you have to ask a gatekeeper. It's not like you have to go to, you know, somebody and say, I would like to participate in this. Can we cut a deal? You're not like negotiating with Oracle or Microsoft. It's like you can just participate and if you have innovations that you can bring to bear, you know, you come up with your own chipsets, you come up with your own power strategies. You have a cousin who works at a nuclear facility in China and they, yeah, they happen to give you a little space for your, for your nodes. All of a sudden you're up to the races. So I'm super bullish on it. Also like if you just think about all the inference that's occurring, it's going to be unlimited demand for this stuff, I think. I don't know that the demand curve ever. Yeah, I don't know if it's going to abate anytime soon. What's your take on it? Like the overbuild versus the dark fire analogy. Do you have a position on it or does all this stuff. Jevons paradoxic sway to a conclusion?
- [12:11]
A
Yeah, I don't have a strong point of view on it. I do think there's some really interesting shifts that are happening. So to your point on the inference side, eventually more and more inference will happen on a device. And so to some extent the distributed compute notion may just be everybody's iPhone in some sense. In terms of at least the inference side. In terms of training, the thing I've been looking at most is the degree to which energy costs will impact where training centers are actually going to get built. And one thing that I think is really under discussed is the fact that given how poorly Europe has governed itself relative to energy, shutting down nuclear power plants in Germany and being dependent on Russian oil, et cetera, it basically means that training data centers are unlikely to go there because it's going to be too expensive to run them just because of energy, ignoring regulation, ignoring everything else. Which means that basically all the training centers should end up in the Gulf and in the US and the Gulf, in part because under the Trump AI plan, basically there's some extension of some rights relative to being able to run training in UAE and a few other places, and those are going to be places with low energy costs. So I think there'll be parts of the US parts of the Gulf where all the training is going to run. Inference is probably more globally distributed in part because of gdpr. So Europe will force it from a regulatory perspective, but they're going to lose out on tons of GDP spend and capex spend because people just won't build training centers there. So I think that's a little bit under discussed right now.
- [13:35]
C
It's quite a paradox because also in the Middle east they're adopting solar and batteries at an extraordinary pace. These are people who, it seems to me the people who understand energy are not being particularly precious about where the energy comes from. In Texas we have the largest installed base of solar, and I can tell you, living in Texas, they're not super precious about where the energy comes from. Is it the cheapest way to do it?
- [14:01]
A
Great.
- [14:01]
C
We can make a business of that. We got a Grid, let's compete. And that's what they're doing in the Middle east, where there's a lot of sun, there's a lot of sun and there's a lot of capital to buy those solar panels. And you're right, their energy costs are going to be so de minimis. And they also have a populace in that region that has been educated in the west over the last 30 or 40 years, very quietly at the top institutions who have now boomeranged back to Qatar, to uae, to Saudi. And then of course you have Israel, which is just a juggernaut per capita in terms of startups. So I'm spending more time in the region. I'll probably spend two weeks, three weeks a year there looking at startups and working with different investors. Are you spending time in the region?
- [14:50]
A
A little bit, yeah. I'm an investor in a few companies, some of whom have been bridged across Israel traditionally. So that'd be things like Trip Actions, not called Navon, et cetera, but then also more recently a company called Eon which is doing data infra one called Kello on the defense side, etc. So I have a handful of investments in Israel and then obviously there's just really fascinating things happening across the entire region.
- [15:10]
C
Yeah, it's going to be really interesting when you can fly direct from, you know, Tel Aviv to Riyadh to Doha to. This is kind of mind blowing in our lifetime that this is happening.
- [15:24]
B
So a lot. You mentioned Navon, which is going public right now. It's currently in a little bit of a limbo with the government because the government shut down down and such. But what can you tell us about the company's road to IPO?
- [15:34]
A
Yeah, I can't really say anything. I mean, usually IPOs have quiet periods. I think in general, I was hoping.
- [15:40]
B
They didn't apply when the government was.
- [15:42]
A
Shut down, but apparently not in general. My understanding of what's happening at a Meta situation is just because of the government shutdown, the SEC for a short period was not reviewing or fully reviewing S1s. And then they actually issued a rule that would allow companies to still go public if they did certain filings and certain disclosures and then within a certain time period they could go. And so I think in general, my sense is that overall companies that have been considering going public have now sort of opted to do that sort of route.
- [16:11]
B
Okay. I was asking because there's a great note in their S1. Jason, you'll love this. They were talking about how they're using their AI powered virtual agent chatbot. And they say that our ability to control customer support costs over time, even as volume has gone up, has contributed to an increase in gross margin from 60% in fiscal 24 to 68 in fiscal 25 and even more so essentially they're automating things, Jason. Static team size, cost savings and it's helping the company look more profitable. So kind of an AI story and a fintech story a lot.
- [16:39]
C
The age of efficiency is upon us. Every unit in every company, particularly in startups who are always resource constrained. They are the ones to first use these tools because they're resource constrained. They can save a, they can turn a nickel into a dollar. Startup's going to do it. Now, big companies like, well we're spending $3 to get a dollar in value and we can afford to do it because we're sitting on a bunch of cash. There's no, you know, imperative, there's no existential dread about money at Apple when you're sitting on hundreds of billions of dollars in cash. So yeah, we'll get to it when we get to it. But at a startup it's quite the opposite. You're seeing that as well, I assume. LAD is startups doing really fascinating stuff with this technology already?
- [17:21]
A
Yeah, it's super exciting. And the ramp on revenue for some of these companies is out of this world. Right? I mean zero to a few hundred million dollars in revenue in two, three years is something that I haven't seen in a very long time and there's multiple companies doing that. So I think this is the big sea change. And you see it in enterprise adoption, you see it in all sorts of things. The company that I started with, Jared and Eric Wu and others that we mentioned earlier is basically working on this for big enterprise. How can you help big enterprises really make that adoption of AI and what applications should you provide for them to be able to really thrive in this post AI world? So I think there's an enormous amount of stuff to be done there.
- [17:58]
C
Yeah, and it's going to be really interesting. The cynical take on this is a lot of people sampling AI solutions and trialing them and they'll be big churn issues and of course competition issues. And I guess the optimist view of it, which I'm primarily in that camp, is these things are providing massive value at very low price. Therefore why wouldn't you try it? It's like, you know, come get three hamburgers that are grass fed beef for 20 bucks. It's like, how is that possible? Okay, I don't need to know. I'm just gonna order the burgers. Like if you want to charge me 20 bucks for something that saves me $2,000 a month. Okay. And it's just wild to see how mispriced some of these products are for the value they're creating.
- [18:45]
A
Yeah, 100%. And some of them are burning a lot of money along the way to get there. So as to your point, the pricing versus the value differential in one or two well known cases are also just leading to odd economics. But I think most of them are very solid in terms of what they're providing, the potential cash flows to support the business, et cetera.
- [19:03]
B
So a lot on that point. We're not seeing a lot of gross margin negative businesses then. Because I've heard bits and pieces about people saying that some companies are so underpriced that they're actually gross margin negative, which is so uneconomic as to make me worry. But. But it sounds like that's not the case too often.
- [19:17]
A
No, I think there's one or two very high profile versions of that. And one could argue in those cases they're basically trading inference for market share. And so you could view it as a marketing funnel.
- [19:27]
C
I've been working with and investing in founders for so long, I've got supernatural startup powers. For example, I know your team is spending too much time on debugging. Your engineers are chasing down bugs at 1am, 8am, and every time in between, instead of building amazing new features or pushing fresh code, just admit it, you'll feel better. The reality is, from time to time, your code is going to break. We all know that. But it doesn't have to be the end of the world. Now they're Sentry. They're the one stop application monitoring platform that's not just going to tell you, hey, something's broke. But what specifically happened? Where? And most importantly, why. Plus they've added Sentry logs. So now when something goes wrong, you get a complete report. No more random jobs that just don't run, no more tab hopping, just actionable information with tons of context. And that means faster fixes. And we want to give you three months free of Sentry's team plan. That's right. F R E E. Just for being a loyal twist Listener, that covers 150,000 errors. Just go to Sentry IO and use the code twist that's S E N T R Y IO and make sure you use that code twist so they know we sent you.
- [20:40]
A
But I think in most cases Most of the companies I know are running at, you know, quite healthy margins and in some cases software plus. One thing I've been investing behind over the last two years kind of quietly is these AI driven rollups. You fund the purchase of services, businesses roll them up and then increase margin dramatically using AI because effectively you're making the people on those teams dramatically more productive. And that's been a thesis that's really been proving itself out quite strongly. There's two companies I'm involved with in particular where we're seeing very strong lift and very strong execution relative to these really big opportunities because services.
- [21:13]
C
Can you say the names of them or are they.
- [21:15]
A
I can mention one of them. The other one's quite stealthy. The one that is non stealthy is called Long Lake and. Or Long Lake Management. So lln and basically clever. Yeah, very good. One of the founders, Alex, is an exceptional PE person. Zach Frankel, who's more technical helped get it up and running and they basically have a really great team of both engineers but also private equity and operations people. And if you think about it, the key for any sort of AI driven transformation is if you're doing a roll up you need a PE person. If you're not, you just need the sort of edict to do it. Then you need somebody to operationally implement it. A lot of the issues with adopting AI is not the AI. A lot of it's actually pretty easy to build given the APIs that we have from OpenAI and Anthropic and others. For Google, it's usually the change management, it's convincing the team to adopt it, it's aligning incentives, it's saying, hey, we understand you really like this tool, but we're switching, you know, it's all that kind of stuff.
- [22:15]
C
Yeah, right.
- [22:17]
A
The resistance will be very real. Yeah. In some cases maybe it's going to threaten that person's job or perceptually do so. So often when you're doing roll ups for AI you need the PE person to be able to do good purchases. You need the ops person to really help revamp the org and then you need the engineers to implement the AI in the workflow. And so very few of the roll up companies that I've seen have this magical convergence of those three skill sets into one team.
- [22:41]
C
Yeah, Athena is in a roll up but they're, you know, people think of them as, and I'm the first investor in that company as well as was the first investor in Thumbtack, Jonathan's other company. People think of it as like assistants, right? Virtual assistants in a different time zone at a different price. But it's really an AI company because every time they can automate a task across 10,000 assistants, if they can help one assistant schedule 5% faster right down the line, if they can help them book hotels faster right down the line, et cetera. And we have an investment in a company called Howie that Austin Peter Smith is doing that. It's just doing one thing, scheduling better than anybody. So there's a lot of people taking swings at this, I guess the company this week of note, the main character, energy, goes to Amazon. Alex, maybe you could tee up the breaking news from the last 48 hours about Amazon and jobs and employment.
- [23:37]
B
Yeah, so the New York Times, yeah, New York Times got its hands on some internal documents from Amazon, the cloud and E commerce giant. These are from the 2024 timeframe, so they're a little bit dated. But essentially the company's investments in automation at the warehouse level in particular are going to help it hire 160,000 fewer people through 2027 and the times.
- [23:57]
A
Right.
- [23:57]
B
So that could save them up to about 30 cents per package that they pick and then deliver. 30 cents is not that much, but if you multiply times billions of packages, it add quite quickly. And all the work they've done over the last 10 to 15 years to build out robots and increase automation in their warehouses will save them about 600,000 hires through 2033. And the ultimate goal is to get to about 3/4 automation, about 75%. One of the major stories we talked about through the COVID era was how many people Amazon hired to build out its warehouses as E commerce took off during that period of economic disruption. Now we're seeing the other side of that and a lot of long term by the company come good. Great news for Amazon shareholders, all of us here who have 401ks. Pretty tough I think. Jason, looking ahead for the people who currently feed their families on $19 an hour Amazon warehouse jobs, what's your take, lad?
- [24:45]
C
Seems conservative to me, but what are your thoughts on automation? Robotics? Yeah. Purpose built, humanoid.
- [24:52]
A
Yeah, it's interesting. It feels like for a long time robotics is one of those areas that was always going to happen and didn't quite right. If you looked at the installed base of robots, it was in the hundreds of thousands of new robots per year, largely for automotive and very specific industrial applications. And it seems like now we're going through a sea change. And I think it's kind of funny actually, if you ask What's a robot? Your dishwasher to some extent is a robot. Right. Like a lot of things that we kind of take for granted now are robots in the modern era. The things that are really emerging now are obviously some of the things that were just mentioned relative to Amazon, but also even self driving, driving cars and Waymo and Tesla and what's happening there, Those are robots, right? Those are cars that drive themselves. There's a lot of startups right now, PI and the bot company and a few others who are all building next gen sort of foundation models for robots. And so it seems like within the next couple years we'll see a lot of these things substantiate. And there's sort of two questions that come to mind for me. One is what proportion of the winners will be incumbents versus startups? So for example, if you look at self driving, there was two dozen self driving companies, but the winners arguably are Waymo from Google and then Tesla. Two incumbents actually effectively have so far at least one self driving. There's cool companies like Applied Intuition and others who are providing pieces of this, but Caros and other things, but I think that sort of clear winners, at least so far, those two. And robots, it seems like Tesla will be one through their optimist work. And then the question is, who else is it a startup? Is it another incumbent? Is it incumbent who buy the startup? Is it at home? Robots, which a lot of people are working on, which I view would be kind of awkward. There's like a robot in your living room and you walk in and it's like mopping or something. I'm joking about that. Or is it an industrial application? More B2B which is sort of the Tesla Optimus sort of view. And so I think there's this big question of who wins and why. And then what are the first application areas on top of that?
- [26:44]
C
Yeah, it was pretty obvious to me. I saw Andy Jassy speak at a private event, I don't know, a couple of months ago, and he's pretty bullish on AI. And then a day later he did this public missive where he wrote, hey, AI is going to transform everything. And we could see different team sizes. That was like a, to me, when a CEO that sophisticated doesn't write that memo without a reason. He was telegraphing internally, externally, with partners, hey, you know, this is going to result in less jobs. And I was on the Bulwark podcast, a political podcast, and I said, you know, I don't see a world in, you know, the next 10 years where a human being touches a package at an Amazon factory or to your doorstep. Here's the quick clip.
- [27:37]
B
Jason, do you want to just run the clip?
- [27:39]
C
Yeah, it's a little self indulgent, but sure, let's. It's.
- [27:41]
B
This is literally your podcast. You're allowed to be self indulgent.
- [27:44]
C
We want to be too self indulgent here. Quoting.
- [27:46]
B
We'll stop you. Anyways, here's Jason quoting himself.
- [27:50]
C
This is where things could get really nutty and we're already starting. And that's where like the doomsday scenarios, Terminator ones kind of come out of that group. So what happens if, you know, all drivers go away? Well, every self driving car is four full time jobs and every humanoid robot in a factory is five jobs, maybe six. Because those jobs you can only, you can work for maybe seven hours as a just physical limitation as humans were driving, you can work a 12 hour shift. And those are, those are going to be here, folks. And before 2030 you're going to see Amazon, which has massively invested in this, replace all factory workers and all drivers. The idea that when you order something from Amazon, a human would touch it at any point in that supply chain is insane. It will be 100%, you know, robotic, which means all of those workers are going away. Every Amazon worker. All those jobs, you, ups, gone, FedEx, gone. Every, all of those are going to be gone. And those companies will be more profitable. And when you order something, it's going to come faster and cheaper and better and your Uber will be half as much. Somebody needs to retrain these people. And that's actually a very valid. What do you think about job displacement, Lad? Obviously I have strong feelings that this displacement is coming. I don't know if anybody disagrees with that. I guess we're all just going to debate timelines. 3 years, 5 years, 7 years, 10 years. Do you think we're going to have societal issues with job displacement? We're already seeing computer science graduates having a hard time recently getting jobs. And I think most people are attributing it to AI, making the senior people more productive. So what's your take on job displacement? I'm curious. Probably the best part of being a founder is watching a project that started as just an idea blossom into a full fledged startup. The worst part, it's bookkeeping. It's boring, it's detail oriented and it takes time away from building your product and hey, growing your team. It's time for you to let Pilot take care of all this bookkeeping. So you can get back to running your company. Pilot's the largest accounting firm, and it was built just for startups. They understand the technology industry and the landscape. They know the stakes, and they're already trusted by amazing companies we love like OpenAI scale, Airtable, and beyond. And you know what? Pilot is about so much more than just getting your taxes done on time. Pilot has a smart dashboard that puts all the essential metrics your startup is tracking at your fingertip. And they have helpful advisors who can walk you through crucial financial decisions you're going to need to make in the coming years. Plus, when it's time to raise your next round, Pilot COO and CFO services are there to help. So it's time to focus on your company and let Pilot handle the books. And Twist listeners get $1,200 off their first year. Just go to pilot.com twist to get started.
- [30:50]
A
Yeah, I mean, I think any technology wave definitely has job displacement that comes with it. And so then the question is, what do you do from a retraining perspective? How do you think about social safety nets? And then also, how well are you governed if the. You know, if we were governed remarkably well, I'd say, well, we'll figure it out and we'll. If we're governed really badly, then it's probably a different story in terms of the degree to which you end up with, you know, different forms of unrest or outcomes, depending on how much jobs are replaced at what speed, and then how much of a safety net or retraining or other things you provide. So I think it's a pretty complex thing, but I do think that it's pretty clear in certain areas that you're starting to see shifts in terms of labor, particularly around customer support, may be one of the earliest places that I think ends up getting impacted to some of the points made earlier.
- [31:41]
C
Yeah, I think your point. Go ahead, Alex.
- [31:44]
B
Oh, just. Salesforce said they're saving 100 million a year now on automated customer support. A lot. Just to back up your point, it's a place where people are finding enormous cost savings already, which, of course impacts jobs.
- [31:53]
A
Yeah. And if you have like a temp or outsourced labor team or you've partnered with somebody about it, as a company, you're more likely to be willing to cut that back than if it was your own employees. Right. And so often when you have these outsourced or delegated systems, you're more willing to shut them off. And so that could be some of the savings without actually changing official employee Headcount, but it changes the virtual set of people that you've hired. And I think that's why we're not seeing it in some cases as much. Right. It's kind of hidden because of the.
- [32:21]
C
Way that Such a good point. Outsourced it already. So you're already outsourced a contract, you're not doing a plan shutdown. So. So in the United States you have.
- [32:29]
B
That it's not visible warren notice.
- [32:32]
C
The Warren notice warn notice. You don't have to warn anybody. It's like, yeah, I just got rid of a vendor and I wonder if that's a pass through where you still have to do a warn notice. Putting it aside, I will say just had this realization as you were talking. I now, as a default, do not go to the manufacturer's website to get an answer. I go to an LLM. Why? The LLM has indexed all the Reddit threads, Quora threads, Twitter threads, message boards of people solving the problem on their own before the company even knows there's an issue. So I think most organizations probably don't even know what their true customer support.
- [33:21]
A
Requests that form. Yeah. And there's companies like Decagon obviously helping with that. And then I think there's two other implications of that or implications of all this. One is there are certain countries that should be impacted sooner rather than later from an employment perspective that we're not actually thinking about. Right. Because there are certain types of jobs outsourced to specific regional hubs.
- [33:40]
B
Et cetera.
- [33:41]
A
Sure. And then I think the second thing, and this is the point that Nikesh Arora, the CEO of Palo Alto Networks, has brought up with me in the past, which I thought was a really good insight. He said this quite publicly, that they're using AI tools and foundation models and everything else to iterate on product because customer support, in some cases not all, just means your product isn't working good enough or the UI is hidden or something else is going on. And so to some extent by iterating on product faster, you're going to cascade into support as well from that perspective. Right. Essentially, if you have bots fixing every problem for you, automatically your support load should go down.
- [34:18]
B
Right.
- [34:18]
A
And so there's all these really interesting sort of second or third order things that, you know, are really interesting to speculate on in terms of the future world.
- [34:27]
C
It's really interesting to see the number of people employed at these companies stay the same as their earnings go up. Exactly.
- [34:34]
A
And this is why.
- [34:35]
C
Yeah, never seen that in our lifetime. And this is if you were going to think about where to park money in the world, I was hearing somebody who's like, super liberal podcaster who hates America and Trump derangement syndrome, the whole thing. And he's like, I'm taking all of my money out of the United States.
- [34:53]
A
Not you, Alex.
- [34:55]
B
I was gonna say, dude, I'm sitting right here.
- [34:59]
A
Just got doxxed.
- [35:01]
C
Exactly. You don't have tds, you just have a love of democracy. It's slightly different.
- [35:05]
B
And another capitalist.
- [35:06]
C
Yeah. And he said, I'm just putting it. I think United States is on the decline. I got to put my money in Europe and every other. And I was like, anarcho capitalism would argue the opposite. If in fact people are going to deploy this technology aggressively in the most vibrant capital market on the planet, the United States. We could see Amazon, to my point earlier, like, what does Amazon look like with no delivery drivers? What does it look like without. With delivery times cut in half and running 24 hours a day? What does that even look like?
- [35:51]
A
Yeah, eventually it's just Jeff Bezos and a bunch of bots back to the.
- [35:56]
C
One person billion dollar startup. Maybe it's a 10 person trillion dollar startup at some point if it's that efficient. Here in Texas, they're delivering stuff from Amazon warehouses so fast that I'm like, I don't need it that fast. They literally like, would you like this from 4am to 8am or 8am to, you know, 10am and I'm like, it's 10pm, sir, I'm ordering a cable.
- [36:26]
A
It's 10pm tomorrow. Yeah, it's okay. At your leisure.
- [36:31]
C
And like literally the drones are coming. Like if you live on a ranch or something like that. I mean, the drones are going to be the ultimate. How crazy is it, Milad, that Larry and Sergey bought every robotics company for five years, what, 10, 15 years ago, and then they sold them all because they couldn't figure out an application for it? I mean, yeah, they were so far.
- [36:54]
A
Ahead on so many things. And so I think a lot of credit is due in terms of being really advanced on the curve. And there's the old saying that. I mean, there's two different things I think that apply here, but one of them is just that A, the features here is just not good distributed then B, often it's less a matter of who but when. And this is a good example of a when. Right. They were just too early. And I think they've been remarkably early on all sorts of things. They experimented with delivery, they experimented with Internet Using weather balloons. They experimented with satellites. They experimented because eventually I think they sold their satellite company to Planet Labs or somebody. Right. But they had satellites, they had everything. They were just too early on. So much kudos to them for trying.
- [37:43]
C
And it's like a venture portfolio. You only have to hit on one and it looks like Waymo will be that. One of the other bets the Loom project was too close to Earth. Turns out, low Earth orbit, better. You look at Google Glass that might actually be landing soon in the form of these Ray Bans that everybody can't shut up about from Meta. So they could have kept going with that. I also think their Chromebooks and the Chrome operating system is eventually where AI lands. If you think about what will be the starting point for every AI journey, where do you think that will occur? Do you think it's voice and wearables, phones, dumb terminals, a new operating system? Where do you think the end game is for mass consumption, mass interaction of AI?
- [38:34]
A
Yeah, I think it depends on how good the companies are who are currently providing all our hardware, because the default state is they should win. Now the question is, are they going to win? Which I think is a separate question. And one could argue that there's been very little on device innovation in AI. And then the question is, is that because they're waiting to see what works? Is it because they're investing and it takes time to build client applications? Is it a lack of imagination? And so I think there's this broader question of why has so little AI been built into these devices so far? And in some cases you look at something like Siri, it was never the best voice based system, et cetera. And so there's a broader question of why is that? And in some cases it's to protect users. So for example, Apple has a very strong data privacy policy, which means that they don't have a lot of data that they normally would have for training that Google has. And so it's been purposeful. But now in the AI era, should that still apply or not? Or how should you think about it as a company? I think those are very interesting, probably ongoing questions in terms of where does that AI future come from? The default should be the devices you use today or the devices that should be the launching point for AI. Whether that actually happens, I don't know. I mean, what do you think?
- [39:46]
C
Yeah, I've been thinking about Apple a lot. They are very good at making hardware. One of the greatest things they ever did was get off of other people's chipsets. And just make their own. Right? That's been kind of the revolution. And they can make devices that are higher margin, more powerful, Leapfrog everybody. Low battery consumption solve so many problems for them. These laptops they build are unnecessarily powerful and unnecessarily long battery lives because they've just optimized so well. So they actually have an advantage in that. And then there will be a time when people have their LLMs come up in, you know, lawsuits and discovery, and Sam Altman will be more than happy to turn them over. He said, like, don't put your stuff in here, folks. Like, we will, you know, have to hand them over. You know, I think it's good that he was honest about that. So encrypted on your device with an Apple phone, Apple, Mini, Apple, you know, M17, whatever they get to you could actually run the LLM locally. All the, you know, context window could be run locally on all your items, and it could be running in the background, indexing your photos without ever putting them in the cloud, like Google, and having your photos, you know, train Google's LLM, which I don't think you really want to do. And Zuckerberg will do that in a heartbeat, obviously. So I think they have a potential there. Yeah, yeah.
- [41:15]
A
It's a really interesting question, because if you look at search as an analog, and I don't think it's the correct analog, right? But if you look at it as an analog, basically, Google ended up paying Apple tens of billions of dollars a year to use its search instead of developing its own. And it was all margin, right? You didn't have to hire all the engineers, you didn't have to implement all this stuff. You literally got paid money to just distribute somebody else's search effectively. It seems like that's the path that's been taken so far on LLMs around foundation models. And then the question is, is that the right approach or wrong approach in this era? Because you're not just searching for things. It's not just a way to interface with the web. It's the way to interface with everything. You're shifting the set, the framework. But from a deal perspective, that deal worked so well that I wonder if that's part of the framework that they're thinking of. So I have no inside knowledge of this. I'm just saying if I had gone through that historical arc, maybe that's how I'd be thinking about things. And it may or may not apply to this different world of AI doing sort of all these Various tasks for you and agents suddenly spawning up, like in the agentic world. That's a very bad strategy potentially.
- [42:18]
C
I'm so bullish on Apple making a big AI purchase. I just think they should just take out whatever LLM doesn't break, you know, stay in the top three. They should just take out Perplexities, a great user interface. They really know how to build slick interfaces, the comment browser. And my lord, just since the antitrust case, think about how much traffic they could spill over to DuckDuckGo brave or perplexity or anybody in that kind of zone of excellence. It would be extraordinary for those companies.
- [42:53]
A
And yeah, yeah, I led Perplexity's first round.
- [42:57]
C
Wow.
- [42:57]
A
And Aravind is one of the most impressive founders I've worked with in a really long time. Like, just a complete. He's incredible at hiring, incredible at execution. Highly strategic, morphed into a very commercial person. I mean, when I met him, he was at OpenAI as a researcher. Right. And I actually met him because I think he reached out to me on LinkedIn and said, hey, I'm working at this lab at Berkeley. And I joined OpenAI like eight months ago. And this is when nobody was doing generative AI stuff. I think I funded him in maybe just 22 sometime. We started meeting up weekly and just brainstorming. And he obviously came up with perplexity and that's what he wanted to do and what he wanted to pursue and all this stuff. But I've just been impressed by how he's really just ramped so quickly on so many areas. So it's been a very impressive journey.
- [43:48]
B
To watch companies sniffing around Perplexity trying to snap it up a lot.
- [43:53]
A
I don't know, I'm not on the board there or anything, so.
- [43:56]
B
Okay.
- [43:57]
A
Look at Alex trying to get this breaking news.
- [43:59]
B
I like it.
- [43:59]
C
Alex.
- [43:59]
B
Well, no, I'm just. I mean, Jason's been talking about secrets. The different, the, the difference.
- [44:06]
A
The worst thing that happened in your childhood. Let's talk, you know, like, no one.
- [44:10]
B
Cares about my secrets a lot. That's, that's, that's the difference. My secrets are news. No, but I'm more just asking the general point, like how, how much interest is there from the incumbent technology companies towards AI startups? Because we've seen some acquisitions, but Jason keeps seeing these interesting possible combinations, and I don't see as many of those in the market as we would expect. So I'm just kind of curious about the temperature or the appetite for that kind of deal making big.
- [44:31]
C
Mm.
- [44:33]
A
Yeah. I think it's because, you know, the, the FTC has continued to be reasonably anti large tech acquisitions and that was started under the Biden administration with Lina Khan and others. And I think that's just been continued into this administration. So I think it has as much to do with kind of a purview of how the government may intervene. And that's not just the US government, it's also the eu, it's the uk, right? The Figma deal was scuppered by the uk. It wasn't the US that did it. And it almost feels like there's this regulator round robin where these different groups are taking turns in terms of who's going to scupper which thing. So it isn't the same party every time.
- [45:10]
C
They must be on a group chat. They're like, we're busy with, with stopping the totally inconsequential Figma Adobe deal. Why don't you guys find an inconsequential robotics purchased by Amazon for a $2 billion company to block?
- [45:24]
B
And it's like, how does the UK have that much power though, Jason? Like the UK is this little bankrupt block off the coast of Europe, which is a natural economic block.
- [45:32]
C
I said this publicly many times. They should just ban you from using Adobe and Figma products in the UK forever and just, just punish them for blocking it. They should just say like, okay, if you're going to block it, we're not going to sell to your citizens. They can go use some open source project. Hey, speaking of interfaces, we have this sesame round which I'd like to go to. But before we do, I think it'd be good to look at a poly market. Will Tesla beat their quarterly earnings?
- [46:00]
B
So I chose this, Jason, just to let everyone know, because we're getting into earnings season and you know, I'm a big dweeb for the public markets and I thought it's interesting that actually as we see sports betting and prediction markets come together, it's important to keep in mind that also people use these for what we used to use the stock market for, which was a surprise to me because why would you bother to use a different vehicle for that? But people seem really into it.
- [46:22]
C
Yeah, 80% chance they're going to beat their quarterly earnings. I think you could really do well on these prediction markets by looking at the historical data. Almost every company tries to sandbag these earnings, right Elad? They're always trying to underestimate them so they can beat them. So, so beat is almost always the default case. But this is an interesting. If you were a large shareholder, I guess that you could use these prediction markets in some way to hedge. You know, just like sometimes people collar stocks or something like that, or getting a little extra juice. How do you think about putting aside just this one specific one, which you're welcome to comment on. What do you think about these prediction markets becoming just so quickly integrated into our lives? I was using one to gamble on the Emmys and made a little bit of money watching the Emmys. And then I did. I was watching a boxing match. I won again. So obviously these things are amazing since I've won two times in a row. If I lose, then I have a different opinion. But for now, my confirmation bias is.
- [47:27]
B
These things are awesome.
- [47:28]
A
They're fantastic. Yeah, yeah. You know, it's interesting because this is one of those concepts that's existed for years and there was crypto projects to do it. And finally it's working between polymarket and Kalshi. And really, if I look at it, it feels like there's two or three big use cases, right? There's basically sports and sports betting. And they took what was a really complicated interface. They did three things, right? One is suddenly it's available in every state when it wasn't before. Because sports betting was regulated on a state per state basis. They simplified the UI dramatically, right? It used to be, what's the spread and the point thing and the this and all this complex stuff versus who wins. Yes, no. So they simplified the ui. It's like swiping left, right? And then there's kind of like, you know, that global liquidity pool where you're suddenly reaching a much larger mass of people. And so you have way more crowd wisdom and that sports is a huge area. Politics, when there's the political season, is obviously a big area. And one could argue the stock market is just a prediction market, right? It's the same thing. And so I always wonder, where does this go? Can you start. Start buying stocks or shares that are basically collateralized via crypto, basically tokenized shares through these markets. At some point, do these things converge? Are they completely separate? In which case a lot of the financial speculation will continue to be the actual markets and this becomes more sports and politics. So I've been kind of wondering, where does this whole world go and what sort of financial instruments emerge either on top of this or does it go the other way? The New York Stock Exchange announced that it was investing through ICE. I think it was $2 billion in polymarket. Do eventually the stock markets effectively become These prediction markets, because they are in some sense already.
- [49:12]
C
Yeah. I think if you want to own the underlying asset, you buy the stock. And if you want to speculate and get extra juice on more acute, specific time based moments options, that's option trading. It's like you're doing it in a simple way.
- [49:32]
A
Yeah. It simplifies options way down. Yeah. And that I think is the real innovation here is you've collapsed these really complex speculative things, sports betting options, et cetera, and you've turned them into a. Yes, no.
- [49:46]
C
Yeah. This is when you're in the professional gambling circles and you are like the bottom rung of it, as I am. And I watch people try to do side bets, what they call string bets. They make them very complicated. Which for the people who are the sharps who are constructing these between the poker sessions where they drop them out, it's like those option traders, they're coming up with some really interesting synthetic thing they want to do, like Michael Burry and the Big Short or whatever. And then there's somebody like Goldman trying to, to make it more complex to try to get an edge. And they're all just trying to get some little edge on it. And you're right. The simplification of it, you know, will OpenAI raise money at a billion dollar valuation in the next two years? Well, I can't own OpenAI shares. It's too complicated. Or people want 10% vig on it. You know what, I'll just bet on that. Get the same. You know, it's like, okay, it's a 60% chance. Great. I can make, you know, whatever percent more on my money. I'll take it.
- [50:47]
A
Yeah. It's a transition from blogs to tweets, but from a financial speculation perspective.
- [50:52]
C
Perfect analogy. Perfect analogy.
- [50:54]
B
Shorter form, fewer calories. You do more of them and they.
- [50:57]
A
Have more people do it. Easier to participate, Everybody can do it, et cetera. Yeah.
- [51:02]
C
Okay, let's talk about the sesame raising $250 million. Because I don't know what Sesame is.
- [51:08]
B
Yes, I thought you might not. So this is a company that came out of stealth back in February. It's done a couple things. One, it's built Voice first AI and its first two models that it put out earlier this year, Jason, were popular enough that they racked up 5 million minutes. But what's interesting about this company and why I wanted to bring it up on the show today is they're also working on hardware. They're building glasses for you. And I have an image here from their. Oh, let's call it their lab or their office. It doesn't really show us much. But the idea here is back to the point about Google Glass being early, is that Sesame wants to give you some hardware that lets you chat with your personal AI agent or assistant or friend or whatever you want to call it, it on the go. And they're taking the hardware approach to it. And I think that explains the amount of capital here. Just as background, how did they manage that? Well, they're founded by Brendan Uribe, the co founder of Oculus and also the former CTO of Ubiquiti 6, which was an AR startup. So they have a lot of that AR DNA in their veins.
- [52:06]
C
Interesting. So they worked at Ubiquiti, so they understand architecture. $250 million. And exactly how is the device going to work? Is there a demo of this device out? Have they made a video of it?
- [52:21]
B
Very, very little. The Virgs got to try it and talked about it a little bit and that's where that image came. And I do have another one from the same post, but it's pretty, pretty nascent. In fact, this company is only putting out the beta of their iOS app now they're actually taking applications for the beta of their iOS app. So it's very much under wraps, if you will. But the idea is, and Sequoia wrote about this in a piece explaining their investment, Jason, is that it's not just another AI voice chat like your ChatGPT instance, for example. It's supposed to be more of like a collaborator with you and with a focus on how it handles speech to make it seem natural between you and maybe your AI. One thing I'm sure you've noticed is that if you chat with GPT5 and you make any sound while it's talking, it just stops to give you space, which makes it almost impossible to converse. So I'm hoping that what they're doing here is making more natural communication with AIs, more personalized AI and then making it functional on the go without having to opt in to the Zuckerberg, I would say spyware empire.
- [53:21]
C
What are your thoughts on wearables, lad? And this idea of it's persistently monitoring what you're doing in the world and what could only be described as the most creepiest Truman show concept ever, but the benefits.
- [53:36]
A
So you're saying you like it?
- [53:39]
C
I love the Truman show in a dystopian way.
- [53:43]
A
Like, that's cool. Yeah, I think. Have you guys ever heard this book called Accelerando? It's like a book, it's like a, it's, it's, it's a book with wonderful concepts. I think the book itself is a little bit roughly written, but the basic idea is it's the, you know, for, it's, it's like the decades leading up into the singularity where you have this technology discontinuity, discontinuation. And you know, eventually at the point where every 10 years you accomplish more than all of human progress. And then eventually every year you accomplish more than everything ever before that, et cetera. And things keep accelerating and in it they have this notion of glasses, which are effectively a form of external memory. They're an agent, they act on your behalf. This book was written maybe 20, 25 years ago.
- [54:24]
C
Yeah, 2000.
- [54:25]
A
Very sort of forward looking.
- [54:26]
B
Yeah.
- [54:26]
A
And one of the really cool questions that it raises, if so much of your memory and all these agents and all these things you do is actually loaded in your device and you lose your device and somebody else puts it on, how much of their personality or how much of your memories or how much of your abilities is it absorbing as you sort of augment as a human and you externalize stuff? And so I always thought that was a really interesting, almost like conceptual book. Like again, there's lots of really interesting ideas in that book and this was one of them. And it's one of the earliest versions of glasses and human augmentation and all that kind of stuff. So it's this interesting old concept that seems to again, people are iterating against.
- [55:06]
C
William Gibson had a similar concept in his Virtual Light Eduru, the Bridge series of like San Francisco dystopian future. And he had this thing called God's Little Machine or something. It was like a little zeppelin that flew above you in like a three quarter shot, like the Sims recorded your entire life, indexed it and you could pull it up and say, oh, when did Elad and I first meet? Or what was it like when I, you know, the first date I went on with my wife and you could just kind of replay everything. And then you had have of course, Strange Days, the Catherine Bigelow film where you can put on a neural headset and record your experience, then somebody else could put it on and live stream your experience and or relive it. So these ideas are there and it really goes to show how much of our consciousness and processing in the world is done visually. That whole system could be the unlock for intelligence and consciousness in some way is what some of the, you know, biology. Yeah, I'm Sorry, Strange days. And it was called the SQUID superconducting quantum interface device. Thank you, producer Claude for coming to the rescue here. I do have some concerns about these things or being always on and recording everybody all the time. I don't know if I'm going to win that battle though. You have concerns about this?
- [56:34]
A
I'm in general a very paranoid person, so I always have concerns. But I think that it just comes back to utility. I think one thing that's been proven to us is people fluctuate between willing to sacrifice privacy in order to get access to more stuff. That was Google Gmail. There was all these concerns around Google reading your email, reading it, not really, but a machine reading it to generate ads. And that was. Was an issue. But then hundreds of millions of people adopted Gmail and then Facebook similarly, people had concerns about their own personal data and privacy. And I think for a while Facebook actually was doing fine on it. And probably what happened is because I sold my first company to Twitter and that was around 2009. And so this was right as Facebook started paying attention to Twitter. And I always wonder if there's some alternate timeline or history or something else where instead of trying to go after Twitter and media and public content, Facebook leaned into its strengths because at the time it was a very trusted privacy brand, Right. It was your identity and it was your friends and your friend group and it was viewed very differently. And what if they'd gone into banking and payments? What if they'd gone into healthcare? What if they'd gone into all the things that were privacy native, almost what Apple ended up doing. That's a very interesting alternative timeline for Facebook. And maybe they will be the trusted identity. Yeah, but it's an interesting concept, right, of like they decided to go after Twitter. What if they decided to go after banking at that moment in time when they were very trusted as like a, you know, they had all your stuff. But that was fine, right? Because it was just your friends seeing it.
- [58:00]
C
It was your. It was your social graph. It was closed. They weren't trying to serve you ads in. Zuckerberg, in fact, hated advertising. That was like. He thought it was terrible. He didn't want to have ads on the service until at some point somebody told him, well, you can make it free and get billions of people to use it. And he was like, okay, fine. And then Sheryl Sandberg comes and the rest is.
- [58:22]
A
And this could have been an alternate timeline. Maybe you would monetize through payments. And this has happened a few times, right. I've talked to some of the people who worked on the early browsers and it was the same thing. They're like, is it going to be micropayments or ads? And they ended up going down the ads route. But there was an alternative universe where it was like micropayments. Right. And you'd pay a.
- [58:39]
B
Maybe not kind of. I mean the technology wasn't there. As the founder of a short lived micropayment startup back in 2000, whatever, you know, I mean credit cards just weren't. Weren't there. I think the idea was good, but I think we needed essentially.
- [58:51]
C
Eli, you had one, you had a micropayments startup. Oh, you did, you did a startup?
- [58:57]
B
Yeah, it's called Contenture. We founded it, TechCrunch wrote about it, we launched, got customers, and then we failed and I had to go back to college.
- [59:04]
C
Oh, okay. Wow. You were just too early. I kind of feel like. Yeah, they were doing the brave attention token for a while. That didn't take.
- [59:12]
A
That's right.
- [59:12]
B
Right.
- [59:13]
A
Yeah.
- [59:13]
C
B A T. There's been a couple shots at this and you know, one of the things I've learned is when we just talked about it with, you know, Sergey being so ahead of the curve that he was too ahead of the curve. And then probably some accounting wonks were like, we're seven years into this. Sergey like let's pull the plug. And he was probably like, all right, fine. When if he had just stayed till year 11, it would have been like, oh, okay, now we're own low earth orbit satellites. We hatched our link.
- [59:41]
A
Which was Waymo, right? Yeah, which was Waymo. But yeah, they could have had a couple others. So again, very forward thinking company. I think they don't get enough credit for that.
- [59:49]
C
I think this is where like Sergey and Larry just need to say we will be willing to lose money on projects for 20 years, not just seven. And we don't care what you think. Remember they were doing Google Fiber and I think that one got deprecated as well. That would have been fascinating if they kept going along and pulled that string. Because then you'd have Fiber plus satellites.
- [60:13]
A
You'D have fiber, we'd have flying cars. Basically everything would be done, we'd be out of business, we couldn't invest anymore.
- [60:20]
C
Well, exactly.
- [60:22]
A
Be like Google's world.
- [60:22]
C
But also that was like Larry had to go off balance sheet and use his personal balance sheet to do. He did three of them, I think.
- [60:30]
A
Kitty Hawk. Right, Kitty Hawk.
- [60:31]
C
Was he also in Archer? I think he might have also Invested in Archer?
- [60:34]
A
Maybe. Yeah, I'm not sure.
- [60:35]
C
He was in at least two of them. Them just fascinating how close being too early is. But I'm always encouraging founders to look in that startup graveyard. I think actually micropayments, as silly as it sounds, built into a social network, a browser for content. This might be like substack or a YouTube competitor might be able to do very well with this. Where you could set your. What if you could set your browser to just say if I watch more than one minute of a YouTube video, send the person one penny. And if I watch, you know, a penny for every second I watch of anything and if I, you know, if I stay on a new a document and I scroll down to the end. Yeah, give them a penny. There's a lot of people who might buy into that if there was some reward for it. I don't know what that reward is.
- [61:27]
B
Well, I mean that was kind of what we were trying to build, Jason. So it was back then you couldn't do one penny, so you would put in $5 to your contenture account and then as you went around the Internet we would hand out the money to people that you visited that had our integration and people kind of liked it. But it was hard to get any sort of momentum.
- [61:42]
C
Think about stable coins. Like I wonder if stablecoins are the unlock here. What do you think about stablecoin?
- [61:48]
A
Fractional? Yeah, yeah, I've been long interested in them. I mean it's interesting to see some of these second order effects of those now. Right. In terms of, of the biggest holders of Treasuries. In some cases, amongst the biggest holders are stablecoin companies. So it's been fascinating to see both their growth. I remember when Tether came out, initially everybody thought it was super sketchy or very uncertain in terms of what the assets are actually holding against the money. And now obviously it seems like they're fully collateralized between a mix of gold, bitcoin and Treasuries.
- [62:17]
C
They've done a great job cleaning it up.
- [62:18]
A
Yeah, yeah. So it seems like it's going to be a really central thing in general. And I think, I think one of the things that people have underappreciated about crypto, particularly people in the US is the degree to which globally, for example, people don't have access to dollars because you're not allowed to buy a government. And this is a way to buy dollars and this is a way to do simple transactions, like all the stuff that we take for granted here. But it also shows the path forward with tokenizing shares and token things we've been talking about for years, it feels like are finally going to happen. So it's a very exciting era I think.
- [62:52]
C
Here we go. Tether is according to producer Claude Tether, 127 billion in treasuries as of Q2 2025. Number 18 global ranking head of South Korea and circle 18 billion is in the top 30.
- [63:09]
B
Yeah, yeah.
- [63:10]
A
It's very under discussed how many treasuries are actually held by crypto companies now. And so one of the biggest buyers of US government debt is the crypto industry. Like think about that. What are the implications of that in the long run? If crypto gets 10 times bigger and that 120 billion goes to 1.2 trillion, what does that mean in terms of the US government government's relationship to crypto? How should we think about that? Yeah, super interesting.
- [63:35]
C
Who's your daddy? It's kind of like the China relationship.
- [63:37]
A
It's tether, it's tether. Tether.
- [63:39]
C
Tether's your daddy.
- [63:40]
B
Paolo is your daddy. I guess in that case.
- [63:42]
C
Well and the interesting thing about Tether is because yeah, they did have that really spicy, I'll use the word spicy. I don't wanna get sued here. But they were banned in a lot of regions. A lot of like regional action taken against them, a lot of questions about those reserves and they have gone like on a five year process of cleaning all that up. And there's like, I think they have according to the bill, the stablecoin bill that Sachs was involved in or talked about on the pod a bunch of times, I think they have a three year window to clean everything up up in order to participate in the United States. So Circle's got the clear lane along with I guess their partnership with Coinbase. But they have to Tether is going to clean up and the, the guy, Paulo, is his name Paulo?
- [64:27]
B
Yes.
- [64:28]
C
Who was like living in some non, you know, the speculation was like none of the executives were available and they were living in jurisdictions where you couldn't be deported or picked up by a red notice. I don't know how much is that were true. There were a lot of tethered truthers out there, you know, speculating. But them figuring out how to go legit if these stories of illegitimacy were in any way true is like a story for the ages. That's going to be the next Sorkin movie I think.
- [64:55]
A
Yeah, it's pretty good. I think one of the most under hyped players in crypto right now perhaps is Stripe who I think have been doing amazing things in terms of stablecoins and Bridges and all the rest of it. And you know I first invested in stripe in 2010 I think. So it's been kind of a long road but what an amazing next act. I mean obviously the payment stuff is going to keep going and the SaaS stuff will keep going but I think what they're doing in crypto is super interesting right now.
- [65:23]
C
They're never going to go public. No, they definitely have. They're never going to go public. This is my new prediction.
- [65:29]
A
Private poly market. I just infinite duration, perpetual, perpetual options. Hyper liquid.
- [65:39]
C
Well here's the thing. If you have that fluid of a market, if you're an expert in transactions in commerce, what's to say they couldn't build and you know the infrastructure to let people who are qualified purchasers trade in and out of their private company and then slowly buy back all their shares based on massive profitability and then have a black box bucks company that nobody understands what's going on there. It's a tremendous advantage. And the Carlson brothers are unique in their perspective of the world. I don't think they're particularly driven by some outrageous personal net worth on a public market and the traditional accoutrement of being an entrepreneur. They seem very mission driven and have very specific principled things they want to work on, you know like their think tanks or science think tank, et cetera. So that could be amazing. I mean I'm trying to think of the equivalent of what's the largest private entity.
- [66:42]
A
Yeah, there's a couple of them. There's like Coke industries, there's Cargill, there's a bunch of these. Yeah, it's kind of interesting. Bloomberg, right?
- [66:49]
C
Yes.
- [66:50]
A
Tons of cash and it's private so. So I think people under appreciate especially you know, Europe. Like a lot of the market cap is actually these middle stock private Companies in Germany e.g.
- [67:05]
C
Vtol. Swiss based Dutch multinational energy and commodity trading company. Yeah. Revenue $400 billion in 2023.
- [67:18]
A
Isn't Coke like a hundred.
- [67:18]
C
Coke is number five, 125 billion.
- [67:22]
A
Cargill should be on there. Like there's a bunch of these.
- [67:24]
C
Cargill number 4, 160. I'm trying to find a tech one. Huawei number 10. I didn't know that they were private.
- [67:32]
A
Yeah, I mean I guess Bytedance is a private Bytedance 14 that company. Yeah.
- [67:37]
B
Yeah. Hey, a lot on the, on the stripe stablecoin points. I agree very much that they're they're leading. They bought was it Bridge for 1.1 billion and they just announced Tempo their their own kind of blockchain or L1 for. For stablecoin payments. I'm not asking for internal information here but like what do you think they're trying to do with that? Is that just to make a better foundation, better rails for stables to fit into the kind of existing E commerce world or are they trying to solve a more I don't know fundamental issue with stablecoins today that is not well served by existing L ones and L2s?
- [68:09]
A
Yeah, I think it's best to just look at what the company says about all this stuff because I don't want to over interpret or misinterpret what they're trying to do. I think fundamentally if I look at what they've been building as a company over the years, it's really this focus on providing a piece of global infrastructure like Patrick used to talk about it almost like container ships and that enable global commerce for the world. And so I think they're very mission driven around this interconnected web of commerce is not good for the world. And how do we help make it really easy for people and in particular developers to build things against that.
- [68:42]
C
That so absolutely extraordinary. What else do we have on the docket here? We got maybe one more story left in us. What do we got?
- [68:50]
B
Well, because we have a lot here, why don't we talk a little bit about the, the, the fracas Jason with Anthropic and your bud David Sachs.
- [68:57]
C
Ooh, that's spicy.
- [68:59]
B
Yeah, I.
- [69:00]
C
We also have the OpenAI browser less spicy. Let's go for the spice.
- [69:04]
B
Okay, let me just pull that up real quick. I gotta find.
- [69:09]
C
I have some thoughts on it.
- [69:10]
B
I'm shocked. I'm shocked that you do I not but I have opinions so I know.
- [69:16]
C
For better or worse.
- [69:19]
B
All right, so Anthropics. Dario Amadei wrote a public post this week responding to his critics. But to understand what's going on we have to go back in time a little bit to the one big beautiful bill. Jason, we discuss here on the show that there was a provision that eventually didn't make it in but that would have banned states from having state level AI regulations for a period of 10 years. Ended up getting voted down 90, 99 to 1 in the Senate. Fast forward to recently Jack Clark, another co founder of Open Sorry of Anthropic wrote a post, a summary of a talk that he gave in which he said that he thinks other stakeholders in the World should have a bit of a say in how we build AI as we build it. He's not a doomer, he's definitely an optimist, but he thinks that we should listen to other people. And then your friend David Sacks comes in and he says that, quote, anthropic is running a sophisticated regulatory capture strategy based on fear mongering. It is principally responsible for the state regulatory frenzy that is damaging startups. Fast forward through some back and forth and other people weighing in and we got Dario's open letter essentially saying, hey, what the hell's wrong with you guys? We love startups, we work with them. We're in favor of a federal standard. And this has become such a big deal, such a back and forth that CNBC is covering it. And so Jason, I'm incredibly curious why we have left the era of founder friendliness and we're now seeing venture capital capitalists attack some of the most successful companies in the history of startups.
- [70:39]
C
You know, this industry left its boutique phase when Facebook hit scale. Facebook just had far, far too much impact on society. Elections, our children, Instagram. I think that was the end of the innocence. Just if I were to pick, you know, a demarcation point, a lot of when, when we became so big that we couldn't be looked at as the weirdos in California building cool stuff. And then it became, oh, how are these things damaging society? We should be questioning everything about AI we just did today, talking about Amazon and job displacement. We should be thinking about kids using these products. We should be thinking about people getting one shotted and getting, you know, and we should be thinking about the truth and the search for truth. I think all these things are valid discussions. I think what's at stake here is there is covert, I don't know, lobbyists and covert political actions to try to regulate things. And though that does make me nervous, I just want to see all the evidence of it, if that's actually happening or not. And I, I told Sachs this, he wasn't too pleased with me a lot. But if this is the most important technology of our lifetime, why wouldn't states have a right in the format of the United States of America to, you know, chime in on it, you know, if they kids access to it or how it's used in healthcare, how it's used in terms of job displacement? What's your take on all this?
- [72:15]
A
Yeah, I think at this point in time, and I think industries evolve over time and the nature of how you regulate those industries should evolve with it. But early on in industries, it's a bit of a disaster to let each incremental state chime in on how to regulate it, because fundamentally there's a handful of states that effectively by fiat govern what the rest of the country does, whether the rest of the country adopts things or not. It's almost the flip, why should California be able to govern AI for everyone in the US which effectively almost happened through the Scott Wiener bill. So I actually view it on the other side of this one where I think the states should not intervene for now. It should be federally mandated. There are both national security reasons to do that. There are national competitive dynamics that drive that. So my view would be. And this literally almost happened, right. California was going to try and regulate how the rest of the country deals with AI, which is a really bad idea.
- [73:11]
C
So. And it was based on the amount of compute you have, as if that is some sort of.
- [73:18]
A
Well, the question is, why should California regulate it for at all A, and then B, for everyone else, which is the fact that, I mean, it would have been national, which means it would have been global. So should California have the right to set national or global AI policy effectively?
- [73:30]
C
Yeah, and the car example is a good one. And so that's one where I happen to be in favor of it. So when you look at these, it's always.
- [73:39]
A
It may have been the right thing to do. The question is, should California have been able to do it? In other words, correct.
- [73:44]
C
Yes, I was in favor of it.
- [73:45]
A
Yeah, yeah, yeah. It's like, let's separate the. Do we agree with the thing.
- [73:49]
C
Yes.
- [73:50]
A
From how should that thing get done? Right. I think those are two separate things. So.
- [73:53]
C
Yeah, and that's what I've been trying to parse in my head because I, you know, yeah, we should have higher gas standards and who cares if cars cost a little more? You know, having smog over cities and people losing, you know, months to years off their life and asthma and all this kind of stuff. Yeah. It's obviously a tragedy of the commons. It's a good thing that California did that in my mind. But should Texas be forced to not buy the trucks they want that have the hauling power and the price performance ratio that they want? I picked Texas as a place to live because I like less regulation. And so it really does come down to the operating system of America where the states are supposed to have some sovereignty, but then you have to super oppose it against. Well, if we are in fact in a global race with China, can we, can we afford to slow ourselves Down. And that's like a very interesting. This is probably there's a global race.
- [74:48]
A
With China and then there's the, there's the political side, there's the military side of that, there's the economic side of that, there's the political and cultural side of that. And so I, I disagree as well with, hey, we should have every country on the planet we in on how AI models should get built in the US I think each country should have its own sovereign AI model and there's lots of ways to accomplish that. If I was running a country, I'd be very interested in that. I want my cultural values represented in local models and things like that. So I agree with that. But the flip side of it is, if anything, perhaps the last few decades have shown us that perhaps we don't need to solicit everybody's opinion on everything, both in the context of a company, in the context of a country. And so this is almost a throwback to things that haven't really been working. It feels.
- [75:36]
C
Yeah. If you were to put this against high speed trains, Alex, like there's a reason why we don't have them in America because everybody gets to chime in on the high speed train. And then the high speed train, you know, is like, you know, but in other countries they're like, high speed train zoo has to be a straight line. Yeah, but it can't go fast.
- [75:56]
B
And you know, but, but in China there's been a lot of studies discussing how both province and central planning has led to a dramatic overbuild of economically inefficient high speed rail. And so we can all give a thumbs up to the Shanghai Beijing line. Why not? But when you see the, what is it, eight, eight pillars and eight cross beams. There's some model that I'm translating poorly here for how to build high speed rail in China. And it's led to a lot of misplaced spend. And so, so I agree that we.
- [76:23]
A
Should be the drivers of that spend. Like that's often also corruption. That's a bunch of stuff. It's not just.
- [76:28]
B
But if you want centralized power a lot, if you don't want people to have saved, you're going to end up with a central power.
- [76:35]
A
Well, really what the question is is what are the things that should be regulated by what level of government? And that is something that has been established and adjudicated in the US for centuries now. Right. And this falls into line of those things, which is, is how much should states be able to govern on their own overall technology progress for the US and for the globe. And that's effectively what this issue really is, right?
- [77:01]
B
Interstate commerce?
- [77:03]
A
Not just interstate commerce. It's effective de facto regulation of the entire country.
- [77:08]
B
Right. No, I'm saying that because it's. That it should fall prey to that and therefore it could be limited in that way.
- [77:13]
A
Exactly, exactly.
- [77:14]
B
But what's interesting though is we're talking almost about the wrong foe here. Cause I think this is why anthropic topic is, you know, being dragged into a bit of the mud here. But the provision to block state level AI regs was voted down nearly unanimously in the Senate and partially because Republicans from the south wanted to protect local industries. So how do we get Congress to listen to this perspective and do something intelligent and pass something that will preclude or maybe just do better than state level regulations? Because Dario said in his letter that they disagreed with the idea because it blocked action without offering a federal alternative. So how do we get to one? How do we fix Congress?
- [77:55]
A
I mean, there are questions. There is, I'd say three things on it. One is I tend to stay away from politics. I'm the least cogent person to ask about how to fix Congress or how to change. That's not the world that I tend to spend time in. I do think there's a number of different approaches that we've taken in the past or historically to federally mandate things. And one route is executive orders, where they can be applied. The other route has been acts of Congress. There's three or four routes that are just standard. And the question is what best applies here? And then I think it's back. There's also the meta issue of where do states cross the boundaries in terms of the areas that probably shouldn't apply to them? And is that a legal framework? Is that something else that already exists? And how do we interpret it? You know, the existing legal structures of the country relative to some of these topics.
- [78:49]
C
I think that's the granular piece that you're talking about a lot, I think is the key piece of this. Which downside? What concern do we have over AI that is not addressed by an existing law? If you were to go in an AI and you ask it right now, how do I build a chemical weapon, how do I build a biological weapon? And says, hey, I'm not going to do this, that it's already been solved by the teams, they've self regulated. So, okay, what's the next piece? Oh, how do we deal with, you know, I don't know, the Bioweapon stuff.
- [79:21]
A
The bioweapon stuff is so overstated. And I feel like that's the way that some of these companies have been trying to cause distress and fear and regulators and others. Because the reality is all the protocols for, or building a virus you can release, et cetera, or on the web, I mean, I worked as a biologist for almost a decade, right. And all these web protocols are out there. The real issue is buying the equipment, using it, actually doing things in an intelligent manner, ordering all the oligos you'd need, et cetera. So the extra information content, in my opinion, isn't high. And the hard part is the physical world stuff. And so whenever people raise that as an example, I view it as kind of a red herring where, where it's purposefully raised in an attempt to cause fear in people so they'll react and regulate something. So I do think those are bad, bad examples that people tend to use. It's purposeful.
- [80:18]
C
I think it's per. I agree. And I always ask people, tell me the next one, tell me the next one. Yeah, let's just get granular here for a minute. I have one like IP protection. I think that that should be done. That should be self regulated by the industry, which they're starting to do. They're starting to say, hey, even if you don't agree with how Sam's doing it at OpenAI, he's saying with Sora, hey, if you're like, you know, the estate of Elvis or Martin Luther King will be gracious and turn off the ability to make Martin Luther King saying inappropriate things. And they did. They didn't have to go the legal route. Sam said, we're good people. We're going to be aggressive in letting people do things, but we're also going to be reasonable in turning them off. I disagree with the approach. I think they should always get permission. But okay, fine, he took a different route. He'll get the legal exposure. But yeah, this is a situation we have to get into each one. And I was just thinking when you were talking about bombs and I was like, remember fertilizer bombs? Everybody was like, oh my God, the Internet's causing this. The Tor network, you know, Dark Web, everybody knows. It's like, you still got to get the fertilizer. Every piece of every bag of fertilizer has IDs in it, like in the chips. From what I understand, there's like some secret way that they watermark every bag of fertilizer. So if you blow something up and then if you Buy them. You need to register. And they've got some, you know, I mean, who knows how the government is implementing this after the 9 11. Because the first before 911 they tried to blow up the World Trade center with the fertilizer bomb. They succeeded in blowing up Oklahoma, Soma City. Two different groups of terrorists, domestic and international, they solved this problem. I bet you they have like ways of tracking fertilizer where like the second the person leaves the fertilizer store, there's like a camera on their license plate. They're being tracked by a drone. Like the government took this seriously. Seriously. There's so many simple solutions to this.
- [82:17]
A
Yeah. That's why I stopped fertilizing my garden. I was worried that I get tracked down.
- [82:21]
C
Yeah.
- [82:21]
B
So you laugh, but the only time I've actually gotten pulled out at the airport and put into a small room was after going home to Oregon to my parents house. And I was walking around the backyard a lot and I popped some sort of test at TSA for possible fertilizer.
- [82:36]
C
That happened too.
- [82:37]
B
And they like, I've never been treated as a criminal quite like that. They're like, is there anything in here that's explosive? I'm like, I don't know, weed, but please don't find that. And it was so. Oh, sorry, we're live. Edit that out, I guess. But it was just, it was so surreal. But it, it can catch you.
- [82:53]
C
And they, they literally asked me like, have you been doing gardening? I was like, no, how's that relevant? I didn't know what they were talking about.
- [83:01]
A
They're like, I've been helping startups grow.
- [83:03]
C
I don't get it.
- [83:04]
B
Gardening the stripes by spreading manure.
- [83:06]
A
Yeah, exactly.
- [83:08]
C
All right, listen, listen. This has been another amazing episode of this week in Startups Aloud. Thank you for joining us today. What a great guest. Where can people find you? I mean, you got a Twitter account. You're pretty active.
- [83:21]
A
Yeah, yeah, Twitter, they can find Melodgo and pleasure to be on. Thanks for including me.
- [83:26]
C
All right, and we'll see you all next time on this week in Startups. Bye. Bye.