The Twenty Minute VCWall St's $725BN AI Question | The Rise of Open Source & How it Threatens OpenAI & Anthropic
EVERY SPOKEN WORD
90 min read · 18,256 words- 0:00 – 1:09
Intro
- JLJason Lemkin
Open Source is a bit of a fake because China's paying for all the training, okay? It's not Open Source like a generation ago. And the reason is, is because there's so much innovation and cost savings.
- HSHarry Stebbings
So what are we to discuss today? Google loses two generational scientists in 48 hours. That's a tough time. DeepSeek closes $7.4B Series A. It's a Series A? At a $50B price, but only China gets voting rights. Interesting. And then finally, the $725BN question Wall Street is finally asking, "Who's actually gonna pay for AI?"
- RORory O’Driscoll
There's only one thing worse than a seat-based model, and that's a model that's based on bodies.
- JLJason Lemkin
You don't get to make $10 million for working 18 hours a week. You get a watch. You get an Omega. You want an Omega or you want to be rich? Make your choice, boys.
- RORory O’Driscoll
The whole reason the OpenAI and Anthropic models work is because other idiots have spent the $300 billion on their behalf.
- JLJason Lemkin
Ready to go?
- HSHarry Stebbings
[upbeat music] Boys, it is so good to be back. Jason, you are back. It is so good to have you back from China.
- 1:09 – 13:28
Google Loses Two Generational Scientists in 48 Hours to Anthropic
- HSHarry Stebbings
Um, we're gonna start with the news that I put at the top of the list, which was DeepMind loses two generational scientists in 48 hours. Uh, namely, we have first Naom Shazeer, who was character AI, and then we have John Jumper, Nobel Prize winner, uh, co-creator of AlphaFold, also leaving to join Anthropic. How significant are these moves? What should we read from this?
- JLJason Lemkin
Look, listen, it's easy to pick a turnover in any organization, right? There's so, there's so much turnover in any organization. Uh, on the other hand, the ... I, I'll... You know, when you talk to, to some of the smartest engineers and developers, uh, in AI, they're really looking to be in a very specific environment, right? Where they get to pursue exactly what they wanna do, especially on the research side, right? Especially they don't, they wanna work on what they wanna work on, and the best of the best. And you know, I was thinking back in the day, I went to pitch Google, uh, for my last, uh, s- startup, and Vint Cerf was there, like one of the creators of the internet, popped into the meeting, okay? I didn't get it at the time, but Google back in the day, right, pre-AI days, had created this environment where the best researchers in the world wanted to be there. And I think that's how they lured the DeepMind guys in, right, was the story, right? This persistent to create this environment. "Look, you're gonna get to stay in London. You guys are gonna get to build your own thing." And I think this is probably just a sign of the, of the cracks of the realities of having to try to be number one in AI and the, and forcing f- forcing an environmental change potentially that your competitors, um, can welcome, right? Anthropic and OpenAI can say like, "Just come over here and work on whatever you want to work on for $500 million, $2 billion." And it's... I just... When I talk to s- folks at the bleeding edge of AI, that's just so appealing. It's just so appealing to only work on what they want to work on on AI. Um, but, but again, I just got back from China, I might be off by a beat. But you, if you want to get the best researchers, you have to give them this environment where they only get to do what they want to do.
- RORory O’Driscoll
It, it, it's funny 'cause that, that feels like a w- a one-dimensional answer, but there might be two dimensions to this. 'Cause in one sense, yeah, you know, you have that whole, you know, researchers just want to go do what they want. On the other hand, you listen to a lot of people who left Google, and it's a little bit of the frustration of not being able to ship. I mean, there's a lot of frustration that they had a ChatGPT alternative, and then the bureaucracy just kind of smothered the product when, you know, OpenAI just jammed it out the door, and as a result, took a lead on them, right? So I think that it, it's funny when you're, when you have a histo- y- you know, an existing business, you're kind of damned if you do, damned if you don't. Sometimes people want to just do research. Other times they want to actually get shit done and ship, and you're getting in the way of that. My sense is, first of all, these two people, in terms of their research pursuits, are somewhat different in that, you know, Naom obviously was at Google, did the original attention paper, left, did character, got bought back to Google in large part to be, you know, it was a very clever acquisition, the brains behind their restarting their, you know, after OpenAI stole a march on them, and I think it was like a couple of billion dollars. So a good slug of that was going for him. So assuming four-year vesting, he's probably left half of whatever he's offered on the table, right? So it's-
- JLJason Lemkin
That's a lot. [laughs]
- RORory O’Driscoll
I can imagine there it's a combination of some version of, "Hey," um, as you say, more research, more ability to do things, plus, you know, a more certain ability to ship and make shit happen. 'Cause I think even though we all did the last year, we did the, "Yay, Google is amazing," because unlike the other three, um, unlike the other hyperscalers, they have their shit together, they have an AI story, and the stock has reflected that in the last year. It's the only one of those -- It's one of only two of the Mag Seven that's up on the last 18 months. But at the same time, when you look at things like having a viable coding model, having really kind of that next level up from just going from just shipping a model to shipping, you know, interesting products, the truth is Google hasn't done an amazing job, and OpenAI and Anthropic have. So if you're in the, into product shipping, which I think Naom might well be, then I can see why going to one of those two makes sense. Totally different, you know, on the jumper... I mean, this is someone who's your pure research science around, uh, um, you know, undergraduate, postgradually, I think University of Chicago, all focused on protein folding. Remember, got a Nobel Prize. Um, DeepMind has been the only company he's ever worked at since his... I don't know if the graduation is the right word after you get your postdoc, but whatever. Since academia, right? So basically, it's been a whole bunch of academia, a whole bunch of time at DeepMind, pick up a Nobel Prize. You gotta believe you're going to Anthropic because there's some story there about being able to do more research, which is almost, you know, on kind of high-end science, which again, they've had an, um, Anthropic's announced initiative on that. So kind of somewhat different. But I think what it speaks to stepping back is, you know, when you're top of the heap, you can promise everyone, e- e- everything to everyone. In a way that you're not c-- And the truth is, top of the heap right now means you're the new model, you're the new companies. You're not unconstrained by history, you're unconstrained by the install base, like that old, you know, proverbial joke about, you know, hell is the install base. And you got a stock and a currency that is huge, and no one's giving you shit about stock-based comp. So if you're Anthropic and OpenAI, you know, you can buy whatever you want, including people, and you can let them do whatever they want, including whatever it is they've been promised to do, in a way that you've much less constraints than the incumbents.
- JLJason Lemkin
You know me, I would never deal in rumors. I don't do rumors. But the rumor mill that I heard was-
- RORory O’Driscoll
Harry, you're a rumor slut.
- JLJason Lemkin
[laughs]
- RORory O’Driscoll
Come on. Come on, bring it on, baby.
- JLJason Lemkin
Is that candidly, Anthropic have clearly had a breakthrough that a small number of people know about, and that's why John Jumper went there.
- RORory O’Driscoll
The reason I'm skeptical of that, Harry, is the lag time between I've had amazing invention and revenue in the kind of LL- core LLM space is a year or two years or three years. The lag time if you have a medical invention, an idea, and turning that into real meaningful economic value is ten or fifteen years. I mean, you know, y- the protein fold... I mean, the amazing thing is the protein folding adva- advance has collected its Nobel Prize and as yet has not had meaningful commercial success or a drug in production yet. So I doubt it's like, "Oh my God, we've discovered something new, and this is magic, and if you don't join in the next three months, you won't be part of this thing." My guess is the initiative, whatever initiative Anthropic has around next-generation science is, you know, a m- a multi-year thing, and it's just a question of where do you want to spend the next five years of your life doing research.
- JLJason Lemkin
It is a vibe, for what it's worth. And listen, I, I really hate to be one of those, like, VC dad tell stories about their kids or what happens in their s- their spoiled kids' school, uh, uh, you know, as, as your, as your thesis for investing. But my son is very good at a certain type of math. I don't understand Jensen, fine string theory and all this stuff. But as a college student, he's one of those, like, top ten in the country. So the-- So my point of the story is the labs find him, right? He doesn't have to apply s- per se. So he, he got an internship offer without looking for it from one of the, from, from, from Anthropic for, for amount of money that when I was in college or grad school would, you know... A- and, and the opportunity is incalculable when you graduate, even today with... And so he instantly turned it down, right? With no job. Um, and he, but he's just, he's just like, "I can't do enough research. I want to do my type of research for my type of math and my type of a- AI." I mean, and so, uh, you know, everything that we do on this show or the other stuff, he kind of makes fun of me because he's so far ahead of how inference works, how open source work. I mean, really you should replace me with him. But, but my point of the personal, the dad VC stories, which I hate, is he instantly just said... Like, they couldn't create... He's not worried about money. He's not worried about any of this. If they can't create the right learning environment for him, he just won't do the job, right? And so I don't mean to, to, to, to map my, my, my family to two of the top researchers i, in the universe, but, but I fe- I feel the same vibe. Like, in today's world, we're on a bull run like we've never seen in our entire lifetimes. There's nothing like this where a, a, a, a so-called startup can pay billions of dollars to acqui-hire you, and then you leave with fifty percent of your billions unvested to where we've never been like this. And it just creates an environment for the best of the best, where you just will only do what you want to do. You just won't do the job that isn't a hundred percent what you want to do. And remember when Harry and I were in London and we were with Maggie for OpenAI, sh- I think she said she was one of the sales leadership. She said she'd never been in the researcher building except for, like, one or two meetings. They didn't allow sales in the whole building with the top [laughs] researchers to leave, to leave. We don't want these pesky go-to-market professionals bothering our, uh, researchers, so. And I hate to be that guy, but it really resonated. Like, you just, you have to, like... And it's really hard for-- If you're not Anthropic, how do you retain... Now, y- it's not DeepSeek, but how do you retain this talent? How do you let people work on what they want to work on when you need the chat bug fixed so you can compete with Sierra? I mean, how do you, how do you contain them? And I, I think it's very hard.
- RORory O’Driscoll
First of all, you're exactly right. I mean, these are two of the most talented people on the planet. One of them w- was one part of the attention paper and has made plus or minus a billion dollars, and the other has a Nobel Prize for medicine and is still relatively young. I mean, these are not top one, they're not top point one, they're top point zero zero zero one percent people, right? Probably one of, you know, out there on the planet in terms of research. And the funny thing for Google that, the so hard thing is they're struggling to keep people like that, while at the same time they're not getting the tactical shit done, you know, and, you know, and, and just making the stuff happen to kind of continue to make progress. And the amazing thing about someone like Anthropic is they're able to do both. They're able to create these wonderful research environments for the leave me alone, let me do research people, while at the same time they're obviously executing tactically, brilliantly in terms of product, making stuff happen, right? That's-- It's momentum. I mean, that's my aha here. It's like everything in life. When you start winning and everything starts going your way, the people start going your way, the breaks start going your way. You know, winners win and they compound until something happens to break that chain, right? And this is just it and magnified.
- JLJason Lemkin
Uh, listen, Google overall I think actually has been on fire. But I think, um, a- as I continue to learn, and I'm only so smart about the role of open source versus closed source, right? The most vulnerable is going to be number three. And so you really, for talent, for people, for revenue, for, and for your ability to do things that maybe aren't core. If you're number three as the closed source LLM, that's where you're going to hit the most pressure from open source, and it's just... It, it-- Sometimes when you're in that, and I'm not saying this is correct, but sometimes when you are in that place It, it, it you feel like you don't have the luxury of, of letting folks do what they wanna do because you're under such intense pressure. Maybe, maybe Anthropic feels like as competitive it is, it has a luxury that its competitors don't have.
- RORory O’Driscoll
I think that's true. And just, I wanna come back on the Google is executing amazing. I was not comment here. No, but just look, objectively, as over the last eighteen months, since twenty-twenty five, only two of the mag seven have outperformed the S&P. It's Nvidia and Google, which is another way of saying, you know, people like Facebook, Amazon, and Microsoft are doing shit. They're not relevant here. Google has done an amazing job of being relevant here, m- and that's true, statement one. But statement two that's equally true is you don't wake up every morning and say, "Let's try the new Google model. Let's try the new Google harness. Let's try the new Google coding tool." You do try Claude code, you do try co-work, you do try OpenAI. So the fact is, they are relevant and in the frame, but they're not yet, as you say, but they are definitely number three in terms of innovation, which is a whole lot better than being Microsoft or Meta and having to say, "We don't have something yet, but we spent seventy billion. We might get something next year." Right? But yeah-
- HSHarry Stebbings
Jason, just, just so I understand, why is number three the worst position and, like, the powerless one in a way that is not for cloud?
- 13:28 – 18:00
Why Being #3 in AI Is the Most Dangerous Position
- JLJason Lemkin
There are two threads happening at the same time. On the one hand, clearly, the price of tokens, token maxing, uh, the budget issues are real, and so the amount of folks rut- routing models, running multi-model is exploding, right? And so at a, at a, at a small level, the open routers and, uh, and your own sources, that stuff's all a big deal, and everyone is, is realizing they just have to get smarter and smarter much more quickly on routing workflows to different models. Okay, that, that is clearly very true today, and in maybe ninety days on the pod, it wasn't clear how big a deal that was. It is across all, everything except the smallest of startups are routing, okay? So what do you route to? Well, you know, if, if, if you have three vendors, and Rory's the professor, especially with his background, you know, nu- number two was often simpler and number three was usually cheaper, right? Some variant of that, right? So i- you know, and, and, and f- frankly, GOO- GCP, old Google Cloud blew up because it was cheapest. It was cheapest. You know, a generation ago, Google, Google Cloud struggled in the beginning, and then it was just cheapest and simple, and so people would move certain workloads a generation ago to Google Cloud. Now you're trying to do the same thing with your massive, massive AI spend, and the question is, open source is really complicated. Open source in, in inference and training is not free, unlike Linux. It's not free. There are substantial costs, but it is materially cheaper, which we could talk about. So should your number three-- forget about the fact in theory with Open Router and others, you could route to ten thousand models. In practice, is the number three thing you're gonna figure out what's the best open source model I can use for my application, or is it the number three closed source? Open sour-- there's so much going on in open source, right? It's fueling these crazy base ten and fireworks and all of the-- There's so much innovation in open source that number three just might get swamped by all the subsidies of the Chinese government, everything else subsidizing open source, because open source is a bit of a fake because China's paying for all the training, okay? It's not open source like a generation ago. And the reason is, is because there's so much innovation and cost savings. So your number three cl-- there just, we just, there just might not be enough energy for the number three closed source model, even if Google has the billions to fund it. They do have all the money it takes to fund it, but developers may lose interest other than it being, uh, what, a, a setting in, in Open Router and friends. That's my thesis, and that's what I'm seeing too, right? Um, and, uh, it's a big deal, and literally this morning, I, you know, I got an email, um, from Anthropic. This is their, this is their sort of a cro-- a shot across the bow for open source saying, "Your prompt cache hit rate is low. Your prompt cache hit rate is out of the blue." I don't know if you guys got this email this morning, but, you know, what Anthropic is doing here aggressively is fighting back at open sources and trying to get you to cache your prompts, which are very expensive, and they offer such a massive discount on cached prompts if they work for you, that it actually can be cheaper than open source. This didn't say, um, you know, fighting Gemini. They're-- This is an email they sent to their, maybe their entire base saying, uh, "Cache your prompts so that it's cheaper than, uh, open source."
- RORory O’Driscoll
Got it.
- JLJason Lemkin
So that's why I think it's just hard. It, number three, you just, you know, you c- you can't be cheaper, and you can't, and you can't keep the researchers, and the projects are less interesting.
- RORory O’Driscoll
No, but, but, but, but, but yeah, maybe I was, 'cause you're-- But I'm, but the short answer is I agree, you're right, Jason, is, is that, look, typically tech markets tend to... I mean, look, you don't end up with perfectly competitive tech markets in the economic sense of millions of players. You end up with small numbers in a tight oligopoly where there's a leader, a number two, and then depending on the size of the market and competitive dynamic, maybe there's a three and a four, and the vast majority of the revenue goes to one and two, right? It's just, it's the structure in the cloud market where AWS was first, Microsoft second, and Google, Google Cloud third. The interesting about Google being third, right, well, there's a couple of things. One is, unlike the typical third, if they were a standalone company and didn't have the Google balance sheet behind them, I think it would be incredibly tough, right? And I think implicit in that statement is it's almost impossible for a number four to em-- yeah, a, a same business model, closed source number four to emerge and catch up at this point. I mean, Elad Gil had a nice piece on that six, twelve months ago. The market is set. The game is clear, right? And, and the only reason Google can keep on punching is because they have a whole balance sheet behind them. 'Cause I agree with you, Jason. The other part of it is, if there is a compelling alternative
- 18:00 – 29:00
DeepSeek's $7.4B Series A at a $50B Valuation: Is China Winning?
- RORory O’Driscoll
to this entire set of c- competitors that's, you know, five X cheaper, which is what open source is, then it's gonna grind everyone down. And when an industry gets ground, what tends to happen is the number one guy makes a little less money, the number two guy makes quite a lot less money, and the number three guy goes bust. Right? And in this case, obviously not Bosch because they got Google behind them. But you're right, it's a real powerful downward pressure on the profitability across all these, and that, and that is, I think, the big story now kind of from open source with lots of caveats.
- HSHarry Stebbings
Rory, how do you think about sovereignty then and like sovereign models and Mistral of the world if there's no room for number four?
- RORory O’Driscoll
There's two ways to answer that. Well, I... We'll answer sovereignty first and then the impact of China. Well, on sovereignty, look, Europe is effectively saying, if someone says we need a sovereign model, re-re-rephrasing that, what you're saying is, "We no longer are part of the market over there in America where our product is third. We are, in fact, first. Instead of being fourth in the worldwide market for closed source, state-of-the-art foundation models, we are first in the European market for state-of-the-art foundation models." Right? So you're effectively saying, "I couldn't compete as the fourth player worldwide, but I can compete, profitable for me, but probably less efficient for the system as a whole as the dominant player in, you know, Europe." And look, a million years of economic theory explains why that's a dumb idea from an efficiency perspective. By definition, now everyone in Europe is getting the less good model at a higher price. But someone's deciding that there are political or national security reasons to pay that tax, and that's a, you know, political decision that's beyond the economic analysis. It's just so-something that government may choose to do or not. Right? But the other part of it, and I'm gonna segue to Jason's trip there, is the other part I thought the sovereignty question is, the fascinating thing is all the competitive models to their foundation models, all the open source models, bar not quite ev- not quite all, but most of, are Chinese-based. And, you know, Jason, you're just back from China. You know, in the context of sovereignty and security, it is amazing that the entire open source initiative is running on, you know, four or five models, all of which are built in China. So what was your takeaway from there?
- JLJason Lemkin
Well, it was, you know, it was interesting. Um, I mean, f-folks probably already know that, you know, when you go to China, in, in interestingly even Hong Kong, where they're-- where Hong Kong has its own sovereignty, it can do whatever it wants, Anthropic and, and OpenAI, uh, don't serve there, like intentionally for security. Like, they don't allow you to access Anthropic or OpenAI. You can't access it effective- it's har- You can do it over a VPN, it's, uh, but it's harder. You can tack it on your phone, but it's tough. They intentionally try to block whatever IP address blocking or whatever. It's not China, it's them blocking it, right? Um, but, but Gemini doesn't. So when I'm in China for two weeks, it's DeepSeek and Gemini. That- those are, those are, those are my friends. Um, it's a parallel universe. Um, and DeepSeek is intentionally crippled in China. It's not as good as it is here. It can't search the web, and it is trained on different data, as near as I can tell. Um, but, but you, you get it. You-- I didn't-- I don't think I got the sovereignty argument until I was in China, right? This is China basically saying, "We, we do not want to be subservient to the United States." Whether this was the original goal of DeepSeek or, and, and others and, and, uh, um, and Alibaba and others, I'm not sure. I'm not a h- I don't have my professorial background on my YouTube, but it clearly is where it is today, right? Massive government subsidies, right? DeepSeek's raising this round at eighty billion. The government's the only one getting voting shares. It says all you need to know, right? Um, and it is, it is a s- an existential sovereignty thing. The Chinese does not want to be reliant on Anthropic and OpenAI to run the next generation economy. It's very smart, and whatever investment the government has quietly made to subsidize all of these providers, it's a drop in the bucket, isn't it, at the, at the, at the sovereign level? What it w- ten billion, twenty billion, even fifty billion. I mean, this is nothing compa- what does an aircraft carrier cost? A lot. [chuckles] It's easier just to subsidize a mo- an open source model and pretend the training costs are ten million. What did DeepSeek claim their training model was back when it launched? [chuckles] Fifteen million to, to train an, an OpenAI competi- like, of course it wasn't true, right?
- HSHarry Stebbings
I, I... The round itself is, is pretty wild in a couple of different ways in that it-- $7.4 billion at a $50 billion, give or take, price. The thing that's crazy is the founder is committing 20 billion yuan himself, which is like 40% of the round. I mean, it's like $3 billion the founder is committing himself, wait for it, and there's fewer than 10 investors, including JD.com. I, I saw this and I was like, "What?" And then none of them are getting any rights. The only people getting rights is the Chinese state, which con- retains governance control. It's wild.
- JLJason Lemkin
Well, it's just, it's a different, I mean, it's a different world, right? These are not, these are not companies, uh, that, uh, you know, it goes back to the old hallway drama. They're not, they're not independent of the government in any way, shape, or form, right? And so you either do it the right way or you end up thrown out of your own company or, uh, you know, having to, to, to recycle and give back your $2 billion that Meta took from you. I mean, you gotta play by their rules, right? And so, you know, this is y- this is the, this, this is the rule for DeepSeek to not be taken from you.
- RORory O’Driscoll
You know, a couple of comments here. First of all, y- y- y- your, your comment on, "Oh, only the Chinese government's getting voting rights," I felt like saying, "So the only people getting voting rights are the only people who don't need them." Right? Because Chinese government doesn't need voting rights, right? Because they have sovereignty, right? They have an army, right? As we've seen by Manus, they can make you disgorge back your money, right, after you've got it, um, just by leaning heavily on you. So it's not like... I mean, s-so in one sense, the r- the round's not surprising at all and, and, you know, so zooming out two l-levels, two other things on that. One is, yeah, it's a $50 billion, but we have two companies, the leading two American companies, closed source models, which by definition yield more, you know, have more economic upside, are trading plus or minus a trillion. So $50 billion for the open source competitor feels roughly right. Right? You know, one-twentieth of the price. And, you know, there is a couple of these... I think, um, Zeppu, Zeppu which-- I should check. Uh, z-z.ai is, which is the Americanized version of it, is actually public now in China, and it's trading, it's trading at a hundred billion or something like that, a thousand times revenues. So again, plus or minus, if you think of these as the sovereign alternative and the open source alternative, they're trading at numbers that aren't crazy at all compared to the US. And then the other thing is just to say, you know, like, we're doing a little bit of the, oh, you know, you ha-- and you do, you have a lot of state interference in China, and, you know, you're seeing it over and over again in your property rights a week. We saw it with Alibaba back in the day with Jack Ma. We're seeing it here and now. But, you know, at some point, we'll have to talk about the fact that you're seeing more of that in the US as well, right? We're seeing, you know, Anthropic unable to ship their most recent model until they satisfy the concerns of the US government. And, and I gotta give, you know, Le-Leo and his 2024 situational awareness credit here. It's like he called it, that about now, national governments are getting involved, and we can argue if that a little bit overwrought. But in both... Yeah, we can't give China grief for this kind of thing when we're doing the same thing ourselves. Both governments are feeling this technology is pretty existential, pretty impactful, and they're trying to figure out, do you regulate it? Do you take over it? Do you stop it from being used by other people? There-there's a whole bunch of, you know, policy questions going well beyond economics that are being raised here, and, you know, China's solving it their way, and we're solving it ours, and Europe is doing theirs.
- HSHarry Stebbings
Uh, Rory, you mentioned Zeppu, or Zip-- I hope I pronounced it right. Zeppu AI's GLM 5.2 beats GPT 5.5 on coding benchmarks. How significant is this? This was deemed one of the most consequential open source AI releases.
- RORory O’Driscoll
I mean, look, to me, all it says is these guys are cranking. I mean, you know, there's probably six total open source Chinese models, three of them, you know, top at or close to US performance, three more just behind it. The, the aha for me is it's compa-- to Jason's point, it's a compelling competitive alternative. You know, many US companies are starting to build their own model based on the open source alternative, so it's providing some kind of... It's providing a, a ceiling on profitability for some of the US closed source vendors. It's really economically significant, and I think the most compelling fact is that there are six of these companies pounding it out there, and there's just a lot of competition on the open source, you know, state-of-the-art alternative. I mean, we can talk later about distillation and how much of what they do is a function of, you know, uh, their ability to learn from what the frontier models are doing, but they're there, they're relevant, and they're providing a competitive, um, well, that's not the right word, a competitive drag on, you know, what Anthropic and OpenAI can charge.
- HSHarry Stebbings
Behind all of these, there is actually the infrastructure, and a core part of that infrastructure, um, is seeing increases in price, memory being one of the biggest. Uh, price of memory has gone up four to five X in, in certain cases. Tim Cook told the Wall Street Journal that Apple faces a hundred-year flood in memory costs driven by AI infrastructure demand. DRAM contract prices rose ninety to ninety-five percent in Q1 alone. How significant are these cost increases? Who feels them? How will we see this? What should we take from this?
- RORory O’Driscoll
I think other than a profound regret of not buying Sandisk and Mic-- and, and Micron technology a year ago and making it 20X. I mean, look, I think it's funny. I was thinking about this and going back to the kind of discussion on the researchers at the start. Very different topics, but actually they're all about the same thing. The investment in AI is commanding resources, right? And then via the price mechanism, everyone el- everyone else is getting impacted by that. I mean, you know, and the impact of this, it's gonna manifest itself, as you say. I mean, I can think four of... DRAM pricing, right? It's gonna manifest itself in the price of your iPhone. It's gonna manifest itself in the price of your electricity. It's gonna manifest itself in the price of your house in San Francisco, right? And it's gonna manifest itself in terms of twenty percent of you losing your jobs if you're one of the companies that wants to kind of go all in on CapEx for AI like Oracle, right? So what you're seeing is the positive side.
- 29:00 – 49:09
Wall St’s $725B AI Question: Who's Actually Going to Pay for AI?
- RORory O’Driscoll
And again, this is not a, "Oh, AI is bad," as much as economics just sends its signal. It doesn't have morality. It just says, "Oh, you want to devote more resources and more DRAM to a spa-- a data center in Tennessee or Mississippi." That means you're gonna have to have less DRAM for Rory's iPhone, and the only way to make that happen is to raise the price of the iPhone, right? And that's just what's gonna happen. I mean, it's this price filter. And the amazing thing about the AI CapEx explosion is it's so huge that it's literally impacting everything, right? And this is just one of the examples. It's just sucking in the money. It's sucking in the resources. And specifically, what it's gonna mean is Apple has clearly decided wisely not to swallow the loss and lower their margins. They're just gonna raise prices. And they'll sell a few less iPhones because of pri-- you know, there's some price elasticity, and everyone will just have to pay more for their iPhone.
- HSHarry Stebbings
Goldman Sachs projected there'd be seven point six trillion in cumulative AI CapEx from 2026 to 2031. But it was fascinating, and it just brought up now the $725BN question. I remember David Cohen from Sequoia, the $50B I think it started.
- RORory O’Driscoll
Oh, $600B. I remember it well. It was a good piece.
- HSHarry Stebbings
Yeah. Are we gonna have a trillion-dollar question next year that we're gonna be discussing?
- RORory O’Driscoll
I mean, explain the f- what you're basically saying is... 'Cause we're, it's worth getting outside the inside baseball part of it. What you're saying is, what Goldman Sachs is saying is the hyperscalers are now spending $700B In CapEx a year, which times five or six years is three, five trillion dollars, some astronomical sum. And you only do that if you expect revenue to be greater than expenses, because that's how American capitalism is meant to work. So that implies that at some point, someone has to spend seven hundred billion dollars in revenue for you to make a buck, right? And right now, they were nowhere near that. We're probably well under a hundred billion dollars. So AI as a whole is m- getting a hundred billion in revenue and spending seven hundred billion a year. That's not a great business. And what you're basically saying is, how does that, how does that end?
- HSHarry Stebbings
And with the increase in CapEx, there's not-- Does that not just increase the revenue requirements to compensate?
- RORory O’Driscoll
Yeah, absolutely. No, people are going the other way. It's like-- L- And, and that's the funny thing to, to say. I would've thought that David Kahn piece a year ago was really good, and it made me think, and I was wrestling with the same things. And then you step back, and you say, "Since then, all that's happened is people have doubled their conviction and doubled their willingness to spend." At that point, it was, "Oh, my God, CapEx is, like, sixty percent of free cash flow for the Mag Seven." Now it's a hundred and twenty percent of free cash flow for the Mag Seven, and they're borrowing to fund it, right? So even though a year ago... And this is the classic thing about bull markets is, is that, you know, y- you can be intellectually right, but the narrative can keep going a long time, right? People's willingness to invest. I, you know, I-- It, it's very hard to know how that ends, right? Or when-- Not how it ends, when it ends.
- JLJason Lemkin
I, I-- Listen, I'm only, uh, uh... Maybe Rory can, again, share some of history where there's an analog. I just think in my, w- in, in my lifetime in tech, being aware of things, I can't think of another time where i- i- i-- you know, this level of demand was infinite. And so how we deal with it, how we deploy capital, but if you're sitting on the other side of infinite demand, you can choose not to embrace it, right? Um, that's like shorting everything, but where does it get you, right? I mean, you've gotta l- W- W- The demand for AI is an order of magnitude more than it is today if it could be served cost-effectively, right? We would all be consuming tokens twenty-four/seven if we could. So the demand is so much higher than, than, than, than is delivered today. And do you, do you play the game, or are you a Debbie Downer, right? I don't know what the right answer is. But certainly in venture, you don't make money not deploying your fund. Like, it's hard to m- I mean, fees are nice and all, but, like, you gotta deploy [chuckles] into, into demand cycles and just hope to God you get liquidity before it crashes down on you. [chuckles]
- RORory O’Driscoll
Diving one level deeper, 'cause you, 'cause you're right, Jason, you can't just say, "Ooh, I think, you know, you're, this is irrational," and go home, right? First of all, that would clearly have been the wrong decision for the last two years, and B, it's not the business we're all in. We're in business investing the future. But you do have to think about it. And the comment you said was interesting, which is that you, you've never seen this level of alm- infinite demand, and then you caveat it in a very important way, pri-- W- 'cause you said price adjust it. And I think that's the difference between this and prior g- you know, kind of prior generations of software technology because, you know, with something like Salesforce, if you were, if you're buying a SaaS product, there was really only two states. If you were, if you didn't need it, you didn't buy it, right? And if you needed it, and you were a big enough organization, you paid the hundred grand, and you said, "It's just the cost of doing business," right? 'Cause it was kind of a fixed price. What's interesting now about, you know, tokens and intelligence is, you know, you, there's an infinite demand for intelligence if it's free, right? But it's not free, and now everyone's just wrestling with some version of, "So how do we allocate that given that it's not free?" Like, no one said, "I'm gonna buy a hundred seats of Salesforce if it's, um, a, a thousand dollars a year a seat, but if it was, you know, only five hundred dollars a seat, I'd buy two thousand seats," right? I didn't buy m- I bought seats for my people if I needed them, and then I didn't, right? Whereas intelligence is this thing, as you say, where the real question is not, you know, you g- is not, "Do you want it?" The real question is, how do you ha-- You now have to have this new skill that didn't have to exist in kind of the prior world if you're a CIO. How do you decide how much to spend and where the cutoff is, right? 'Cause I think, and I think, I've always felt that the slowdown here doesn't come from, um, you know, some kind of technical barrier around all the models c- you know, 'cause the people who articulate the models can do everything are correct. The models are gonna keep on doing everything. They're gonna get better, blah, blah, blah, right? The real question is price. Price is gonna be the arbiter of how much you can do, and I, I, I, I think you're right on that. To make it in practical terms, even the last month and a half, there was this suddenly this, you know, recognition that maybe token maxing wasn't a good idea, and now you're seeing people try to be more efficient. The question is, do you think that shows up in actual slower revenue growth for Anthropic and OpenAI? 'Cause that's the rubber-hits-the-road question. If they're sending you an email, Jason, trying to effectively cut your bill by saying, "If you're more efficient, you won't be spending as much with us," do you think that reduces the growth from 10X to 5X? Or do you... How does all that shape out in terms of revenue growth? Do you have an opinion?
- JLJason Lemkin
Yeah, I don't know. I mean, I don't know if... You know, it's interesting. I got that email. You guys got that email. But also, I'm on the, I'm on the Max plan, where for two hundred dollars, I get ten thousand dollars of inference a month if I can use it properly.
- RORory O’Driscoll
They hate you.
- JLJason Lemkin
Right? So, so for me, they may have more of an incentive to send me that email than someone on an Enterprise plan, where for ten thousand dollars, I spend ten thousand dollars a month, right?
- RORory O’Driscoll
Oh, right.
- JLJason Lemkin
Anthropic and OpenAI are running an AB test where they have a pro su- they have a free segment of their base, which, which they are subsidizing, right? For sure. It's just bigger at OpenAI. They have an enterprise base which is, has, you know, between forty and seventy percent in, you know, gross margin just on inference. And then there's this prosumer one, the Max, the Max guys, right? Um, that some of us they're making a profit on, but some of them is massively subsidized, right? The prosumer, the kid v- uh, vibe coding ten thousand dollars of tokens a month, paying a hundred dollars, two hundred dollars for a Max plan is massively subsidized, and that, that's not a joke, right? That's the whole OpenCloud drama. Uh, it's a real issue. And so we have, we have this weird spectrum of free, massively subsidized, and actually, if you just look at inference and not training, fairly, fairly profitable enterprise customers, right? Fairly profitable. Well, I don't know what the term they use is, right? It's not, it's not... It's, it's an unblended gross margin, but the inference margins are attractive, right? And that- And that's what, uh, that's what open source is attacking, that enterprise customer, the, the lucrative customer
- RORory O’Driscoll
Yes, 'cause it's, it's lucratively profitable for the vendor, right?
- JLJason Lemkin
Yeah.
- RORory O’Driscoll
And the question is, is it profitable for the customer? In other words, ha- ha- have people said, assuming you're not on some kind of cap plan, and I think those are increasingly harder to get for the enterprise, assuming you're on the API, can you really afford to give everyone free untapped intelligence all the time, and does the math of that work? And as I say, the question is, what's happen... I mean, I'd love to know what's happening in real time now on usage and revenue across those companies, 'cause that's, that, that's the be all and end all question.
- JLJason Lemkin
Yeah, I think the story, you know, it's interesting, the token maxing isn't really the story, I think, to, to, to what Rory said. I think the big story of 2027 in AI and the enterprise, and all the margins in the enterprise, right, is, um, w- show me the ROI next year. Right now it's token maxing, because 2025 into early 2026 is, "Guys, just go do it. I don't want to be behind," right? We can make fun of token maxing, but it was the best way to get teams a- AI fluent. Just go build it, guys. Here's, here's 100 million, 50 million, 10, 5 million of to- whatever. Guys, just go build it. We've even seen it in our portfolio companies. Then the reaction was, "Guys, it's actually not a joke. You spent too much." Right? You blew through the, uh, the, the, the, the IT budget is bounded it turns out, right? Then, uh, the, the reve- uh, you're already starting to see it, but I think going into 2027, CIOs and others are gonna be like, "Okay, just show me the F and ROI. We're not gonna, we're not just gonna ration tokens based on who we like the most in the company and, and who makes the best PowerPoint pitch. Show me, show me the ROI. And if you have, if you have ROI, if you are able to lay off 20% of your department, or you have the highest growing s- division in our company, we will give you more tokens. And this group that, that can't ship or can't get anything done or is in decline, we're just not gonna give you the tokens." And so for the first time, ROI is really gonna have to connect to a lot of the spend.
- RORory O’Driscoll
Uh, I, I, I agree with you, Rory. I mean, it's... Sometimes I think saying the obvious is really helpful. I think you're exactly right, Jason. If you're running a big company and you said, "Okay, I wrote off the first half of 2026. We spent way more than we thought, but at least my people now know how this shit works." Write it off as a one-time thing. You're exact- And the question now is ROI. And then bringing it back to, you know, the $750 billion question, the Goldman CapEx question. I mean, I, I did this math a while back on the fly on the podcast. I'm kind of revisiting again. Let's round up to a trillion, 'cause if you're spending 750 on CapEx, you gotta pay for the electricity too. So you need a trillion dollars of revenue, right? And, and if companies are gonna give you a trillion dollars, they gotta be getting more value than that from that spend, right? So a trillion and a half plus or minus, you know, right? The total spend on labor in the US across everything is, you know, uh, GDP's 30 bill, 30 trillion, you know, it's sub 20 trillion. So you're really talking about 7, 8% of the labor force being replaced by tokens for the math to work. It's... I mean, when you look at that, you go, it's a dauntingly high bar, right? There's a little part of me that says, "I think this is amazing, but I don't know if that last dollar in CapEx is gonna earn a return." Right? And if it does, it's going to be because of a huge amount of labor force disruption and product... Let's call it, labor force disruption is the negative spend, the positive spend is, of course, productivity improvements. If the trillion dollars is gonna earn a buck, you're gonna have to have huge productivity improvements from AI, and huge productivity improvements, you know, result in labor displacement and hope- and eventually new jobs for that labor that's displaced. But it all has gotta... If one's gonna happen, the other's gotta happen.
- JLJason Lemkin
The problem with productivity calculations is there's also a parity tax, right? Like, you know, if you're the only one using a tool, you can achieve certain types of productivity, even if it works. Even... A lot of software productivity was BS. But let's assume it even works, but then all my competitors deploy Salesforce and Anthropic, and all of a sudden it doesn't show up because we have parity. So you can't, so you can't go back, but we may reach a situation in 2027 where we just cannot prove any of these productivity gains. And so going to your math, everyone may just need to lay off 10% of their company in addition to all other layoffs, right? Maybe the Oracle layoffs and others are just early views of where it's going, and Robinhood, and everyone's just gonna have to say, "Listen, we have to fund this. We have no choice," right? This isn't really... We can talk about ROI and productivity, but if we don't do it, we're gonna lose to our competition. So our team has to be 15% leaner next year, guys. It's just this simple. It has to be 15% leaner.
- RORory O’Driscoll
I, I, I'm gonna... And now I'm gonna do the professor thing just for a sec. That is productivity. Y- you, you're right. You can have an improvement in productivity and a, a massive improvement in productivity in industry by virtue of the advent of an enabling technology like AI, and no improvement in profitability for that industry if everyone adopts the technology. You're right. So if every ba- if, if every bank adopted ATMs at roughly the same time, everyone could, in theory, save on tellers. In fact, they used- ended up taking more tellers 'cause the business expanded, but we'll skip that for a second. But the problem didn't change, and the same could happen here. Everyone adopts AI, and everyone's cost function reduces by the same amount, and ev- You're right, but the people who don't adopt it are dead. That's your point, right? Which is why the people... Yeah, which is why you have to lean in, and I think that's probably likely in a bunch of these industries. If, you know, if management goes segueing to Accenture, if consulting or any of these companies, any of these white-collar jobs are 20% more efficient with AI, then everyone should adopt it, and then everyone will be... Then in the cost of those white-collar jobs, like audit, should reduce by 20%, all other things being equal.
- HSHarry Stebbings
My question is, like, do we see the same acceleration in model progression capability that we saw in coding in legal, in accounting? 'Cause if we do, where Andrej Karpathy says, "I moved from 20% to 80% in a six-month period," if you're able to see that same advancement in legal, whoa.
- JLJason Lemkin
I think, why wouldn't... Listen, obviously coding is uniquely well-suited, but we just built an AI VP of finance while we were in China-
- HSHarry Stebbings
Wow
- JLJason Lemkin
... and it's already better than any of the humans on our team It does. It, it creates, it creates the quote, it builds the contract, it ships the contract, it gets it signed. It updates the opportunity in Salesforce. It logs into Bill.com, where Rory's an investor. It sends the invoice. It follows up on the invoice. It gets it paid. It interacts with Brex, and it closes out the entire transaction. Then it logs into QuickBooks and makes sure for the first time in 10 years our books are accurate. They've never been accurate before, and the revenue's properly characterized. The agent does all that. This is with the models today, and which are not tuned for this workflow.
- RORory O’Driscoll
Can I be a customer?
- JLJason Lemkin
It's so good. It's, it's so good.
- RORory O’Driscoll
That'd be awesome.
- JLJason Lemkin
We just got... We had a human that for Sastry AI annual this year, where Rory was at, forgot to invoice $80,000 of revenue we had to write off. Now listen, I can afford it, but it was so frustrating that this person just didn't do it, and had no... Didn't even explain any reason why, just didn't do it. So Amelia's like, "I'm gonna fucking build an agent to do this i- in Ch- in China when we're, when we're there for Sastry." And then boom, like it's just, you know, it's, uh, i- it's, uh, it's better. It's better than the humans.
- RORory O’Driscoll
Did you have a full human being doing all that? Did you contract labor as a result of this, or was that person doing different stuff now?
- 49:09 – 53:00
Gross Margin Is Now the New Growth
- HSHarry Stebbings
Fast growth on revenue you don't keep is a trap." That, that was the tweet
- RORory O’Driscoll
I, I like Nicholas. We've looked at Algolia back then. I think he's awesome. I disagree with him on this. I, I... Look, I'll, I'll, I'll go further than that. The objective reality is that's not what was happening. In other words, companies with tough gross margin profiles and hyper growth have been getting funded and frankly have been im- able to improve their gross margins and pull it off. And you can... The foundation... That describes the foundation models, it describes the inference providers, and it describes the coding agents. So historically, what he's saying is not correct. Now, he may be picking on something in real time, which is that ability to build a company with a tough gross margin profile and then fix it over time probably makes the most sense in that kind of, you know, kind of big bang stage of AI, which I think was the last three years, where, you know, it went from nothing to something and you had 10X growth and it just paid you to, to grab the ground, like it paid Cursor to grab the ground. It may be that investors now are saying, "Hmm, as a first generation coding environment, you know, you can be Cursor, you can have negative gross margins, you can still be worth $60 billion 'cause you just grabbed the space." It may be that next generation they're seeing, 'cause they see a lot, you know, obviously, given the volume at which... They're seeing a little more focus on gross margins now, which is plausible. But there's no doubt that to date, the bet that my gross margins are shit, but my growth will cover it, and I will figure it out, even though it sounds stupid when you say it, has in fact been 100% true.
- JLJason Lemkin
I think he was synthesizing all the learnings across the Y Combinator portfolio and the, and this change, right? W- ca- is it okay for startups to have negative gross margins or no clear path to positive margins 'cause they'll figure it out? That's Rory's point. And I think you could argue both sides to Rory's point, but maybe that age is ending partially, right? May- maybe it shou- should end, right? You know, we're sitting here and, and Menlo just raised th- a $3 billion fund after 50 years on the back of a crazy bet on Anthropic that, that, that went very well, right, when the margins were crazy. But maybe that era doesn't last forever, and I would s- I'm sure all three of us are sitting on a portfolio company investment we made in the last 12 to 15 months where inference was the marketing strategy, and the gross margins were negative, and we're sitting here today and we're like, "Company's doing okay, but I'm not sure we're gonna get right side up on that investment." We're all sitting with a couple investments like that, and then we see a few others that w- are wildly efficient. They're wi- they're... Like, they took advantage of, of AI i- in other ways that's wi- wildly efficient, and we're wondering d- in that two by two, do we really wanna be in the, in, in the bottom... Is it the bottom left? Right? Highly inefficient and not, uh, better growth, uh, not top 0.001% of growth. I don't think we wanna be in that, that two by two, do we?
- RORory O’Driscoll
Everyone will have some deals where you go, "We thought we'd earn our way out of the gross margin problem, and for whatever reason, we couldn't, either because we didn't grow quickly enough to get to scale or because the foundation model company started grinding us, our competition bundled it in, and now you're just in a shitty gross margin profile business which will go bankrupt."
- JLJason Lemkin
Yep.
- RORory O’Driscoll
Because when things slow down, no one says, "Hey, let's do an acquisition of an adjacent gross margin negative thing and it's gonna be fun." Right? So-
- JLJason Lemkin
Yeah, that's a m- they'll-- all these startups will just fail.
- RORory O’Driscoll
Yeah.
- JLJason Lemkin
Right? If they don't achieve massive growth, they will all fail, right?
- RORory O’Driscoll
That's exactly right.
- JLJason Lemkin
That's the s- good simplification, right? And so it's just, for an early s- for an investor, that's not that f- that outcome is not that fun.
- HSHarry Stebbings
Jason, you mentioned Menlo. Uh, obviously, uh, one of the biggest winners from Anthropic. Raised $3 billion. I'm sure we've all got friends at Menlo. Awesome, good people. Happy for them. They're also in Lagora. They're in Lovable. They've been some of the, you know, best AI investors of this wave, I think, categorically.
- 53:00 – 1:04:30
Menlo Ventures Raises $3B: Why Not More After the Anthropic Win?
- HSHarry Stebbings
The interesting thing, honestly, for me, and you're gonna kill me for this, guys, child of the boom, whatever, whatever, why is it not more? They, they legitimately are one of the best. They'll deliver billions and billions of dollars back. In the wake of Thrive and GC and Lightspeed, this three billion's pretty conservative.
- JLJason Lemkin
I, I thought that. I actually don't, don't have an answer. The only answer I could come up with, right, uh, I do think... Look, not only is Menlo wildly successful, but they're, they're omnivorous. They'll invest at every stage. They will take lead positions, but they will also do smaller checks in small positions, right? That, that even a... And so they'll... You know, if they lose the round, they'll do 2% into the round or thr- t- maybe they'll invest more later. Maybe they won't. Like, they, they... If it's a great one, they'll just do it, right? Uh, kinda like how Felicis got going, but s- you know, any, any size, like 20%, 1%. So my own... A- at first, my answer is it, it's all they could raise. Poor guys could only raise $3 billion. [laughs] Okay, this was my first read. Rory's shaking his head, the professor. A- and listen, I don't know. I just... I shouldn't have gone first with the answer, but I think my secondary read is, look, they'll just raise another $27 million in other vehicles like they did for Anthropic. Like, this is just one or two funds, and they'll raise another $10 or $20 billion in, in SPVs, uh, or sidecars or other funds. The headline number of the fund size is not always correlated to the amount they'll end up deploying over the life cycle of those investments, right?
- RORory O’Driscoll
I think they were smart, right? I think the interesting... Yeah, to, to, to raise that amount, because yeah, they al- given their amazing performance, they'll always have access to more, and by not, you know, starting the clock on a much larger growth fund, you know, with the fees and the drag that that entails, you're just set up for success. I mean, remember, all other things being equal, provided you've got a big enough check size to play in any round, um, y- as a GP, you actually, quote unquote, "have higher risk-adjusted return" if you have smaller funds because then you have l- less cross-deal aggregation. In other words, y- y- two separate $1 billion funds, um, versus a single $2 billion fund, all other things being equal, you have more probability of winning on at least one of the, one of the two funds than on the single $2 billion fund. So that's not the reason, by the way, they did it, but just to point that out. I think the, the zoom out question is this: $2 billion is a lot of money. Like, how many deals are there going to be like Anthropic All OpenAI, all Android, where you can put one or two billion dollars to work. And the question is, do you, do you create your main vehicle for that, or do you accept that there are anomalies and you know you have access to that capital via SPV? So you run your business on what you think you can deploy across normal cycles on normal deals, including wildly successful, you know, ten billion, twenty billion dollar outcomes. And then you accept that, you know, two or three times a decade there's going to be a trillion dollar outcome and you wanna have access to that capital, but you can get that via an SPV. I think it's a reasonably rational structure. I mean, your fund size dictates your strategy. Once you raise ten billion dollars in a fund, you're signing up... There's relatively few places you can put it. You are signing up to put... You know, what are the statistics in Q1? Like 70% of the venture dollars went to four or five deals. You're really signing up to put a whole bunch of money in those deals, and that's the only place you can put it.
- HSHarry Stebbings
I, I, I kind of like it in the way that you're like, "Hey, we're conservatively sized and we can just take advantage of deal by deal carry on SPVs if anything does pop, and we'll have great fund returns if not."
- RORory O’Driscoll
Yeah. Yes. I think those-
- JLJason Lemkin
As long as you can get-- As long as you can spin up the SPVs on demand, it's a great mo- it's the better model.
- HSHarry Stebbings
Oh, dude, they'll be able to spin them up like never before. They've made a fortune with their LPs.
- JLJason Lemkin
No, I know. I, I, listen, I, I, I, as long as you can... Like, when I started investing, I had two hedge funds at LPs and they said, "Listen, we'll each give you a blank check SPV in your winners." And I'm like, "Okay, that sounds like a great LP." I took it. I called them up for a winner. Both of them said, "Well, we need to meet the founders. We need to do some diligence." I'm like, "This is the worst deal I ever got." [laughs]
- HSHarry Stebbings
Oh.
- JLJason Lemkin
You need... For the SPV work, it needs to be one, one, one WhatsApp message, and, and I get the, I get the two hundred million to invest. Then I'm all-- Then it's the best model in venture there is, right? I just need a hundred million. I know you haven't heard. It's a good one. Uh, did you read my last investor update? I need it by 5:00.
- RORory O’Driscoll
I think the impressive fact that I haven't internalized about that announcement, 'cause obviously I've known folks at Menlo since the mid-'90s, is that they're fifty years old, right? As a firm-
- HSHarry Stebbings
Which makes them thirty years younger than you. [laughs]
- JLJason Lemkin
Is Menlo still with the firm? Is, is he... Is Mr. Menlo still with the firm? Do we know?
- RORory O’Driscoll
Mr. Menlo is not with the firm. He-
- JLJason Lemkin
Does he still come into the office a couple times a week? Does he have an office in the back or is it in the front of the office?
- RORory O’Driscoll
No, but stop, stop, guys. Keep it sane here. My point is this. They're pursuing a strategy that I think is designed to survive and be successful across the cycles rather than... You know, my worry would be if you were reaching and doing a ten billion dollar fund. You've got to be really sure you can put that somewhere and without changing the strategy dramatically. It's not impossible. I think Founders Fund have done it very successfully. But again, to repeat, the number of companies that can ingest billion and two billion dollar checks is pretty limited, even on a decade-by-decade basis. So I give them huge credit. I think they're- they've done an amazing job with Anthropic, and they've done an amazing job lasting fifty years.
- HSHarry Stebbings
But speaking of people crushing it, Kalshi passes two billion run rate, starts prep for IPO rumored to be. How do we think about this news?
- RORory O’Driscoll
Americans like to gamble, and we weren't allowed to gamble for years and years and years and years. And then the Supreme Court said, "Screw that," so Americans started to gamble, but the states regulated it, and that's why, you know, FanDuel and DraftKings did well, but there was always regulations. And Kalshi found a way to pretend that it's a prediction market, found a way to get, you know, US jurisdiction from the CFTC, and have convinced everyone that it's predictions, which is different than betting. Ninety percent of what they do is sports betting. They're on a roll. They found a regulatory arbitrage to a wildly popular pursuit, says the person who's betting as we speak on the World Cup, right? So it's ninety-- it's eighty to ninety percent sports betting. Sports betting is really popular in America. It was illegal 'cause of the Puritans for the longest time, and these guys are riding a wave. End of complex analysis.
- JLJason Lemkin
I do think if Meta really can, um, copy it, right, for real, like if they're comfortable, if they're comfortable going as far as Kalshi has gone in poly market, I do think there's a chance they'll clean up. It would be so convenient to go into Facebook, right, in my feed and or, or, or Mess- or whatever, and just instantly bet on anything my friends are betting on. Like, the social aspect for Facebook could be so powerful. I just don't know if they'll, if they'll cut the same corners or not. I'm, I'm ignorant, but I, I actually think this is one where Meta could win big.
- HSHarry Stebbings
Uh, I, I-
- JLJason Lemkin
Because, look, I'm a little bit slow to Kalshi, but if I lo- but, but, but if I logged into Facebook and I saw that Harry was doing something for five bucks or ten bucks, I might just do it with him on Facebook, right? Uh, Grandma and Grandpa might. Like, there's a whole... Kalshi's scary. What does it mean? Is it safe? Right? The whole world... F- like, Facebook's still doing pretty good, right? I think it's a great place to bet. It's a great place to bet.
- RORory O’Driscoll
To, to remind folks, Harry started by asking the question, "Why is Kalshi doing well?" And Jason, when you jumped in, you, you, you covered something that we hadn't explicitly said, but just to say it, Kalshi's doing amazing and is talking about going public. It's doing two billion in revenues. Facebook, as in Meta, just said they might offer a competing product, which right now doesn't exist. But you're speculating on will that take the business from Kalshi. If that's Kalshi's only problem, I think they'll be just fine. All right? I think-
- JLJason Lemkin
Yeah, no, no, I'm not saying it'll hurt Kalshi. I'm just saying I could imagine-
- RORory O’Driscoll
Yeah
- JLJason Lemkin
... it being wild, wi- instantly wildly successful. If they're comfortable legalizing betting on their own platform and making it as elegant and as fun and as social as these platforms are, um, I might do that rather than mess around with these llamas.
- RORory O’Driscoll
I don't know. I think the demographic for sports betting is young and male, and I don't think that's the Facebook demographic anymore within a million years, right? To be honest. I think, I think it's actually smart of Facebook to think about it because, yeah, I think the, the f- the, the core Facebook demographic is aging fast and this is, um, this is where, as I say, especially young males are playing. I don't know if it, it'll turn out to be good business for them, but I, I I, I don't think they launch it on Meta and they get a whole bunch of traction. I mean, my son's an avid bettor. I worry about that sometimes. And let me tell you, he ain't-- I don't know, he's been on Facebook for, you know, five years.
- HSHarry Stebbings
But le-le-le-let's, uh, just be direct and chronologically correct. Kalshi, what price does it go out at when it does go out?
- RORory O’Driscoll
Who the hell knows? I mean, you know, they're talking about ten times revenue-
- HSHarry Stebbings
Well, it's the Prediction Marketplace, Rory, so let's predict. [chuckles]
- RORory O’Driscoll
That, that's actually a very fair response, Harry. You win on that. And, you know, and that would be good because it would be one of the ten percent bets that isn't about sports betting. Look, the growth is so amazing that, you know, if they're talking about ten times revenue, if it's at two billion today, by the time it could go public, it could be, you know, that could only be seven or eight times. So it could be a really big win. I, I looked at it before I came on, but I can't remember. FanDuel and DraftKings. So one obvious question is to what extent, why do they get revenue from FanDuel or DraftKings, which are a hundred percent betting? Those guys have some kind of geographic issues. They have to have state-by-state licensing. Kalshi, because it's regulated as a prediction market, has been able to avoid all that, so they might have an edge on that. And obviously, it's not quite the same thing because it's-- Just to be clear, the structure of the business is not quite the same thing as FanDuel and DraftKings are a classic betting house, and in the case of Kalshi, they're just a clearing house, and they match buyers and sellers. So it's not quite the same thing, but to a rounding error, the experience is much of a muchness. So I don't know. I mean, I think it-- right now it feels like a really good play. I-
- 1:04:30 – 1:12:22
Accenture Plummets 19%: Why AI Is Destroying the Consulting Business
- JLJason Lemkin
I don't know, un-understand all of it. I just think it's interesting i-i-that, that, uh, you know, thirty days ago it was a co-- it was an AI beneficiary. Now it's not, right?
- HSHarry Stebbings
The context is Accenture plummets nineteen percent. That's on already being down I think about twenty percent, so it's down about forty percent for the year. And to Jason's exact point there, I thought these were meant to be services that benefited from AI. Why is it down forty percent year on year?
- RORory O’Driscoll
And yeah, and I think I can chime in on that because I think two separate things are happening at the same time for Accenture, right? One is the business of helping companies adopt gen AI is probably exploding because all the companies need help. So if, you know, that was a zero part of their business a year ago, two years ago, three years ago, it's probably exploding and doing really nicely. Every company in America needs some help on gen AI, right? Separate comment. The core business they're in that, you know, ninety-nine percent, ninety percent of their revenue is, is other consulting, and there are few markets more primed for disruption by AI than consulting in general because it's white collar work. It's very-- It's, it's already been outsourced. That's why [chuckles] Accenture has it, right? And we're seeing a whole series of companies kind of-- The mental model for us is AI for SI, right? So the whole systems integrator market is prone to disruption. If you're ex-- And so if you look at some of the bus- the core business that Accenture used to do, you know, SI consulting for an SAP deployment, there are companies like Tessera and Conduct that are doing that. AI, um, SI, systems integration for Salesforce integration, there are companies like Swantide and others doing that. In all these markets, A-Accenture was probably billing, you know, maybe twenty, forty million dollars for an SAP implementation, and today they might only get twenty million dollars for that because twenty million of that can be done via using LLMs, right? So I think what happened is their core business came under pressure. So even though the new business of, um, you know, kind of helping com- other companies adopt AI [chuckles] is exploding and doing nicely, ironically, the core business defending yourself against AI is in a pretty tough place, right? It's not the end of the world, but it, it totally makes sense, right? Because I actually think one of the investment themes we'd love to find an interesting bet on, and have talked to some, is around this AI for SI space, right? Because all those consulting dollars, they're massively vulnerable to com- you know, compression from AI. Because if you look at the kind of tasks they're doing, it's gathering requirements, it's building statements of operating procedure, and then it's writing, you know, fairly simplistic code to say, "How am I gonna deploy Salesforce? How am I gonna deploy, um, SAP?" Right? Those are precisely the kind of roles that AI will replace.
- JLJason Lemkin
Yeah. When I was at a, a VP at Adobe, there was an entire floor of Accenture for five years deploying Salesforce, an entire floor of people Right? And I don't know what they made off that deal, but it was $26 million a year for Salesforce, so I'm guessing they charged at least $26 million a year for five years to get Salesforce up and running, right? That business has to be partially disrupted, uh, in general, and, and, and also seat compression hurts them, right? Anyone, you know... The, the one thing I would add to the list we get, I didn't know Devon was on such a roll. We talked about it right before we started. Devon, up 30% this year, one of the big winners. You know, we talked about it. The IPO crashed. It's up 3X from the bottom. And when I kind of did the other day, I was just trying to do my own little amount to oversimplify the public markets. Everyone selling seats, for the most part, is getting crushed. Everyone selling variably, one way or the other, is winning. Not all of them, but pretty damn close, right? Variable might be because it's directly attached to AI spend, right? But it might just be because it's attached to the economy that's doing well, right? And so a lot... The Salesforce Accenture thing is tied to seats, man, so that's gonna get crushed by definition, right?
- RORory O’Driscoll
I, I think you're right, Jason. The seat element is part of it. The long-term trend isn't great, but I think the short-term crushing is less because the number of Salesforce seats at Adobe went from 1,000 to 900. That sucks, but to your point, the real point is you had a, you know, you had literally a floor consultant sitting there getting $20 million a year for three years, and when you look objectively at the work they're doing, it's probably the, some of the easiest work for LLMs to do. It's automating-
- JLJason Lemkin
Yeah, it was migrating from Siebel and, and your own databases to Salesforce. It took five years to lift data, right? And now Databricks, which is on fire, we won't hit it this week, Databricks claims they can do that lift in 30 days for their customers, an entire lift to Databricks for anybody using L- It is actually under-discussed story that Databricks promised... What are they growing, 80% at 6 billion or something like that? Like, accelerating. I didn't really get this until one of the founders came to Saster and, and then I researched it, but they tell you, "In 30 days, we will do an LLM lift of all of your da- all of your data, whatever, how much it is," right? I'm sure there's exceptions and answers and whatever, but compare that to a five-year lift with Accenture to go to Salesforce. Um, and we just did one, like, we've been trying to get off Marketo for five years. It's our worst software. And then believe it or not, Salesforce used their LLM thing and moved us to their product in a couple weeks. They just lifted, they just lifted it with no humans. It's a big deal, this lift. Like, it's under-discussed, and it's a moat destroyer.
- RORory O’Driscoll
It is a moat destroyer.
- JLJason Lemkin
It's a moat destroyer when LLMs will lift you from one vendor to the other. And, um, I, you know, I don't even a- like, I, I literally was doing a pitch this week, and the founder was going on and on about their moats, and I, I, um, I immediately didn't want to invest. Like, I just... E- enough. I don't c- you don't... Your moat can be LLM lifted away. [laughs]
- RORory O’Driscoll
And the interesting thing, staying with the kind of Accenture problem is, is that, look, I mean, one of the things we had done about two years ago is just say to yourself, you know, when you, when you think about what work AI will replace, you know, white-collar work, you can say BPO is a good proxy for that, 'cause anything a company's willing to outsource to India, they're probably willing to outsource to AI, right? So in fact, you can literally just take the BPO spend and look at the BPO and start saying, "Okay, there's a whole bunch of places where AI will win," right? And you know, when you do that, Accenture's top of the heap, right? They have all that business. And what's challenging for those companies, you would think that they would want to adopt this technology, but the problem is they, their business model, there's only one thing worse than a seat-based model, Jason, and that's a model that's based on bodies. If your business as Accenture or any of these SI companies is, "I bill out 100 people, I pay them 200 grand a year, I bill them out at 500 grand, and that's how I run my world," if I don't need 100 people, if I only need 40 people and some AI, that makes my head hurt. I got to get rid of these 60 people or find another project for them. I got to figure out how to bill that. My whole margin structure collapses, and my take being on top of the heap goes down because... So what we're seeing is some of the biggest consulting companies, you know, they're adopting AI technology, but they're really struggling to pass on huge price increases, which leaves the room open for newer companies to come in and say, "Look, we are AI first. We are using these tools. You got a bid for 80 million from Accenture. We'll do it for 15." And that kind of thing gets the attention of the CIO, right? And we think that's a super interesting place to invest. So yeah, A- Accenture has a lot of structural questions. I mean, yeah, this is gonna be a constant theme there for the next couple of years, the pressure on those kind of businesses.
- HSHarry Stebbings
Welcome law, selling hours.
- RORory O’Driscoll
I- i- it's a little like the law comment, but even worse, I would argue, because I think law at some level, and I know, yeah, a law at some level, yeah, you want the work done, but you also
- 1:12:22 – 1:18:10
Work From Home Is White Collar Fraud
- RORory O’Driscoll
want the wise guidance. You know, if you're doing SAP migration, you just want the damn thing migrated to the new version of SAP, and you don't want to talk to these people ever again.
- HSHarry Stebbings
Okay, we're gonna do a rage bait but real. Um, we, we had a little clip that went slightly, slightly rage baity this week, uh, five and a half million views. Ryan Peterson at Flexport said, uh, kind of jokingly, obviously, um, "Work from home is white collar fraud. Um, I have two kids. When they come home at 3:00 from school, of course it interrupts my work. Yes, I have a home office, but of course it does. It's not the same as in person." Rage bait or real?
- JLJason Lemkin
I, I, I, I saw it. Um, I thought it's just dated, was my read for what it's worth. What I mean is, listen, I, I get in a little trouble on this show. Uh, of c- I agree with him. Um, I, I think a lot of work from home, the era was working 15 to 20 hours a week. It, and it also involved a lot... I, I, I saw it on my own team. Uh, and I, and it involved a lot of distractions from home, right? On the other hand, plenty of folks do make it work, right? Um, the reason I say it's dated is companies just, like, companies aren't hiring people. The companies we want to invest in, j- and this is a very narrow set of the universe, they're not hiring folks that want to work 20 hours a week from home. We're just not hir- like, our portfolio companies, at least our newer ones, our older ones still are, but our newer ones just aren't h- like, the whole Way you build a startup to your first 100 or 200 employees, I think is radically changed and is under-discussed. It is under-discussed. When we started this podcast 60 weeks ago, it was toxic to be running cognition and telling folks they had to run seven days a week and laying off half of Windsurf when you acquired them 'cause they weren't willing to work hard enough. That was utterly a toxic thing for the founder to say. Today, it is how you build a winner. You can't win in your marketplace if people are working 20 hours a week. You can't win. And so the only thing I thought, Ryan's running an old company, right? It's old. Flexport is old, and I think he is struggling with trying to modernize his team and being competitive with the way startups are today. He's struggling with the fact you can't change out all-- your entire team. What do you do with the folks that are unwilling to change at these big companies? What do you do with our 10-year-old portfolio companies where 90% of the folks are unwilling to change? And that's what he's rai- really railing against, is the folks that, that, that really just wanna work 20 hours a week. And so, but I, I think it's dated, right? I literally, I, I hope that no new investment I make is structured this way, and I feel like the majority of my older ones are.
- RORory O’Driscoll
When you say structured this way, can I be direct? Is that, 'cause I, I, I-
- JLJason Lemkin
Yeah. I want small, high-paid teams that work in the office over six days a week. That's-- I'm not interested in investing in anything else. I'm just not interested. And it's not because I don't have empathy. It's because they're gonna fail.
- RORory O’Driscoll
Yeah. I, I, Jason, empath is the word I think about when I think of you.
- JLJason Lemkin
I am an empath.
- RORory O’Driscoll
It's okay. I, I... Let me just ask a question here, right? Let's do a simple, I'm a c- uh, pretend I'm a consultant, poor guys. Um, a two-by-two, right? Work from home. Do you th- um, is your dimension of objection is, there's, there's two dimensions is, do you wanna work hard or not? Uh, in other words, yeah, hardworking, not hardworking, and then effective, not effective, right? Do you think the problem with work from home is that people don't wanna work or do you think it's not effective?
- JLJason Lemkin
Yes.
- RORory O’Driscoll
But which one?
- JLJason Lemkin
It's both.
- RORory O’Driscoll
I'm just curious. In a-
- JLJason Lemkin
We need to hire, uh, teams, teams that are half the size they used to be, that are paid top of market. They get double the equity because the teams are smaller, and we, we come into the office, frigging, the Corgi guy, what- what was his name? What's the insurance guy you had on the show that people made fun of?
- HSHarry Stebbings
Nika. Nika.
- JLJason Lemkin
Yeah, with his cafe. I wanna invest in startups with a cafe that runs 24/7 because you can't win when Corgi's running 24/7. You're not gonna win. It, it is a, here's the pro- here's the problem, and I'm struggling with this, uh, intellectually. It is not a sprint, it is a marathon, but it actually is, today it's a series of endless sprints. This is so hard because it's a sprint. You get about five minutes to relax and then today OpenAI released the Jalapeño chip. We didn't even know it was c- coming out today. Now they have their own inference chip. May- maybe we don't need Cerebus anymore. May- maybe the whole market's gonna be disrupted in 60 days because OpenAI is gonna run its own inference. Do we get to breathe, guys? You don't get to breathe anymore. Sorry, guys. We no longer get to breathe if you wanna make any money f- do you wanna make money from your equity or do you wanna make $180,000 a year? Like this is gonna be your choice in tech. Do you wanna make money from your equity or do you wanna be well-- or do you want a watch? Honestly, this is the choice though. You want a watch or you want to make 10 million or 100 million? There's nothing in the middle. You don't get to make 10 million for working 18 hours a week. You get a watch. You get an Omega. You want an Omega or you want to be rich? Make your choice, boys. And, and pick. And I, "Pick your path," I say to people. Pick your path. Go work for that old software company growing 8% a year. Go work there and make your one eight or two twenty and wear a really nice-- they all have nice watches at these companies now. Or go work in the office six and a half days a week at the Corgi cafe with a chance to make eight figures. That's the choice today. This is a different world, and I don't want to invest in anyone in the middle, and I can't afford to invest in the folks that want watches. Enjoy your $12,000 Rolex or your $8,000 Omega. I ho- I hope it's meaningful to you. I ho- I hope it's meaningful to you.
- RORory O’Driscoll
Okay. I'm letting it go.
- HSHarry Stebbings
Boys, do you want, do you, do you wanna have one final addition in, in the five minutes remaining? And it's OpenAI have announced that they are doing a custom chip. If Google has TPUs, Amazon have Trainium, well, OpenAI now has Jalapeño, co-developed with Broadcom. OpenAI says this chip beats current state-of-the-art GPUs on performance per watt. Broadcom CEO on record saying it cuts costs by 50% of a typical GPU.
- 1:18:10 – 1:19:07
OpenAI Launches the Jalapeño Chip
- RORory O’Driscoll
Inference is 50, 60% of revenue, plus or minus, and we know, yeah, peering through CapEx, when you look at CapEx, GPUs are well over half of that, so that 30%. So maybe if you could replace all of them and c- and NVIDIA makes 70% margins. You can make some kind of intellectual case for you can save some money with this. But honestly, my real response to it is you've got plenty to be doing elsewhere, OpenAI, winning on the top line side. I'm not sure I'd spend all my time optimizing. I, I'm not, I hope this isn't distracting you from the things that matter, right? So yeah, I'm, I'm trying to give a shit. Um, I-
- JLJason Lemkin
I'll throw out one thought and maybe we can talk about it next week for what it's worth, 'cause we didn't, one topic I thought we were gonna do more this week, which we didn't, is talk about open source even more, right? I think if you're, I, I do think that open source as token maxing comes under mo- becomes
- 1:19:07 – 1:27:51
Open Source Is Hollowing Out the Middle of the AI Market
- JLJason Lemkin
a bigger deal, I think open source is more and more important, and I think the, how do you win for certain workloads if you're open AI or anthropic? Well, it's you cut your inference costs cheaper, actually cheaper than commodity open source providers can provide it because they still have to buy the GPUs. Like j- open source is not a ma- again, it's not free, right? Listen, if I can s- if I can, if I can run my own inference farm d- one way or another buying somewhat expensive NVIDIA chips, which is that I, how I have to do it, right? And, but, but, but the advantage is I get to keep the margin, right? Or I get to keep most of the margin. Or OpenAI can provide me with a cost competitive product because its inference costs are half of what it costs me to do it on open source, this is not just about cutting costs. This, this, this could be about, uh, shoring up yourself from open source, which is go- which has the potential to kind of destroy the middle of their market. Like, the f- the-- people are always gonna use the best models for frontier applications, right? And actually, the, at the bottom of the market, th-they're, these guys are pretty competitive, but the middle is open to massive disruption. But if you cut your inference costs in half, the truth is open source is probably only twice as cheap for a lot of use cases, and if you cut your inference in half, your, your model still makes sense. I-- that's why I think in theory it's a, it's a big deal. It's not just about capacity, um, and driving down costs. It's about this m- this, this flabby middle. The flabby middle in open-- in AI is at, is at risk.
- RORory O’Driscoll
Okay. Um, I, I couldn't articulate-
- JLJason Lemkin
Next week you're gonna agree with me.
- RORory O’Driscoll
No, no, no. I, I, I'll disagree right now. I, I understand the bet, right? OpenAI and Anthropic have discovered the single best tech market in the, in terms of consumer demand in the last twenty years, and they should put all their effort into meeting that demand, right? Vertically integrating backwards down the stack, right, two levels down. Not just vertically integrating to own a data center, but vertically integrating to own a chip that goes into a data center doesn't strike me as the highest and best use of resources. Why do I say that? You've got three or four cloud providers who are dying to do business with you. You've got Oracle, you've got Google, you've got Mic- Microsoft, you've got CoreWeave. So you've got a whole bunch of vendors one level down for you. Some of those vendors themselves have chips. Google has a chip, um, Amazon now has a chip, right? There's a whole ecosystem people breaking their picks to provide you with cheap compute and taking on all the capital risk of that. God bless their poor little selves, right? Oracle will rue that day sometime, as will Microsoft. And you're sitting in here, and you should be putting all your effort right now into, you know, grabbing that cons- grabbing those end customers and enterprise. If you've got additional time and resources, then you can also take on building a chip. Have a go and knock yourself out, right? And I get it intellectually. You're right, Jason. At some point, to the extent it's become a cost business, you know, optimizing the whole vertically integrated stack might, may make sense. But the whole point-- I mean, if you look at it, you said it, but right now, if OpenAI and Anthropic had to do vertical integration today, they'd be fucked. Because the whole reason they worked is because they have outsourced three hundred billion dollars in CapEx to other poor fools, and the definition of vertical integration is taking in-house things that were done outside, right? The whole reason the OpenAI and Anthropic models work is because other idiots have spent the three hundred billion dollars on their behalf, right? At f- I mean, at some point, maybe you have to vertically integrate, but it just, it doesn't strike me as the right-
- JLJason Lemkin
Well, look, I, it's not that I argue with you. And first of all, the fact they launched this as we're recording this is obviously a decision that was made in a different era.
- RORory O’Driscoll
Yeah.
- JLJason Lemkin
Right? It was even before the TBPN era, right? It was an era of abundance. And so if this decision was made today, w- it would be a more, it would be an interest- very different conversation, doesn't make sense today. So all that caveat aside, we're actually looking at a, you know, some sort of e- early twenty twenty-five decision when the world was radically different, okay?
- RORory O’Driscoll
That's a good point.
- JLJason Lemkin
Having said that, maybe we can pick it up next week. I think the, I just think that OpenAI and Anthropic have massive existential risk that didn't exist at the start of the year, which is we di- we glossed over this GLM five point two open source whatever. W- what, it's not just performance. It's the fact that as we come under cost pressures, the middle of their market is under massive threat. And what you would do if you had a high-margin product like software is you would just have a, a, a cheaper offering. You would just subsidize the middle of the market. Th- but what ha- actually happens is they only have a h- re- okay, forget about... They really only have the high end and the low end today. They have Sonnet, Opus, and Friends, which is great. And then they have Haiku and Mini, which are these super small mar- mar- mar- m- models. They don't have this middle product. And the problem with the middle product is it's too expensive to provide. And so if you can cut your inference costs to below open source inference costs, then you can have this middle market where every single workflow you can serve rather than have your middle, your flabby middle hollowed out for you. I think the flabby middle is high risk for these. Just when they're all ready to go IPO, they have a n- brand-new existential risk, the flabby middle, just when things are getting good.
- RORory O’Driscoll
Agreed on the cost side, and I totally agree that. I'm just questioning is, if in order to minimize your cost, you have to vertically integrate backward two steps, I not only own the data center, I query whether you really need to vertically integrate it back through the hyperscaler and also integrate into the chip level. It just feels like a lot of backwards inte- backwards vertical integration that where in a competitive market, and there are three or four chip providers, and there are three or five or six hyperscaler providers. Surely you can just beat the crap out of them on cost, 'cause you are the largest buyer on the planet of this shit, right? And remember, you ju- I mean, how-- Okay, let's actually check, let's check if the market cares. Um, why do I say that? Well, we should actually see how much the stock-- what happened to Cerberus today. If you remember, Cerberus went public. They're-- had a, you know, decent quarter, going really well, and their big traction is a twenty billion dollar chip order from, um, OpenAI. Uh, did the stock go down a lot? Because presumably that is now... I mean, you know, yeah, it did. It went down sixteen percent. So there you are. The market said, "Poor old Cerberus. Um, OpenAI's gonna build that chip instead." Now, the question is: Would they have done just as well by saying to Cerberus, "You know, we're paying you off versus Titanium, TPUs, and Nvidia. We need a twenty percent discount." Did they need to build their own chip to do that? I don't know. Right?
- JLJason Lemkin
Well, it w- it doesn't matter-
- RORory O’Driscoll
Do we-
- JLJason Lemkin
-because it's so many generations ago. The decision was made is irrelevant today. The world's changed so much, right?
- RORory O’Driscoll
That's totally fair. Um, which by the way, gets to how these cycles end. Peop- I mean, I really believe what you're saying is if they s- The next stage is when someone comes in and says, "What we need to do is build a foundry and make our own DRAM." At that point, you'll know the cycle's about to go really badly down, you know? If you vertically integrate back into memory, then you know it's over. Just you and the, you and the Koreans.
- JLJason Lemkin
Look, I'll just say one thing maybe for the... As the- we'll f- we can find out next week and as the months go on. I, I firmly believe, especially for... We just want to tie this to B2B software and stuff for some of the heart of the show. I firmly believe we need these middle-level models from the closed source providers for the business to work, okay? We can't... Not, not everyone can afford to run every workflow on Opus, right? And Haiku and Mini are too small, and this is where open source is gonna disrupt the market. And so this is imp- this is more important than it, than it looks if it can work because you gotta head off this middle disruption, this middle layer.
- RORory O’Driscoll
Jason, first of all, I want to say something. I totally believe it. I think the number one question is w- how does software companies, how will they access kind of mid-priced, high-quality intelligence that's not got frontier pricing? Because we're seeing people start to gag on pricing. I totally agree with you on that. In my comment, I don't think it's a question of, you know, optimizing the stack that gets you there. It's, you know, what's the business model for these guys to provide it? And maybe with only two frontier providers, there's not enough competition. If I was someone like Google, I'd be looking at... This is why I go back to, come on, guys, wake up. I'd be looking at this market and saying, "How do I put a lot of pricing pressure on Sonnet, on Opus, and on, you know, GPT-5.5 by providing a just almost just as good US-based, you know, um, competitively priced product?" I mean, to me-
- JLJason Lemkin
I think they have.
- RORory O’Driscoll
Yeah.
- JLJason Lemkin
Yeah.
- RORory O’Driscoll
They just haven't made it happen. Okay.
- HSHarry Stebbings
Boys, it's so good to have you back from China, Jason. We missed you. Um-
- JLJason Lemkin
Well, l- I, I thought the guy from Benchmark was pretty good. I don't mind being replaced
Episode duration: 1:28:01
Install uListen for AI-powered chat & search across the full episode — Get Full Transcript
Transcript of episode LDBff24uaDQ
