(Preview) Meta Continues Its AI Spending Spree, More Fun with OpenAI and Microsoft, ‘Apple in China’ and Related Matters
14:22 Share

Hello, and welcome to a free preview of Sharp Tech. Hello, and welcome back to another episode of Sharp Tech. I'm Andrew Sharp, and on the other line, Ben Thompson. Ben, how you doing? Doing well, Andrew. For the record, employed by Stratechery LLC, not Meta. For now. Some people were asking questions.

Yeah, it's pretty wonderful, man. I mean, tumbleweed season has been interrupted by a spending spree that just keeps on going for Meta, and that's where we're going to kick things off. From the information, Meta Platforms is in advanced talks to hire the prominent artificial intelligence investors Nat Friedman and Daniel Gross to help lead its AI efforts, according to a person familiar with the discussions.

As part of those talks, Meta is in discussions about partially buying out Friedman & Gross' venture capital fund, NFDG, which holds stakes in top AI startups and is worth billions of dollars on paper, the person said.

If the talks are successful, Gross would leave Safe Superintelligence, which he co-founded with former OpenAI chief scientist Ilya Sutskever last year. At Meta, Gross is expected to work mostly on AI products, while Friedman's remit is expected to be broader. Both Gross and Friedman are expected to work closely with Meta CEO Mark Zuckerberg and Scale AI CEO Alexander Wang.

So, Ben, when we discussed this last week in the wake of the Wong news, you alluded to Meta's interest in other big names. What do you think about Nat and Daniel potentially being thrown into the mix at Meta here? Well...

This is definitely a disclosure podcast. Nat and Daniel have appeared on Tritechery multiple times to discuss AI. We initially started the series actually before ChatGPT where our perspective was, okay, GPT-3 is unthinkable.

unbelievable. Like, why is no one building products around this? So even we're going back to 2022, we're saying there is a product overhang. Like there is a new capability, a new technology that exists. And why aren't people building products around this? And it was interesting because I was hesitant to do it. As the other disclosure, I am personal friends with both Nat and Daniel. There are some of my closest friends in tech.

And so... And it comes through in the interviews. The AI interview series is one of the best AI resources I've had over the last two years or so. They're great. Right. So I... So...

Just that's on the table. I'm obviously going to proceed with, you know, just it is what it is. So I just but normally I don't talk about executive moves per se, except that. Well, I mean, you know, it's not just my friends. These are big deals, but they're it's interesting because they're headline deals that are not about AI researchers.

What didn't happen is Meta didn't buy SSI as a whole. Ilya Sutskover is arguably the greatest AI researcher in the field, or he's in the top five. And so it suggests, I think, one of the challenges for Meta, which is you go to Meta because you see an incredible terrain to build products.

you don't go there because you're like inspired you're like an ai true believer sort of thing you kind of know what you're getting into when it comes to meta and so in ilia obviously had that conflict whatever it was with sam altman and with open ai ended up going to do his own thing it's very highly valued it has no product uh they say they're not going to release a product until they have

whatever that is. So I just think the fact that Daniel's involved and not Ilya is interesting. And by the way, I've not talked to them about that specifically. So this is just sort of my surmising. But Nat was the leader of GitHub when GitHub created GitHub Copilot, which was...

Arguably the first major AI product. Again, it's not AI. It's the product. It's something you actually use. That was a huge deal. And he left Microsoft and GitHub Copilot's kind of fallen well behind. Things like Cursor and things along those lines. And so it's interesting that what sort of meta is hiring to date is really more about management than it is about sort of researchers. And

Maybe that's because they can't get researchers. I don't know. TBD. Again, there's a couple big names that are rumored to be headed in that direction. But I do think it's a pretty smart and humble sort of – this is some classic Mark Zuckerberg stuff where he's taken a heavy interest in AI. And to me, this says –

My interest in AI speaks to the fact I'm not capable of managing AI. I need someone to come in who really gets the space, who is not just a researcher, but is a manager who can actually direct and set direction and align incentives and do all those sorts of things and manage.

Again, totally biased here. If you're looking for the best manager of AI, it's hard to imagine someone better than that. He has the management experience. He understands the space. He knows what's going on through their sort of investment fund. And that's the other sort of exception with them. I typically don't have VCs on. When they started, they were just doing like a startup fund sort of thing. They were seed investors. I'm like, I'll just kind of squeeze it a little here. And then they got more and more like...

big into space. I'm like, well, this is so much for my no VC interview principle. We had a good thing going. So we've kept it up. I think the last one was in January. I was going to say, it's sort of like a biannual tradition with those guys. And it started six weeks before ChatGPT launched. So it's been fun to trace the evolution of the space through those interviews. Yeah, we'll put a link to the interviews. I assume they still hold up pretty well. It'd be interesting to sort of go back and read through them. But

To me, that's just the big takeaway. I've met Alexander Wong a couple times. It's interesting what the fit's going to be with him and Nat. But Nat is my age. He's experienced. He's managed large organizations. He's operated in large organizations. He was inside Inadela's inner circle at Microsoft. He understands what it takes for an entity like Meta

to operate in a space like this, probably better than almost anyone that you could imagine. There you go, yeah.

Obviously, I think it's a great move by Meta because I like Nat and Daniel. And Daniel, if you go back, Daniel's brilliant. You read through those interviews, and he's been on the product point from day one, talking about how the – like people – number one, there are no products. And then number two, the people doing products don't understand and aren't thinking about how AI transforms product capabilities along the way. So Daniel going to do products, Nat doing business.

Now they still need to tell it like at the end of the day, it's the engineers and the researchers that actually do this stuff. So this story isn't complete, but yeah, this is a big deal. Like it was, it's pretty surprising. I think it's probably safe to say they're buying out their venture fund. Like if we thought that $15 billion for, for,

scale AI was a lot. You know, I mean, whatever. Clearly, there's no holds barred. This is pretty nuts. I can't imagine what...

i don't know what the budget is it's got to be very large are getting pretty astounding man and you know i too am interested in the interplay between nat and alexander wong and we got yon lacoon still in the mix like if this were sports i gotta think yon's being sidelined to a certain extent i mean i think i expressed my my my yon take earlier again salute to him giant of the field um the wrong person for meta even if

All he is skeptic about LMS is correct. Correct.

That has nothing to do with the possibilities in front of Meta. And again, this speaks to why maybe someone like Ilya doesn't want to go to Meta because he's not interested in making a more compelling Facebook experience. Making everything on your Instagram photo clickable and an advertisement. That's not necessarily what Ilya is trying to do for the next five years. Right. And so we'll see what all this stuff sort of ends up looking like. But again, that goes back to our initial interview.

What I did with Nat and Daniel, their point from the beginning was about productization of AI. That's the possibilities that they were focused on. And so they're the sorts of people that make sense in a meta contest, the possibilities of product ties within metas,

you know, surface areas is massive. But again, you know, they, they, they're not, they're not the ones actually doing the programming or the research. Like, like, you know, I say this as an analyst who is not doing the program research and they will freely say as VCs. Now they can do it like Nat had net.dev, like when the earliest playgrounds for AI that he actually funded for ages, uh, very large bills where you could go and just play with the stuff and sort of figure it out. And he did that all himself. He did a super early, uh,

like a web scraper, like where the AI would surf the web for you. And he did this like years, like he's capable, but

He's not like the... He's not Ilya. He's not Ilya Satsukov. Like, are they going to get those sorts of people? Well, look, maybe Meta's offseason isn't over just yet. Well, yeah. It clearly wasn't when we talked about it last week, so... Yeah, over the next couple weeks here, Austin sent us this tweet. Breaking Meta to trade Jan and four first-round picks for Shulman and Sholto Bricken in a three-way trade. Shulman contract expiring at the end of the year.

and expects to sign a three-year, $200 million deal with Meta. Karpathy says, quote, he is happy where he's at, but CEOs around the Bay expect a signing to come soon. Should we get on the trade machine and start putting together the next moves for Zuck? Is it time for you to enter your Bill Simmons era? I saw the rundown of the entry of the Bill Simmons era. I wasn't sure what you meant. But yeah, I mean, there's a reason I don't generally talk about executive movement at the top. It's just like,

Even here, I'm talking about like, look, Daniel's going to be a great product guy and that's going to be a great manager. That

And it's interesting because in tech in general, there's a real bias and appropriately so towards engineering and the people that actually build it. And obviously me, I do appreciate management and the sort of fuzzy aspects that go into creating the circumstances and the incentives and the environment for great engineering to take place. But it's really hard to sort of like

know and talk about if you're not there, if you're not sort of in the building. And I do think that Meta is the only company that could have pulled Nat and Daniel in precisely because Mark's clearly in control. The ability to sort of start afresh, cut through red tape, is going to be higher there than it would be most other places. I

I'm sure that they were not the only company that's tried to hire these guys. But yeah, we'll see. So in terms of meta and what's attractive about meta, it's essentially Mark has unilateral control. Number one is Mark is in control, and if you have a direct line to him, and we just see evidence of meta being willing to take the risks.

the products get pushed to 3 billion people. So it's a massive opportunity if you care about making great products that the whole world is going to use. And Meta has a lot of great talent, right? Like even if the question about the very high-end AI researchers is out there and they lost some of the people that they did have, they lost a lot to Mistral in particular. Like when it comes to actual productization and engineering, like they...

They got a lot of talent. Well, and they're going to keep spending. It sounds like they're going to keep spending because the FT had a report. OpenAI chief executive Sam Altman has accused Mark Zuckerberg's Meta of trying to poach his developers with the promise of $100 million signing bonuses and even higher pay as the social media platform races to catch up in the artificial intelligence battle. Altman said Meta, which has a market capitalization of $1.8 trillion, is

had begun to make the, quote, giant offers to employees on his team after falling behind on their current AI push. And he said that on a new podcast that Jack Altman has started. A lot of new podcasts out there in the world these days. Yeah, why is everyone coming for me? Go for someone else. I mean, come on. Exactly. Well, my question is,

Are we entering an era where we're going to start to see the barbell effect hit comp for tech talent going forward? Because the engineering, it seems like there may be a downward push on engineering salaries over the next five to 10 years as AI proliferates. And then the type of talent that can really make a difference for a company like Meta

I mean, given the stakes, it isn't crazy to spend $100 million or $150 million or open AI throwing $6 billion of stock at Johnny Ive to get a hardware business off the ground. Like, is this going to be a new normal in tech going forward?

All right, and that is the end of the free preview. If you'd like to hear more from Ben and I, there are links to subscribe in the show notes, or you can also go to sharptech.fm. Either option will get you access to a personalized feed that has all the shows we do every week, plus lots more great content from Stratechery and the Stratechery Plus bundle. Check it out, and if you've got feedback, please email us at email at sharptech.fm.