AI Foundation Models

“What matters is that no lead is secure. And projections don't mean anything, right? And so I think that's the real story is it's too easy to switch. Leads don't matter. It's going to be impossible to financially price any of this stuff.”

“Our belief is that there will not be a, my belief is that there won't be one model to rule them all. It's just impossible.”

“These markets are never dominated by one player. There's going to be an alternative.”

“And the capital intensity for every dollar of run rate revenue, it probably takes four or five dollars of CapEx to support that, right?”

“It's like whoever has cheap energy eventually wins as a commodity business with no defensibility.”

“And I feel like that's actually where Apple could come in, and without, even without necessarily owning, like, a fantastic foundation model, Could come in and just be like the interface by which normal people interact with AI day in and day out. For sure, I th…”

“Elon did an amazing job of throwing money at the foundation model problem, and he caught up very, very fast. And that's something that in the whole history of our industry could never have happened before.”

“I think at this point it's about products. And if people like the meta products that they put out there that are run on these foundation models, to be honest, they could probably be run on, on a lot of different models that are out there. I think it's partly e…”

“Ultimately, these things become capital gains for certain types of approaches or models. Because we know scale really matters, which means that eventually you have to collapse into a handful of players because capital will aggregate the things that are working…”

“We fully changed our mind. I think there's going to be application software companies built on top of models in pretty much every direction.”

“On the model side so far, the way it looks like it's played out is It will be more like the cloud industry. Like it's not going to be a winner take all. Certain technical advantages seem limited in timeframe, right? Like there's kind of, you know, always this …”

“These systems have this other effect where they're probabilistic, not deterministic... And so you can't just say, oh, deterministically, this is how much it's going to cost us. And so the companies are in this area of exploration right now, where that's causin…”

“That's why I think the open AI's of the world will continue to move towards applications because it's quite hard for them to make money longterm on like their API, for example, because There's like such high competition, all the labs are building, and it's lik…”

“And this is true even in AI, which I did not know. But people tell me now, because I am not that tech savvy. But people tell me now that even in the AI, the foundation, you know, models of AI, it's Rajiv's mind that's at work.”

“The reasoning capabilities matter a little bit less. I would say the models as they stand today are generally good enough at solving most of the meaty inquiries in customer service. I think it's like, instead of reasoning, it's more about instruction following…”

“Well, no, so I'm a firm believer that no, no one is going to win it all. The consumer and enterprise are different, that different, you know, like the act of software engineering is different than the act of marketing. And I'm definitely not a God model. One m…”

“Foundation models can be distilled relatively quickly. When I was on your show last time, I talked about how foundation models were sort of going to be the fastest depreciating assets in human history, like weeks. It's almost days now.”

“Building applications on top of these models becomes, it is a lot easier, and you can go from zero to one, and you can be more nimble than even these labs that are gonna all have, like, you know, tens or hundreds of millions of users, and you have to move slow…”

“If you are doing something at the very edge, I mean, I definitely am constantly trying to advise founders and encourage founders to think about what the models might do in six months or nine months and start to chart there.”

“I, you know, I think that ultimately the model layer, and I think this is not a hot take anymore. I've been saying it for a few years, but I think it will become commoditized. I think that a lot of value will accrue at the hardware layer especially And we coul…”

“If you're an AI investor doing very big checks and foundation model up, foundation model layer up, you haven't really had many write offs. You've in your losers, or I wouldn't say losers in the companies that haven't broken out, you've made your money back.”

“And this could easily take 10 years before we start putting these things into the clinic. But I think it will happen, 10 years, a lot can happen in 10 years.”

“That might happen. I think that's a very serious possibility, and I think this is actually one area where regulators should be paying attention. You know, what does this mean for market concentration, antitrust, and so forth?”

“Early stage funds can't invest in this stuff. We're already boxed out. We can invest on the streams that come off of it.”

“So it seems like the, you know, in general, I wouldn't be surprised if 20, 24 and 20, 25 is the year of proliferation of models where we're going to start to see an expansion in terms of the different types that are covered, you know, chemistry and material sc…”

“The idea that, like, we know how to, like, exactly, like, design for that sort of thing is low. You could try something, but it's not like, oh, I know for sure it's just someone's waiting to do it. Remember, you need data. It's hard to accumulate that. Yeah, i…”

“These are strong words, but I think they're getting commoditized until the next big innovation comes and you allocate some more value to the model.”

“We think that there's not going to be a single model to rule them all.”

“So there's a lot of focus on models today, but those models are going to remain important. But over time, there's going to be additional capabilities like agents, like reinforcement learning, like the ability to be able to understand and vet The responses that…”

“Because there is no doubt this is more important than five G.”

“So we saw image pretty much be solved last year and then text continually get better and better.”