You can gather around for a little chat. Well, welcome to day two of the summit. Mark needs no introduction. I think when Sam Altman was introducing Mark to us, Sam Altman calls Mark Chen the most important person in OpenAI. So I think the theme of this conversation is AI is something that we're all thinking about, whether we're actively working on, investing in, building around. But I think there's a lot of misconceptions and also just hot takes that I think there's very few people other than someone like the people in this group, but also Mark, to give hot takes on. So excited to have this talk as we just warm up our brain for a really awesome day. Any other welcome thoughts?
No. Yeah. It's been great. Thank you, Mark, for doing this. And obviously, hope everyone's having fun. If you guys want to do anything specific or any other activities, just let us know. But that's all I had to say this morning.
Sweet. Yeah, thanks for having me. Let's make this conversational. I feel like it's going to be boring if it's just one-sided. So, you guys should ask spicy questions. Let's just have a conversation.
Do you think people will fall in love with their AI and is it already happening?
Oh, it's very much already happening, right? I think even if it's like half a percentage of the users that get really attached, it's still a lot of people, right? And there is a growing group of people as the models get better, they can't tell that the model isn't a real person and they kind of engage with it in that way. And I think the dependence we've seen kind of grow, and the danger is just a company can really lean into that and make a super engaging product, right? Like you can imagine if you just optimize for the thing that makes you feel really connected to it, that would be a great business, but it's probably really world negative. So, I think it's just got to tread the line really carefully.
Maybe I'll switch to more of a business question. Like there's so much capex going into data centers. I think that there's the big data center deal that popped Oracle stock like crazy and that was because OpenAI has a big multi-hundred billion dollar deal with Oracle. I think when there was that tech dinner with the president, everyone was talking about multi-hundred billion dollar data center buys and then you look at the revenue numbers of even you guys are growing super quickly but it's like tens of billions of revenue. So how does that check out? What are the people not understanding in media that you guys are seeing where the smartest people are deploying hundreds of billions of dollars into capex?
Yeah. So I think there are two levels of the game. So the first level is all the hype you see about getting researchers from one lab to the others. But I think the deeper game is getting the compute and the energy. And that's actually the real battle. I think your researchers can be however good, but they can't be productive if you're not running at some multiple of compute better than your competitors. And so I think everyone's seeing that this is the bottleneck, everyone wants to secure that compute. The situation is all the labs, they're well-funded on compute, but everyone still needs more. That's the very strong feeling. If I had 3x, I think 3x I could deploy immediately. Like 10x, it would maybe take me two weeks and we could probably eat up that compute. That's how much the need for compute is. And so I think everyone realizes that and they're just trying to get all these bottlenecks. You have to prepare a long time in advance, these data center buildouts take a long time. There are bottlenecks in energy and certain components, in cooling and turbines. And if you miss one part of the supply chain, I think you're just completely cooked. So, yeah, I mean, everyone's kind of playing that game behind the scenes. And it's a very different game than talent where there are a couple key players, you know, TSMC is a big part of it. In memory, Korea is a big part of it. So, it is just kind of systematically going through identifying who's really important and just securing all the things that you need.
So, do you think it's fully priced in or are these things underpriced still?
I don't know. I think energy is still a really good investment, right? And I just think most people don't understand what's coming still. It's definitely not priced in. I think there are a lot of bears here. Like even just to talk about bears, right? We look in academia and a lot of people, they don't have any idea what the models can do today. You have like one or two people, let's say in physics, I tried to get someone to use GPT-5 Pro, they put their recent paper in it and it just solved the whole thing and they were shocked and they're like my colleagues have no idea this is the state of the world. And I think that's actually still the case in the world right now, right? Like most people just don't know what's coming. They don't know the implications. And so I think if you believe or you see the future just a little bit more, none of this is priced in.
How important do you think physical embodiment and 3D environments are to understand reality versus a language approach?
Yeah. I mean, I think there's a gradation. It's not just you have robots or you don't. There's kind of like digital environments which make a lot of sense too, right? Like you could have a model which acts on your computer and that is a form of embodiment too, right? It's like your world is the digital world and you could do whatever you need in the digital world. And I do think if you're talking economic value, a lot of that's locked up in physical labor, right? But when it comes to intellectual work, almost all of that is in the digital space. So I think it depends on how you define the scope of what you're trying to target. If you want to target knowledge work, you might be okay with kind of putting robotics later in the road map. But if you want to target just economically useful physical work, you have to get the robotics. So different firms have different approaches. OpenAI is very intelligence centered. So we want to crush the digital task first.
How far out are we from an affordable, to the average person in the United States, an affordable robot task doer in the home?
I think the scary thing here is that China's so far ahead on robotics. You see the Unitree robots, right? They're really good. And I think we're at least two years behind that here. They just have everything in house, right? They build up the manufacturing, they get the prototyping, it's all done, you know, I think maybe in the same city. And we don't have that supply chain here. But just given how good those robots are, I think in some places in the world maybe like two years out or something like that.
Would you trust a robot in your house to just, you know, farm off physical tasks to it?
Well, like one from China or... I don't know about that one. I would trust it if it was from a reliable company or, you know, like OpenAI or, you know, anyone Google or something like that. But it is still a weird concept that something is always there and listening. But I think that's where we're moving anyways. I mean, we talked about hardware and like having something to always wear that's always listening that knows what to do in advance. I'm very excited for that as well.
I think everyone having a house elf, like I think if it's a 7-foot tall humanoid, it's too scary. But I think if you have like a cute 4-foot tall humanoid, so like everyone has their own Harry Potter house elf, that's probably the winning form factor.
You couldn't put the cups up.
Maybe like the legs unfold.
They can jump up onto it.
So basically I think humans are quite primal. So if you feel like you can dominate a 4-foot thing, it's probably more accepting as the Trojan horse into the household. And then I think once people have house elf servants that are doing their bidding, I think that it's too much economic value.
Yeah, there's a real point to what you're saying. When people are thinking about development of robots, there's a difference between robots with tendons and just fully mechanical robots because you're not going to be able to push back against the fully mechanical thing but if there are tendons that can break or snap you feel at least like hey this thing's not going to be able to overpower me or something like that. So I think there's a lot of design elements here. Actually, yeah, like when you guys talk about this kind of wearable or something like that, I think that's really interesting because the way AI is going, I feel like it's going to look like your ChatGPT is just going to be your personal assistant or kind of like a digital house or something like that, right? And you want it there to be some way where it's seeing the same world that you're seeing. So, I think what you guys are thinking in that world makes so much sense, right? You should just passively be observing all the things that you're going through and it should learn about you, right? I think the models today, it's like every time you open up ChatGPT, it's like you rebirth a whole new thing. It doesn't have any real context on what you've been doing all day, right? And you have to prompt it and tell it all the things. But just imagine it's just soaking all this stuff up and it just kind of knows what you need, right, at any given point in time. I think that would be a really cool world. Just circling all the way back to the first question, just do I think you guys actually have more insight into this kind of stuff. Like, do you think people are going to fall in love with these things? Like, are they going to replace their human friends with AI friends? Is that what the world... I feel like you guys know the social world so much more than I do. So, I'm just curious how you think it's going to play out.
I'll let Jake start. I have some thoughts here.
Yeah. No, I think for a pretty large percentage of the population, I think specifically males that are seemingly in a weird place in life. Even more so, you know, more women are graduating college, etc. And so I think there's this male crisis going on a bit. And I think that they will find love within the AI sphere. And especially if there's physical companionship, I think it's inevitable for sure. I don't know how big it will be and I'm sure society will maybe frown upon it, but I think it's definitely still going to happen and people will have AI girlfriends in their house that they're in love with for sure.
Yeah, I would double down on that because I think when the next generation of people grew up in their formative years in COVID, their main human interaction is through text or video, which is pixels, which language models can mimic and be slightly more sycophantic and friendly and encouraging than any possible human because they have infinite patience and infinite context. I think it goes back to something like we were exploring yesterday, which I think there's just going to be... I think there's already in modern economies like producers and consumers and I think the consuming category of humans is going to be like 99%. I think it trends to infinite basically and there might just be one productive human left in like a hundred years and everyone is just consuming the labor of a superpowered running 50,000 agents superhuman or like that's the limit right? So I think the aggressive prediction is that that is where we're trending where there's fewer and fewer productive people that are creating experiences and products for everyone else that is turbocharged by AI agents and then more and more of us are consuming and maybe that's scary or maybe that's okay or maybe that's inevitable. Like I'm actually curious, what do you have a hot take of what human society looks like in a hundred years?
You're probably right in the limit but I do think shorter term it's going to get everyone more creative because all this manual, routine kind of execution work, the model could do it for you. And that's kind of the thing that people trust the models to do, right? I think we kind of enjoy the slice of the economy which is creative work. Humans just innately feel some kind of affinity to that. So the thing we will want to automate away is the routine and boring work. And I think short term you're just going to have a lot more people creating. Just imagine if the mechanics of actually creating a video are easy, but the ideation and that part, you're going to kind of enjoy doing that part and shift towards that. So, but yeah, maybe in the limit, the models are just going to create much more engaging content for us, right? We're all just going to have our AI friends and we don't really care what other people make. So that could be a shift longer term. But I do look forward to this world where we can all be creative and the mechanical execution that's the easy part.
Yeah, that reminded me, I think ChatGPT has made me more creative. The Paul Reserve logo is actually made on ChatGPT. I actually made it myself. So, that's actually a fun tidbit I forgot about. So, you can sue me for ownership rights. It's technically Mark's logo.
Yeah, let's... I know there's a couple questions. Steve. Yeah, but I just Googled it. Says like 34% of US adults have used ChatGPT and 58% of US adults younger than 30 are using ChatGPT now. So I feel like that's a lot of power that OpenAI has for the public benefit. Let's say maybe there's a declining fertility, would OpenAI ever tweak the models if they think it could benefit the public?
Yeah, that's a good question. And so I think actually one structure that we have right is we have a nonprofit entity which has as much control as essentially the production or research arm, right? And what we want is to just give it the resources such that it's basically figuring out these kind of questions for us, right? Like what way do you tweak the underlying behavior? We will often sacrifice profits for the right thing to do from a behavioral point of view. And it's kind of scary too, right? It's like you can, like I was saying earlier, you can build the most sycophantic glazing model that you can and really get a lot of users that way. But I think we decided to roll that back and we don't want to cause a mental health crisis and I think you pay a long-term cost for these kind of decisions. And yeah, I like that, that is the kind of thing that we're used to doing and there are people who shape the model behavior in ways that are counter to growth. Yeah, it's kind of crazy that we're honing in on a billion weekly actives. It's every week like one in seven people in the world log on to ChatGPT. It's crazy. Go for it.
When you also think about output as OpenAI now the company's about 4,000. I think a lot of even sentiment in SF is that the company's a lot more bureaucratic and slowing down now. So people go to other labs or other places. As a leader, how do you think about if that's true? What are you doing about it? Are you just going to remove half the people, focus people, do something else?
Really good question. I think one of the hardest challenges leading a lab like this is to fight the big company sink, right? That if you don't actively manage things, it just evolves into that. I think when you look at size, the rest of the company has grown a lot. Research hasn't. I hate hiring and it's about 400 or 500. And last year we were maybe a little bit over 300. And so, you know, like the rest of the company's doubled, we've gone up like 20-30%. And it's very deliberate. I don't want it to be one of these cultures where you just have this spiraling bureaucratic work. I run a lot of experiments like I earlier this quarter I said no hiring, manage out your people and really elevate the talent, right? And I think you just have to always hold everyone to a high talent bar. I tell people like you don't hire to the point where someone's just like zero marginal add. Everyone has to be extremely high positive marginal add. And I actually do think like on a talent perspective we have out-competed people over the last two years. We've basically gotten our pick of the litter. When you look at kind of these talent raids on OpenAI, I want to provide maybe a different perspective from what you hear in the media. So Meta, they went to every single one of my direct reports. They all said no. And we're actually very... when they come with that kind of money, you're not going to be able to say like we're going to retain everyone at OpenAI, right? But we can be selective about who we retain. And it's a strategic game there, too, right? They're going to be able to get some people with the amount of cash that they're willing to spend. But culturally, we've kept all senior leadership. They're all still very bought in. And I think we are strategic about the people who like to go to other labs.
