"The Cognitive Revolution"
01:09:29 · Marc Bhargava
Can we debate if, you know, robotics is certainly getting better and enable robotics is getting better? Are we gonna eventually go to this other world, etcetera, etcetera? I mean, that's a very hard question to ask, and there's all kinds of technology and automation improvements. I would just say, me personally and General Catalyst, we are very focused more on what are the opportunities in front of us today. And we have been really mindful of the companies we back. We're a large investor in Anthropic. We think they do an excellent job of thinking about the risks around AI and AI safety. And so, you know, responsible innovation, AI safety, those are really key to our mantras and the companies we're investing in. And we think like every other technology, as we adopt this, there will be challenges. It's the mindset you bring to those challenges.
01:10:16 · Nathan Labenz
[01:10:16] S0: Yeah. I sort of have a vision of that classic Simpsons episode where Homer's just, like, hitting the enter button at every prompt as the nuclear power plant birth... Bill... Boots up until such time as he finally decides to put the the dippy bird in to be hitting the button, and then he kinda just leaves. The plant, I think, does end up having a meltdown as a result of that. But this sort of managing agent future, I'm a little bit like, you know, we... We're talking the morning after the Grok four presentation. Right. Mhmm. And I don't know. It feels like it's coming at us really fast. I mean, I'm like, I want everything you're saying to be true. I... And I genuinely believe that the consumer surplus is gonna be amazing. Just the quality of service, the access to expertise, the democratization of that, I think is all, you know, just super exciting upside. But then I see Elon saying things like, new science next year. And I'm like, I don't necessarily think I can count this guy out because for one thing, you know, they keep climbing these... You know, the... There's no benchmark seems to be, you know, immune to their ability to climb the hill. And as they make this transition from other training signals to just, like, learning from reality itself, you know, and they start to equip the models with, like, the same high powered simulation and other tools that their best engineers at Tesla and SpaceX and Neuralink are all using, and it starts to actually compete with them to, you know, redesign a part, you know, or, like, find a find an efficiency or, you know, can we eliminate this part? If the... If those systems can start to do that in the next year, it seems like we are in for a really choppy ride even if the, you know, even if the consumer surplus story, you know, is is totally true, which I I do think it it is. Is there a ratio of, like, how much you spend on AI to to how much you save on labor that you... I usually tell people expect 10 to one. But do you have a rule of thumb for, like, if you can take some task and automate it, like, what's the savings? And of course, that could be. You could do 10 times as much at best. But what is that ratio?
01:12:32 · Marc Bhargava, Nathan Labenz
[01:12:32] S1: Yeah. We're not really focused on the savings, to be honest. So we're focused on grinding up. No. If you [01:12:37] S0: flip it and say, how much more can you do with the same inputs? What Yeah. Average. [01:12:40] S1: There we there we have an objective. I think ideally in a three year period, we would like folks to be with the 20% cost increase, be able to double your revenue. So that would be like a broad target. But in three years, if we come in, we spend 20% more than you spend today to make sure you have the right software, the right engineers, the right AI tools, increase the cost 20%, can we increase the revenue a 100% in a three year period? And if we can underwrite, yes, we think that is a plausible case, then we think buying that company could be a good investment. But we are not looking at AI as a cost saving even though it of course can be that. That kind of limits your upside. We're looking at it much more how do we grow and have venture like returns and fast growth and scale by AI enabling our workforce to take on more revenue. So that's... So maybe our mindset is a little
01:13:28 · Nathan Labenz
[01:13:28] S0: bit different on that metric specifically. Is there anything that big tech could do that you are afraid of? I mean, one that I always point to is, and they swear they'll never do it right up until maybe they do it, is not allowing... You know, basically, using their best models for their own products first. Right? Like, so far, we've had this sort of generally even playing field of ChatGPT as a model and the API gets it roughly the same time, maybe even a little before, whatever, but it's, like, pretty much parity. But you can imagine a world where Claude five and Grok five and GPT five are all proprietary products only for some significant period of time, if not indefinitely. And that could potentially really shift the power. Now from a consumer surplus side, like, they probably still can only charge me so much a month because I've got four or five of those options. Of course. Yeah. How do I... How does that play into, like, the app layer? It seems like it could be a real problem for the app layer.
01:14:30 · Marc Bhargava
[01:14:30] S1: I think that's true if there were two or three players. So even two years ago, you could argue OpenAI in the lead followed by maybe Google that kind of fumbled the back there they were the OG with the transformer paper, but still credible and Anthropic, which had got many of the early OpenAI best folks. And, you know, there were two or three players. But then x AI came in and enter stage left DeepSeek and Alibaba and all these other folks. Now I don't think there's a world where, you know, because you have seven, eight pretty top quality model providers, obviously, Facebook with Llama is doing a whole refresh on talent, which is in the news. But there are just so many players now that the idea that all of them are creating closed models and shutting the door is extremely unlikely, and customers can always go and use deepseq. The Chinese labs are also quite good, and there's a very good technology coming out of there. So there are just too many players at this point who are open source, and it was really like a shortage of researchers. Two or three years ago you could kinda count who had worked on these projects because they were really Google OpenAI and starting to be anthropic. Today a lot of PhDs are being done in this space. There are a lot of other folks hiring Facebook, x AI, I would say both have kinda caught up as well to a certain degree. The companies out in China and Mistral in Europe. And so I think we're at a point now where there's a lot of kind of AI talent, AI researchers, AI companies that are creating these models for all of them to turn off and go close model. Maybe we can trade that on Polymarket or something, but I will put the odds at, like, sub 3% there. So
01:16:01 · Nathan Labenz, Marc Bhargava
[01:16:01] S0: Yeah. Okay. Yeah. I mean, the dynamics do change a lot from one to two to eight. [01:16:07] S1: Exactly that. [01:16:09] S0: Because, I mean, some other interesting factors around just like safety, reliability. I mean, there's a whole... There could be some other factors that could make it shift, but I do think I I tend to agree on a purely competitive market basis. There is a strong incentive to try to win the API market, and so there's always a... Even if you have something, you know, better in reserve, there's always a incentive to leapfrog your competitor and, you know, try to win share. Why are you telling this story publicly? You know, one first rule of having a great investment strategy is keep it to yourself. Right? Like Yeah. You know, coming here and talking to me, you're just talking to a bunch of AI obsessives. So maybe there's like a talent angle, but like what... I've also seen you on CNBC. How do you think about the public comms that you're doing?