This Week in Startups
00:30:40 · Oliver Cameron, Jason Calacanis, Lon Harris
[00:30:40] S2: I mean, our focus is primarily on these as environments for agents to learn in. And so, really, one thing that strikes me as really interesting to pursue is this idea that we humans exist inside this world model and all of these systems that we are raised in shape us, shape our intelligence, shape what we do and how we do it. And language models as they're currently trained, I don't feel are faithful to that idea. They sort of short circuit a few of these and have, versions that I'm not quite convinced by. And so what I think is really exciting is this idea that you could place agents, both mature or nascent agents, inside these environments that to the agent are utterly real and actually place it with other agents too and say, hey. Learn the world. Learn how the world works. Try to accomplish tasks. So [00:31:39] S0: Percentage you're convinced that we're in a simulation right now. If you're being honest, like, and you've had a range of this from, you know, over time. And let's just assume 0% of us before seeing The Matrix or having read some of the philosophers who the books were based on. If everybody starts at zero, where are you at now in terms of this reality we're in? Is the base reality or we're in a simulation. Base reality would be we're the ones running the simulation versus [00:32:10] S1: Core core reality. The matter you
00:32:10 · Lon Harris, Jason Calacanis, Oliver Cameron
[00:32:10] S1: Core core reality. The matter you [00:32:13] S0: Core reality. So where are you at, Oliver? Be honest here. You're not gonna lose investment or anything. People don't need to know where people are building world models sit. How... Where where do you sit? Be honest. [00:32:23] S2: I'm 90% certain that this is a simulation, a very advanced simulation of some sort. [00:32:30] S0: This is why I'm the world's greatest moderator, and I can convince people to actually answer the question. I am fifty fifty. I'm fifty fifty at this point. Where are you, Lon? [00:32:38] S1: I used to be a lot more skeptical. I sort of thought this was like, people read too much sci fi, and then they get, like, wacky ideas. Because it's always like, well, it'd be like, well, what's the point? You know? Like, that that was my way in. It's like, why would an alien species do this to us? Like, why would you run this all day? What would be the point? But now that we are humans building them and putting animal brains in them, it's like, oh, the point would be to see if you could do it like that. It it unlocked that for me. So now I'm a lot more open to it because it'd be like, yeah, if you have the technology, someone would be like, well, let me put 6,000,000,000 human simulated brains in here and see what happens. Like, maybe we're gonna see what happens. I can't rule it out.
00:32:38 · Lon Harris, Jason Calacanis
[00:32:38] S1: I used to be a lot more skeptical. I sort of thought this was like, people read too much sci fi, and then they get, like, wacky ideas. Because it's always like, well, it'd be like, well, what's the point? You know? Like, that that was my way in. It's like, why would an alien species do this to us? Like, why would you run this all day? What would be the point? But now that we are humans building them and putting animal brains in them, it's like, oh, the point would be to see if you could do it like that. It it unlocked that for me. So now I'm a lot more open to it because it'd be like, yeah, if you have the technology, someone would be like, well, let me put 6,000,000,000 human simulated brains in here and see what happens. Like, maybe we're gonna see what happens. I can't rule it out. [00:33:20] S0: I'm gonna I'm gonna show you a paper. I have a friend, Ali. We used to play cards in LA when I was there. And, know, when you play poker, and this, paper, we'll put the link in the the show in notes so you can search for it, Beyond Perception, colon, Proposing Our Reality as an ASI Alignment Simulation. [00:33:38] S1: Sure. Okay.
00:33:39 · Jason Calacanis
[00:33:39] S0: So it's proposing that our reality, Lon, is an ASI simulation, a superintelligent simulation. And the concept is like, you know, we... Our universe is a simulation designed to train and test superintelligence. In other words, somebody created this in order, and it's getting kind of meta here, to see what happens in a simulation when the participants in the simulation make superintelligence, and can they reach alignment, or will they blow everything up? So this is like, as part of the evidence in this study, the proposal would be the fidelity of the world. And and Oliver, we'll see if this dovetails with your work. So you've been listening to this show for long enough. You've been inspired. You're starting your own company. AI maybe pushed you. Superintelligence maybe pushed you to the edge, and now you're gonna jump in. You need to have a partner you can trust to help transform your idea into a real business, and I recommend our friends over at Northwest Registered Agent. Northwest has just launched free identity services. They're going to give you a free domain name, plus open source website hosting, a business email, a phone number, and all the other important assets that make a brand new startup feel like an established professional company, all with no purchase required. And this is free done right. Northwest doesn't hand over your data to any third party vendors, and they will never sell it. So whether you've got an idea you can't shake or you're already knee deep and building something amazing, start with Northwest for free. Find out more at northwestregisteredagent.com/twistdomain. Would you, as a creator of a world, a world creator, would you produce the entire world? Let's say your world is Italy and what we just saw was, you know, Venice. Would you produce a simulation of the entire tens of millions of people in Italy in every single nook and cranny under the bridge, 10 feet down in the water and a fish swimming, and you would have that entirety simulated. Or when you create a world model, is the fly's path high fidelity and everything else is dormant? The latter. And why?
00:36:06 · Oliver Cameron, Jason Calacanis, Lon Harris
[00:36:06] S2: Simply that if you want the fly's path to be the best possible performance, that's where you focus your GPU time. And so everything outside of that, unessential to be computed. And so I think we... You know, some of our work is around how do we scale up agent learning to massive scale. Right now, we can support something like, 20 agents inside a world model session simultaneously. [00:36:38] S0: Got it. [00:36:38] S2: And so we can generate 20 viewpoints, unique viewpoints, simultaneously. Now how do you scale that to a billion? How do you scale that to 8,000,000,000 to Yeah. [00:36:49] S0: If you were actually simulating eight eight billion people's reality, half of them or a third of them would be asleep at any point in time and dreaming or not dreaming. [00:36:58] S1: But, I mean, if we really are in a simulation, simulation, how do you know there really are... Like, you only know there's 8,000,000,000 people on earth because you've read it, because you've been told, like, we don't... Maybe all of us who are being simulated are in Texas and nothing outside of Texas exists. And you know, [00:37:13] S0: like Yes. When I go to Japan, it pops up Japan as a concept and a [00:37:17] S1: it because now one of the people with actual perspectives are there. There could be just 10 of us in here.
00:37:17 · Lon Harris, Jason Calacanis, Oliver Cameron
[00:37:17] S1: it because now one of the people with actual perspectives are there. There could be just 10 of us in here. [00:37:22] S0: And and that's a limitation of the tech. Now if we were to look, Lon, at what we experience as humans in evidence, we try to understand things like dark matter. We try to understand the universe. We try to get off this planet. Pretty hard to get off the planet. It's pretty hard to get to the core of the planet. It's pretty hard to get to the Mariana Trench, but we have these little interludes where we can get to the moon or we can send a probe out to the end of our solar system. We can do our solar system, but we don't get much further behind that. Then whatever else we get is part of the big bang or not the big bang, this crazy story we tell each other that we're kind of guessing on, but it's incredibly low fidelity. That is evidence that you're in a simulation and that what Oliver just described to us, like why would I do the compute of more than 20 people in the simulation? I don't have that kind of compute money. Nobody does. And we took all the compute on the planet, Oliver, we took all the computer on the planet, we gave it to you for simulation. How many people would you have in the simulation? A million? 2,000,000? Yeah. Have you done that calculation? [00:38:31] S2: Yeah. But we we haven't. No. But I I do think... So LOGO is the double slit experiment, or it's inspired by the the double slit experiment. [00:38:39] S0: Explain.