This Week in Startups
00:19:44 · Jason Calacanis
[00:19:44] S0: And somewhere in there is like the bar chart I just shared that you could show. This needs to be on the top level. It should literally be part of the interface, just showing it to you. And that every single one of your jobs should give you a little piece of feedback at the end when it finishes its job. Point 7% of your weekly total, this many tokens. Just educate us in real time, and you could turn it off, I guess. But I think the reason people didn't have this on is they didn't want to give people compute anxiety. But if we're in this dance with the providers of this technology and we want to use it, I think we just have to be transparent about it. Which for a business, you kind of want to let somebody run out and then have to upgrade, like video games or whatever. I think in this case, got to switch it since it's a business case and just say, you really based on your work, you're at the $20 level. We highly recommend you go to this $100 level. It's gonna be this much more per year and you're worth it. You know what's even harder than writing great code? Consistently recruiting people who can write good code. And if you're working on a project and you need to hire great engineers, but you don't want to go through a painful, time consuming hiring process, G2i can help. G2i has a network of more than 8,000 pre vetted software engineers, the creme de la creme, all with more than five years of coding experience. These are the elite developers you need. And these aren't people who hack some automated coding test. The team at G2i puts engineers through live technical interviews to ensure these are folks who really know how to get their hands dirty building stuff. And the best part for founders, they make the entire process fast and easy. You get a dedicated Slack channel where you communicate directly with the G2i team so that you can share exactly what you need and start getting matched with engineers today. So go right now to g2i.co/twist and get $1,500 off your first hire. That's g2i.co/twist.
00:21:41 · Lon Harris, Jason Calacanis
[00:21:41] S1: I think it it stung a little bit maybe extra because it I I think a lot of people don't think of it as an enterprise application. It absolutely is. We use Grok Bot for our business every day. But it it also had a consumer friendly vibe, and I think that's what maybe people were like, oh, I've just been enjoying my Grok Bot, and now all of a sudden I have to pay more. But, for me, I would pay... I would happily pay more. It's it's worth it. Of things, I would pay more for but don't have to, Jason. You'll notice we're wearing our Plaud pins. I don't know if you have a Plaud pin on you. [00:22:12] S0: Well, I mean, have it right here on the charging station for a [00:22:15] S1: reason. Extremely useful. It [00:22:17] S0: on here. [00:22:17] S1: Record whatever you're talking about, whatever you're meeting about, whatever you need, it's gonna store it for later and not just record a transcript, but then you can apply that transcript, use it, reformat it, however you want to just sort of keep track of everything that was said in whatever format that you wanted to take away in. [00:22:36] S0: You know what we need to figure out by the next episode? Is there a connection between Grok Bot and Plaud? Alright. Good. Do that for the next episode, Lon. Do a little extra work here. Put a little extra work on you. Like, get that set up, and and show it next time, and show us some real world examples of how you got this work. Like, maybe it's giving you to do list ideas, and it sucked it out of your transcripts.
00:22:36 · Jason Calacanis, Lon Harris
[00:22:36] S0: You know what we need to figure out by the next episode? Is there a connection between Grok Bot and Plaud? Alright. Good. Do that for the next episode, Lon. Do a little extra work here. Put a little extra work on you. Like, get that set up, and and show it next time, and show us some real world examples of how you got this work. Like, maybe it's giving you to do list ideas, and it sucked it out of your transcripts. [00:22:56] S1: Know what I'm gonna do? I'm gonna double up. I'm gonna use this for the show, and then I'll write about it on the This Week in a I newsletter. So that's it. We'll do [00:23:02] S0: for This Week in a I newsletter. Great one. [00:23:05] S1: Great one. If your work depends on conversations, interviews, meetings, or calls, or you have a Grok Bot and you want it to be able to listen in on your meetings, you could check out Plaud at plaud.ai/twist. That's blaud [00:23:17] S0: Don't wait. [00:23:18] S1: .Ai/twist. Use the code Twist, and they're gonna give you an extra 10% off on your Plaud. [00:23:23] S0: Do me a favor. Don't wait. Go buy one now. Tell them your boy, Jake Al, sent you. Okay. Let's keep going. [00:23:29] S1: We have a... We have an interview guest. Do you wanna bring in our guest? Yes. We do. [00:23:33] S0: Let's, let's, let's talk to our guests. Of course.
00:23:33 · Jason Calacanis, Lon Harris, Oliver Cameron
[00:23:33] S0: Let's, let's, let's talk to our guests. Of course. [00:23:35] S1: Let's bring in Oliver. He Oliver Cameron. He is the cofounder and CEO of Odyssey. Of course, they build world models, AI trained on video that learns how the physical world works. So things like robots, cars, drones, physical AI, they could practice inside simulated worlds instead of a real one. We also wanted to break it up because a few weeks ago, Odyssey posted a video of a simulated fruit fly brain flying around inside a world. Its AI was imagining in real time, and we we gotta dig in a little bit more into this. [00:24:05] S0: Thank you for getting Oliver on the program. How are doing, Oliver? I I saw this. I was quite taken by it. And I said to my team, bring me whoever is running this Odyssey ML. Because we're we're playing the video and they summoned you, they found you. This thing went viral. You got 6,200,000 views on it. X.com is where, like, everybody is talking about AI. Like, if you're on... I saw Bill Ackman said the other week, Lon. If you're on X, you feel like you're two weeks ahead of the news cycle because whatever's happening on X, two weeks later, the regular press gets to it, it's true. And so we're looking at what looks like, I don't know, Spain maybe, Portugal, I don't know what town this is, but there's a tiny little bug there flying around that world. Is that world a AI constructed world, or is that a video of the real world from a drone? [00:25:04] S2: So this is entirely imagined by a world model. And importantly, it's, imagined in real time. And so
00:25:04 · Oliver Cameron, Jason Calacanis, Lon Harris
[00:25:04] S2: So this is entirely imagined by a world model. And importantly, it's, imagined in real time. And so [00:25:12] S0: Imagined in real time. So it's building the simulation on the fly like a video game might. Lon, when you saw this, did you initially think that across the uncounty valley, did you think this was a real video? [00:25:22] S1: I did. From a drone? The the first... My first impression was the simulated fruit fly was the AI part, and this was just real world footage. And then you read it, and you're like, oh, no. That's just... Like, it looks like a river in a city. And... But, no. It's just a... It's just all made up. [00:25:38] S0: So the question, I guess, is why did you do this? Why did it go viral? And what does it say about simulation theory?
00:25:46 · Oliver Cameron
[00:25:46] S2: To speak a little bit about Odyssey and what we do. And again, thank you, for having me on. So we build world models, and we believe that world models are a true foundation model. What we do is we train them on vast visual observations of the world, and then they gain this really cool ability to predict the possible futures of the world. And so through this training process, what they do is they learn physics, they learn dynamics, cause and effect, all of these fundamentals of the universe. And so what is this useful for? Well, we believe that WARP models are essential to enabling a billion robots and a trillion agents. For robots, WARP models are this general brain, and it allows robots to better predict the consequences of their actions, and to thus better interact with the world. And then for agents, world models generate infinite environments for agents to learn within, and this increases their performance on tasks that they would learn inside those environments. And so what we did here is say, hey, there is this really cool, you know, fruit fly project. Someone's mapped the brain of an... You know, a species. That's incredible. Let's now put that brain inside an environment that's being generated in real time. And will the fly decide to fly? Will it just sit there? What will it do? And as you can see here, the fly flew around as a fly does. And I I think the reason it went viral, you know, is that it's this perfect perfect moment of saying, are we living in a simulation? Right? We we have increasingly capable artificial intelligence. This dream of superintelligence is no longer a dream. And I think it's on lots of people's minds now that with increasingly advanced technologies, whether all of this is in fact a simulation. And so I I think it just sparked that thought in people. It's like, wow. Okay. Now now we're starting to see real visual evidence that this might be a simulation.