Dream Machines: Physical AI, Robotics, Being Human
00:21:32 · Charles Wong, Aravind Kandiah
[00:21:32] S2: I I remember, like, you know, like, in Newtonomy, I was working on, like, feed management as well as perception systems. Right? And so when me and Aravind came together, like, the the key thesis we're trying to figure out was what is the most scalable path to get AI into the real world. Right? What is the enabling... The the best enabling tool or technology that that we could work on? Right? I think simulations and data very much came to the fore for us. And coming from the industry, it just felt so obvious to us at the time. Like, why wouldn't you tap on an unlimited source of data or, you know, a real world, you know, stimuli for robots to learn from. Right? This makes more sense. And, you know, the funny thing is that when we when we actually got together and started trying to find the company, tried to raise our first round of pre seed, that was that was a challenge. Right? Mhmm. We were going to all these VCs and trying to explain a, what AI was, because nobody really understood what it what it meant at the time. Right? Everybody was just thinking Terminator. Like, you trying to build Terminator? What's going on? [00:22:32] S0: Right? And
00:22:32 · Aravind Kandiah, Charles Wong, Evan Helda
[00:22:32] S0: Right? And [00:22:33] S2: then eventually, we moved over to we've been at the pitch evolved into like, we're actually working on autonomous systems and robotics. Right? And then when investors heard the word robotics, they were like, oh, robotics, terrible ROI, not investable, economics don't make sense. Right? And to add, you know, fuel to the, you know, to add, you know, insult to injury, we we told them we're not building robots. We're building dev tools for robots. Right? Which is like, how many layers of abstraction do we have to go before we become totally uninvestible? But, you know, we had a bit of a break, me and Aravind. So we did... We got connected to folks at NASA JPL pretty early in our journey. Right? And this was... We got connected to... Which is a warm intro just to one of the planetary scientists working on the Mars rover. Right? [00:23:26] S1: Cool. [00:23:27] S0: And I [00:23:27] S2: still remember it. Like, this was in the middle of COVID. Like, we like, we were staying at our parents' houses. Right? Like, we were in Singapore at the time because we were all locked down. Right? Mhmm. And we were like, man, there's a... We we we when we got the meeting, we we over prepared. Right? We were [00:23:45] S0: like, okay. [00:23:46] S2: One... We have this one shot. We have to nail it, and Aravind went away and created these beautiful, like, lunar terrain simulations and everything. [00:23:53] S1: That's awesome.
00:23:53 · Evan Helda, Charles Wong
[00:23:53] S1: That's awesome. [00:23:54] S2: Then when we got there, you know, we did the we did the call. He was so impressed. He then said, I'm gonna organize a seminar for all my fellow planetary scientists to come in and listen to you guys. [00:24:04] S1: Wow. That's cool. [00:24:05] S2: They're like, well, shit. Shit just hit them. [00:24:08] S1: Talk about a tail, as they say. [00:24:10] S2: Exactly. We're still faking it till we make it. Right? And so we got in the room there and, like, I like, this was like, you know, 3AM in Singapore. Right? We were pitching these NASA scientists, you know, in our parents' homes. Right? And I remember like, at some point during the the meeting, like, my mom, like, came into the room and wanted to give me, like, a drink. And I was like, no, mom. I'm on the I'm on the phone with NASA right now. Casual. What did come out of that was that we... NASA became our very first pilot project. Right? And it was for validating martian landing systems. So, like, the train you see behind me is from from one of those, you know, like, terrain generations. Right? Nice. Essentially, that the new the new Mars helicopter at the time, they were trying to validate, you know, where it's safe to land. Right? You know, regular debris, all that stuff. And they used our data to validate those systems. And I think that gave us a lot of street cred when later on we talked to more customers and, you know, the pitch ball from there.
00:24:10 · Charles Wong, Evan Helda
[00:24:10] S2: Exactly. We're still faking it till we make it. Right? And so we got in the room there and, like, I like, this was like, you know, 3AM in Singapore. Right? We were pitching these NASA scientists, you know, in our parents' homes. Right? And I remember like, at some point during the the meeting, like, my mom, like, came into the room and wanted to give me, like, a drink. And I was like, no, mom. I'm on the I'm on the phone with NASA right now. Casual. What did come out of that was that we... NASA became our very first pilot project. Right? And it was for validating martian landing systems. So, like, the train you see behind me is from from one of those, you know, like, terrain generations. Right? Nice. Essentially, that the new the new Mars helicopter at the time, they were trying to validate, you know, where it's safe to land. Right? You know, regular debris, all that stuff. And they used our data to validate those systems. And I think that gave us a lot of street cred when later on we talked to more customers and, you know, the pitch ball from there. [00:25:10] S1: That's amazing. So you have this Lighthouse customer. What did... So how did you go from creating some data and something in a very custom way? What came out of that? And then what was required to turn it into a product? And, like, what what does the actual product look like today? And from, like, a user experience perspective for someone who's just getting started using using Bifrost?
00:25:10 · Evan Helda, Charles Wong, Aravind Kandiah
[00:25:10] S1: That's amazing. So you have this Lighthouse customer. What did... So how did you go from creating some data and something in a very custom way? What came out of that? And then what was required to turn it into a product? And, like, what what does the actual product look like today? And from, like, a user experience perspective for someone who's just getting started using using Bifrost? [00:25:38] S2: Yeah. I think, you know, the the thing that people underestimate significantly about simulations in the data is how much setup is required. Right? Mhmm. Like, people think you can just download Omniverse or download Unreal Engine, chuck stuff in, and boom, you got simulation. Right? But, no, there's so much support required. Right? There's three d assets and environments. There's rendering pipelines. There's scaling up. There's compute as well if you wanna support such a heavy workload across such a large dataset. [00:26:04] S1: Mhmm. [00:26:05] S2: All of that has to come come into an experience that, you know, an AI developer can use. Right? Yep. Like, an AI developer isn't gonna wade into the Unreal Engine docs and try and figure it out. Right? They have... It's a totally different, you know, vernacular. They just... You know, their time should be spent building robots. It shouldn't be spent figuring out quaternions and materials and textures. [00:26:27] S0: Right?
00:26:27 · Aravind Kandiah, Charles Wong
[00:26:27] S0: Right? [00:26:27] S2: So a lot of our work was how do we abstract that away, present it in a way that an AI developer can understand. So today, our product is a web app. Right? It's a it's a platform called Stardust, where on one side is it's like a three d preview of what you're building being streamed into you on your browser, so you don't need compute. And on the right is a Python notebook. Right? Where you can... That's how you spec your scenarios and build your generations and all that whole. Right? So the idea here is that if you're a company and you wanna adopt simulation in your company, you're... Instead of spinning up a 30, 50 man simulation team, you know, with, like, three d artists, technical artists, three d developers, just use Bifrost. It's just one person, one seat, one experience as an AI developer, and you don't need all that additional support. Right? We are currently working on a new product that's going up to market soon. Nice. Product is specifically focused on robotics evaluation. Got Where the previous product start us was more around, like, focused on perception, general autonomy. This one is hyper focused on solving what we feel now is the hardest problem, which is evaluation. Right? How as a robotics developer do I go from my initial policy next to, like, a parameter sweep across all the things that I care about, doing it in a scalable way, doing it in an easy way that I don't have to, you know, again, wade into documentation, figure out Docker containers, and stuff like that.