Sources with Alex Heath
00:57:14 · Tony Zhao
[00:57:14] S0: I don't think that that's how I'll put it. Is I think it's true that right now, I think robotics is still at a at a fun part that, people are debating. Right? And these days in in large language models, people don't debate about, like, whether you need training. Right? They're like, this is... Like, give me money. I'll give you the the corresponding results. Right? There's sort a fun part that people disagree that there are many fun ideas that are... If you are, like, kind... Trying to, like, combat each other with narratives, right, essentially. But given the amount of information and the concrete evidences that we saw, we think dealing with gloves, like, collect the data, transfer it into the robot is the best way. But this is not... It's a strong opinion that is, I would say, like, weekly health. As in if there is another breakthrough that we see that changes some of the assumptions, we're more than willing to, like, update our prior and figure out how to best leverage that. Right? And that can come in a lot of different flavors. Let's say, world models that people talk a lot about these days. It could be simulation. It could be, like, anything. But the piece that will never change in my opinion is, like, real high quality data. And that is going to be the concrete thing that we'll have either way that would differentiate us no matter what's the downstream algorithm is. Let it be like training the robot directly with, like, reinforcement learning, with imitation learning, or more models. I think we're largely open minded about how to use it.
00:58:45 · Ellis Hamburger, Tony Zhao
[00:58:45] S1: And in five years, do you have a sense of what market it's going to hit in first? I mean, I feel like intuitively, I would think, you know, elderly folks, like people who have the most trouble with the chores and may need it most. Is that your instinct as well? [00:59:05] S0: For the early customers, we do think there will be a period of time, especially in the beginning, that the robot will give them more trouble than it helps. Right? And in the beginning, maybe it's like you need to learn how to do things in a way that the user wants and in the... It might make mistakes. So I would say when it comes to the early customers that we'll be addressing, I don't think we'll be aiming at people who, like, almost, like, desperately or, like, have to use it. But we'll start with people that the robot is very nice to have, but, you know, just elevates their their standard of living by a lot. So I don't think we'll be going for, like, elderly care from day one, but that's eventually the goal for sure. [00:59:56] S1: What's the robo narrative at Davos, Alex?
00:59:56 · Ellis Hamburger, Alex Heath
[00:59:56] S1: What's the robo narrative at Davos, Alex? [00:59:59] S2: Oh my god. I mean, it's all like jobs. You know? It's like, this is... Yeah. This is such a different group than what we normally talk about on the podcast. But I I do think everyone's like, oh, you know, if robotics really hits and there's millions of these robots out in the world, you know, these are a bunch of policy folk. Like, how's that gonna impact hiring and GDP growth and all these things? And I don't know. Everyone in tech that's building this stuff seems to think that it's actually gonna be accelerant for everything, that it's not going to hurt economies, but it will free up more people, you know, to do things that they wanna do or just create net new jobs. So, yeah, there's a lot of, like, dueling narratives and dueling viewpoints, but I think everybody agrees that, like, robotics is at an inflection point. Right? And that's like... That seems to be the consensus here for sure. And, Tony, like, I I assume you agree with that. You wouldn't be doing this company if you weren't, you know, believing that. But, yeah, like, how do you how do you feel about the jobs piece of it? Do you do you feel like we're gonna find different kinds of jobs, more fulfilling kinds of jobs that get taken away, you know, from robots, or how does that work?
01:01:16 · Tony Zhao, Ellis Hamburger
[01:01:16] S0: Yeah. For sure. I think the the whole mission for us is to give people time back so that they can enjoy, they can spend time with their families, they can spend time with their interests. Right? And I think there's almost like a, like, not human part of doing chores that is so repetitive, that is so boring. It doesn't highlight what makes us unique as as humans. But at the same time, I think it's... Like, because of our positioning and the current type of work that we want to tackle, it's really not going to replace anyone's job at all. It's all about, like, saving people's time because they need to do it themselves. Right? We're replacing, like, software engineers' time to do, like, chores as opposed to people who really need to to feed their families, to do, like, housekeeping. I don't think we're touching their... Kind of their unique proposition at all because, again, it might be funny to say, but we as a robotics company, we do not bet against human dexterity because it's so good. And we always find place that we can complement them as opposed to replacing them because, like, from hardware, software, AI, we're not close in terms of getting, like, a human hand, let's say. That's why people are talking about it. Like, the hand is the bottleneck. Right? Yeah. [01:02:42] S1: Yeah. I wonder what simple chore do you now have the most respect for after digging into the hand dexterity or otherwise. You... It's like, oh, man. Cleaning a toilet, unbelievable skill.
01:02:42 · Ellis Hamburger, Tony Zhao, Alex Heath
[01:02:42] S1: Yeah. I wonder what simple chore do you now have the most respect for after digging into the hand dexterity or otherwise. You... It's like, oh, man. Cleaning a toilet, unbelievable skill. [01:02:56] S0: Yeah. I I I think just... Even just using, like, shirt folding as an example. Right? We can fold these shirts in in the midair. We fold it in a very dynamic way. We, like, fling it. We use our hand to, like, kind of tuck it in. And these are things that are very hard to learn for the robots. The... I think I definitely come to appreciate that more, like, as we work on these tasks. The other thing, like, as I said, for for bathroom cleaning, there are, like, harsh chemicals that we really don't want to get to the robot. Right? It might, like... They have beautiful paints on it. If it got, like, the the bleaches on it, it will not look great. And so when they are, like... I I think when they are... You need to interface with very harsh chemicals and you need to apply a high amount of force, those are still, like... I think remains very difficult for the mechanical side. [01:03:52] S1: Just had a vision of somebody walking into my apartment, me saying, sorry, guys. My cloyd is covered in shit. I gotta go wash them off. [01:04:01] S2: Oh, god. Maybe we should have met there.
01:04:01 · Alex Heath, Ellis Hamburger, Tony Zhao
[01:04:01] S2: Oh, god. Maybe we should have met there. [01:04:03] S1: Oh, Cloyde. Well, Tony, thanks for joining us, man. It's been great to reconnect after all this time. And, yeah, super fun to see what y'all are building. I mean, it just, you know, there's so many problem statements in tech that either, I mean, there's so many reasons things can be weird, whether it is someone who just comes straight from business school versus a researcher like yourself who spent, you know, a decade on this stuff or just attacking problems that just kind of don't really need to be solved quite yet. But I think we could all agree chores are annoying, so nice to have someone doing that on the show. [01:04:41] S2: For sure. Good to meet you, Tony. [01:04:43] S0: Great to see you guys. Yeah. Enjoyed it. [01:04:50] S1: That's it for this week's show. Thanks, Tony, for coming on the podcast and Alex for staying up late. I know you have very important people to be meeting with. So, thanks for giving us your time. [01:05:01] S2: No one is important as you Ellis always. Don't forget to like, subscribe everywhere you get your podcasts. We're access.show too. We have a website. You can find us in video at access pod on YouTube.