Gabriel Hunter Chang (2:49)

But it's really hard to build a

Jacob Goldstein (2:51)

robot that looks like a guy. The Physics are complicated. You need a lot of moving parts. Which is why, despite the amazing demos we keep seeing on social media, and despite all the smart people and all the money trying to make robots happen, almost nobody has a humanoid robot at their house. Maybe another approach would make more sense. Maybe instead of starting with the human body and saying, how can we make a robot that looks like this? We should ask, how can we build an affordable robot to help people solve real problems at home right now?

Chase for Business Advertiser (Alternate) (3:37)

Jacob.

Jacob Goldstein (3:37)

I'm Jacob Goldstein. This is what's yous Problem? And my guest today is Aaron Edzinger. Aaron has been building robots for decades. A while back, he sold a company to Google. Then he became the head of robotics at Google, and then he left to start a company called hello Robot, where he is now the CEO. Aaron's problem is this. How do you build a robot that is useful and affordable for at least some people to buy and use at home? Hello Robot hasn't really solved that problem yet, but they're working on it. They have a robot called Stretch 3. They've sold hundreds of them. And the robots are proving truly useful to some people in very difficult circumstances. Also, the robots don't look anything like metal people. They're not humanoids. In our conversation, Aaron and I talk about why home robots have barely progressed since the Roomba, what it actually takes to get a robot to feed someone, and why the physical world is so much harder for AI than language.

Gabriel Hunter Chang (4:41)

One of the interesting things to me about your approach about Stretch three, the robot that you have in the world now is frankly, the things that it's not right, it's not autonomous, it's not a humanoid. It's a robot arm that's operated by remote control. And, you know, you've been in the field for a long time. You were the director of robotics at Google. So why did you leave Google to build this?

Aaron Edzinger (5:05)

Oh, it's. Yeah. Complicated answer. It's a good question. In the end, we envisioned robots being something different than what they were becoming at that time. One part of that is just about simplicity, right? So more and more, the robots that we were building at Google were becoming more complex. And you see that with the humanoids today, there's this push towards kind of human level dexterity capability. And our premise was, you know what? You could do a lot today if you push on simplifying the design, the approach. So the whole premise with hello Robot is how do we practically deliver value with a robot to help people and forget about, should it look like a human? That's really focus on just that core question. And from that came Stretch three, which is a very minimalist robot. We like to think of it almost like robotic cubism. Right. So if you take a human form and decompose it, put it back together in some other way, if you squint, it's recognizable as a human. Maybe it can do 80% of what a human or a humanoid could do, but it's much more practical, deployable from a business perspective, but also just from a usability and also autonomy. So the robot is autonomous. We don't highlight a lot of that, but the autonomy gets easier too, when the robot gets simpler. So that's the basic premise of starting the company.

Gabriel Hunter Chang (6:39)

Yeah, I mean, it's interesting, as you were saying, that I was thinking of sort of the classic software thing is the minimum viable product. Right. Like, what's the simplest thing we can get out into the world and then learn what's going on with it, iterate, et cetera. It's like you're building the minimum viable robot.

Aaron Edzinger (6:56)

Exactly, exactly.

Gabriel Hunter Chang (6:57)

So you have built a robot and it is in the world. The current version is called Stretch three. Tell me about Stretch three.

Aaron Edzinger (7:07)

So Stretch three, third generation. It is fundamentally, think of it like a Roomba with a big mast on it. Right. It comes from that perspective. Right. And then you add to that mast a telescoping arm. You know, telescoping the way a car antenna might telescope, but turn sideways and you put on the end of that telescoping arm a hand.

Gabriel Hunter Chang (7:30)

So it's basically an arm. There's a Roomba with a vertical pole going up, what, three feet or something?

Aaron Edzinger (7:36)

About four feet?

Gabriel Hunter Chang (7:36)

Yep, four feet. And then basically a robot arm. What we think of as a robot arm on top of that pole.

Aaron Edzinger (7:42)

Exactly.

Gabriel Hunter Chang (7:42)

That can drive a robot arm around your house.

Aaron Edzinger (7:44)

Drive around and it can reach up to the countertop underneath the couch, that type of thing. Very, very simple design. And what we saw early on is just by remote controlling it, you can do quite a few things around a house that you care about. And so our first launch video showed it picking up toys. Teller operated, but picking up toys, wiping down countertops, putting laundry in a laundry machine with a very simple device. You know, I just come from Google where we're building a very complex device and to do the same type of thing. And I just thought, you know, if these robots are ever really going to get out to the world as real products, it's going to be and look and feel more like this.

Gabriel Hunter Chang (8:25)

How much does it cost?

Aaron Edzinger (8:26)

Right. Now it's about $25,000 and that's because we run the company off of sales. We're not trying to, it's trying to, you know, it's low volume.

Gabriel Hunter Chang (8:37)

So does that mean your marginal cost is about $25,000? Are you selling it for about what it.

Aaron Edzinger (8:42)

No. So the opposite. So what I'm saying is it's actually can be very low cost but because we're funding R and D and engineering and production, so you have a margin

Gabriel Hunter Chang (8:52)

on it, your marginal cost.

Aaron Edzinger (8:53)

We have a real operational, sustainable business selling robots today.

Gabriel Hunter Chang (8:56)

So you're a real company selling real

Aaron Edzinger (8:59)

company selling robots to people.

Gabriel Hunter Chang (9:00)

Amazing.

Aaron Edzinger (9:01)

Yes. And that was part of the strategy from the get go is, you know, I've been in robotics for 25 years now and seen a lot of companies over promise what's possible with the technology and then their capital, their investors get impatient and they end up closing up shop. And we see this time and time

Gabriel Hunter Chang (9:24)

again we're selling to Google on a happier version.

Aaron Edzinger (9:26)

Yeah, well you could do that too which. Yeah, but really we wanted to have a viable business, build a product that people liked and loved and then from there then you have permission to take the time that it needs to really to do it correctly.

Gabriel Hunter Chang (9:43)

How many have you sold?

Aaron Edzinger (9:45)

Ish. We have about 300 robots in the world across 23 countries and about 45 different institutions from top tier research labs to small startups, to people looking at vertical farming, to people looking at healthcare, a whole range of things.

Gabriel Hunter Chang (10:03)

I saw a video about a man named Henry Evans who uses one of your robots and I found it quite compelling. Tell me about Henry Evans.

Aaron Edzinger (10:17)

Right. Henry and his care partner and wife Jane. They're in California, not too far from us. They've collaborated with my co founder Charlie Kemp for at least I think 15 years now. So it's been a long relationship. Henry is quadriplegic from a stroke. Jane is his primary caregiver. But Henry is an early adopter. He was really excited to find ways that technology can give him agency. Not just that, but make life easier on chain. He's been looking at robotics as a solution, partial solution for that for quite some time. We've collaborated with him and really when we founded hello Robot, you know, he was one of the key, well I'd say inspirations and end users that we had in mind. So through some work that was funded by the nih, we spent the last three years working with him and some other people in similar situations, looking at how stretch can benefit him. And so Henry isn't able to communicate, to talk, except through eye gaze. So he uses eye gaze on a letterboard to spell out words, and then he can also use eye gaze to control a computer mouse. And through that, he's able to control Stretch to do simple tasks, things that might seem trivial or not so exciting to other people, but for him, it's really life changing. And so a really nice story with Henry is we came in thinking, well, you know, Stretch will do very practical tasks very quickly. What we learned is there's a whole range of things that are much more social, emotional about connection that we just didn't anticipate with a device like this. And so one example is his granddaughter at the time was probably about three or four. She was a little bit scared of him because, you know, he's immobile, he's in a power chair in bed, can't talk. And so they didn't really have a relationship. And what we found is through Stretch, because he could control Stretch, he could start to play with her, interact with her in the physical world. She would decorate Stretch, put stickers on it, they would have a tablet on it, so, you know, his face would appear. And soon they started to play. So they'd play basketball, pick up a little Nerf basketball, play different games. And, you know, she started to really kind of see him in a different light and form her relationship with him. And so now she calls him Papa Wheelie because he's in his chair, and she, you know, can be excited to see him. So it really bridged that gap between the two of them. And it's the type of thing that, with robotics that has so much potential. And I think we really don't quite know how far we can go on the sort of the social side of it, really. This can be this empathetic technology that brings so much to people, and that's kind of what we're excited to explore.

Gabriel Hunter Chang (13:28)

As you were talking about the robot playing with a very young child, the question of safety crossed my mind. Like, what do you have to worry about in that setting? How do you deal with it? Like, how much does the robot weigh?

Aaron Edzinger (13:42)

Yeah, it's a great question. So this robot weighs about 50 pounds, a little over 50 pounds. Most of that weight's down in the base, Right. And really, it's an important question because there's so much happening in robotics right now where we're seeing humanoids in the home doing dishes, right? And these are scary machines. And really the fundamental thing you have to think about is just the physics of it, right? When you have a humanoid that's balancing particularly a human scale, one, the amount of weight that's up high, the potential energy that's there is significant. And when things go wrong, if it

Gabriel Hunter Chang (14:23)

falls over, it could smush things. That's fundamentally what you're talking about.

Aaron Edzinger (14:27)

That's fundamentally it. And so just the pure physics of that are fundamentally, we can't cheat it. Right? You can't and things will fail. So really from the get go for hello, robot and stretch, we thought about safety from just the physical properties of it. How do we make it so it's very hard for this device to cause harm. So first off, it should be lightweight. That potential energy, the mass should be down low so that if it falls over, it's not going to smash and break something it. And then you think about the robot

Gabriel Hunter Chang (15:04)

joints themselves and is there a pincer? We haven't actually talked about the sort of gripper. Is it like a two piece pincer?

Aaron Edzinger (15:12)

It's basically like a grabber that you'd buy on Amazon, you know, like to pick up trash. Yeah, it's one of those that we've attached a motor to essentially. But if you think about, you know, the robot kind of motors, if you have to hold your arm out against gravity all day long, you know you're gonna need a big motor there. Right. Cause that's a lot of energy to support that. If you don't, if you can design the robot in a minimal way where you're not spending all of your time fighting energy or first off, you're more energy efficient. Second, you have a smaller motor which will be less dangerous. And so the arm moves in a way. It's designed in a way that it's enough to fight gravity and that just makes it inherently safer. Right. And so when you're working around older adults, people with disabilities, kids, pets, these things really, really matter. And so that's, that's just the starting point from the design. The intrinsic safety has to be there, the physics of it has to be. Right.

Gabriel Hunter Chang (16:12)

So people with quadriplegia, people with extremely limited mobility are an obvious early use case. What else, like, what else are people using this for? What else do you have in mind for early use cases?

Aaron Edzinger (16:26)

Yeah, one of the most interesting use cases is in the home around caregiving. It's a very challenging one, it's further out. And so, you know, there's a lot of use cases nearer term which I think will be in service, you know, backup retail logistics. And we're seeing data centers, a lot of stuff like that, that we are seeing interest in and people exploring, like picking and pecking.

Gabriel Hunter Chang (16:51)

I mean, there are specialized robots for those things, right?

Aaron Edzinger (16:54)

Well, yeah, there's some places where there's not yet. Right. So for example, micro fulfillment, little fulfillment, you know, like even like a, you know, stocking of 7 11. There's lots of things where a robot like this could play a role. There are other options.

Gabriel Hunter Chang (17:09)

And just to be clear, my understanding is that your robot is mainly not autonomous, that it's mainly remote controlled, tele operated. Is that right? And what does that mean for that use case?

Aaron Edzinger (17:20)

Right, it can be autonomous. There are autonomous. It's completely open source. There's a whole ecosystem of open source software people have built for it, including autonomy. And the generation of robot we're working on now will really be autonomy powered. Right. So in terms of the sensing the software, but back to that sort of really minimal approach. We're trying to say, what can we do today that has value? And for someone like Henry Evans, he controls the robot, it has autonomy mixed in there. So he's not controlling every motion necessarily. And so what we see is a steady progression from direct teleoperation to mixed autonomy to full autonomy, but doing that really deployed. Right. So this is not something in a research lab, it's actually in the product. And you know, we also see a path from people with really severe mobility impairments like, like Henry. They're interesting because they have a high motivation to use the device and spend the time to control it. Right. They've got time and motivation. And then you can layer in the autonomy over time. And as you do that now you can broaden the set of people who can benefit from it. I think from people like Henry, the next step in terms of, in home care is remote operation by a family member, a care provider. For example, I'm in New Orleans right now, my mother is back in California. And I'd love to hop on and check in. And of course you can zoom, but sometimes you need to go into the other room, open the cabinet door and you know, find the medication or something like that. And, and so again, there's a human in the loop. But they're not going to be controlling every motion. They're going to be clicking on, on a map and saying, go to the bedroom, open the, open the drawer. That type of thing.

Gabriel Hunter Chang (19:23)

Yeah.

Aaron Edzinger (19:23)

And. And then from there, of course you can bring in more autonomy. And suddenly you have in home autonomous care and eventually in home, robots that are doing your laundry and dishes and everything that, you know, people are excited about. But for us, again, we're taking this very pragmatic approach about how do we stair step, how do we kind of climb our way up to that. That future

Gabriel Hunter Chang (19:47)

stair step is an interesting choice of words for a robot on wheels. But, you know, one of the things people always talk about in the context of robots in the last few years is getting the data to train the model for autonomy.

Ryan Reynolds (20:04)

Right.

Gabriel Hunter Chang (20:05)

And the obvious analog is large language. Models got so good because the Internet is full of text. Right. And there is no Internet of the physical world. There is no giant data set that you can just train a model on so that robots can understand how physics works in the real world. Right. How are you dealing with that?

Aaron Edzinger (20:27)

You know, I would say in terms of AI, physical AI, as it's called around robotics, we're taking, you know, an approach that's really about taking things that are mature and integrating them into our autonomy versus what's called end to end learning. Where, you know, you might collect a huge data set. Right. You might train the robot, teleoperate it to open the dishwasher a thousand times, collect that data and train a deep, learned model to do that task. And you see very exciting results from that. And I do think that will continue to progress, but in terms of being ready to be deployed, it'll take some time and the amount of data that's required beyond just the dishwasher. Well, now it's a different dishwasher. Now it's a dishwasher in different lighting conditions. Oh, there's something in front of it. That complexity gets so large so quickly

Gabriel Hunter Chang (21:23)

and it doesn't generalize. Like there are not models where you can teach it one dishwasher and it can make inferences about other dishwashers.

Aaron Edzinger (21:30)

There's inklings of that happening. Right. But we're not there yet. And researchers are still really exploring what's possible. You know, some are saying it's going to learn from YouTube, some are saying it's going to be all in simulation. Some are hiring teams of operators to collect that data. Right. Our view is, you know, that world is progressing quickly and it'll sort itself out. Soon we will be able to train the robot to open that dishwasher. I think what's more interesting is how do you get it to do the five or six important tasks you care about and actually do it in a robot that people feel is trusted in the home, is safe in the home, it's is affordable, you know, all the things that actually they have a business.

Jacob Goldstein (22:17)

Yeah.

Gabriel Hunter Chang (22:17)

So you're basically, it's again the, like the physical intelligence is the bright shiny object that you are not chasing so that you can build a robot that works.

Aaron Edzinger (22:26)

Exactly. And, and you know, think about, you know, you know, is Apple building a big AI team? Well, no, it's partnering with people who are doing the AI and I think, you know, for us, we're, we're more on that side of the fence.

Gabriel Hunter Chang (22:41)

Are some of those AI gains that you are sort of piggybacking on coming from autonomous vehicles? I mean, when you talk about saying go to the other room and have it go there, I think of that's where my mind goes. Right. Those are the robots out in the world right now, driving around San Francisco.

Aaron Edzinger (22:58)

Yeah, I mean, it's slightly different in terms of the requirements, the sensors, but it's certainly related. I mean, we've all benefited from these rapid gains in sensing and compute and you know, even LLMs is they're being used to help the robots plan and think and replan. Right. So, you know, when I started in robotics 20 years ago, I couldn't imagine being able to say to the robot, hey, go to the other room and find the blue mug that's next to the microwave. Right. Like the context of that was just like, how are we going to solve that? Are we going to write every rule

Gabriel Hunter Chang (23:36)

into the robot and that's like commodified like those are, that's, that's the easy part now.

Aaron Edzinger (23:42)

At some level, yes. I mean, certainly there's more to be done. I'd say, you know, the real challenges today are going to be around dexterous manipulation and being able to do things that are really in dynamic, you know, kind of home type environments.

Gabriel Hunter Chang (23:59)

So when you, when you talk about, you know, dexterous sort of fine motor skill type things, what are you thinking of? What do you want the robot to be able to do in the relatively near term that it can't do now?

Aaron Edzinger (24:11)

I think just broaden the range of tasks. Right. So one thing that's really exciting to this user group is feeding. Right. I think feeding, feeding is one of the things where right now someone, your partner, care partner has to come and feed you. And there's a number of reasons why that's just they would prefer to do it themselves.

Gabriel Hunter Chang (24:29)

Of course, the basic, it's a basic function of adult life is you feed yourself.

Aaron Edzinger (24:35)

You feed yourself and depending on someone changes that relationship with that person. Right?

Jacob Goldstein (24:40)

Yeah.

Aaron Edzinger (24:41)

And so feeding is one that we've looked at a fair amount. It is challenging. There are safety considerations. So we're careful about what utensils and what type of food, but I think if it can feed, if you can feed yourself, if you can get yourself a glass of water, you can do basic, you know, hygiene maintenance things for yourself. I think that's kind of the baseline of regaining a sense of agency. Right.

Gabriel Hunter Chang (25:06)

Yeah. And what do you have to figure out to make feeding really work that you haven't figured out yet?

Aaron Edzinger (25:12)

Yeah, it's a mixture of Food is kind of messy, mushy. Right. It's not like a rigid object.

Jacob Goldstein (25:19)

Well.

Gabriel Hunter Chang (25:20)

And it's super heterogeneous.

Chase for Business Advertiser (25:22)

Right.

Gabriel Hunter Chang (25:23)

Food is many different kinds of ways.

Aaron Edzinger (25:25)

Exactly. And so. But there is, you know, one thing we've seen is, you know, the caregiver is willing to prepare the food, maybe put it in a special tray. Certain types of food that we agree it can work with. So. But of course, you'd love that to get broader. But, you know, food in itself is challenging. Then there's the safety challenges. But also, you know, a really interesting part of this which kind of relates to, you know, where AI is in general, is understanding the person. Right. Are they ready for another bite? You know, are they, you know, what side would they like to be fed from? There's all these little things that are about that human interaction that make the product useful, exciting, acceptable in the home. And so much of what's happening in AI right now doesn't really get at the human side of the story. And I think that's one thing where we'd love to see advancements is really having better understanding of these sort of more subtle cues and interactions that the robot will need in order to be accepted.

Jacob Goldstein (26:35)

We'll be back in just a minute. I'm not like a coffee connoisseur, but recently I tried coffee from a company called Perk, and I loved it. And I'm not just saying that because this is an ad for Perk, though. This is an ad for Perk. I. I really did think the coffee was delicious. Perc has lots of different coffees to choose from, and they color code their bags. The blue bags are mild coffee. The pink bags are wild coffee. You can find the coffee that matches your vibe and get 15% off your next order with promo code problem@perkcoffee.com that's P E R C coffee.com promo code problem.

Amazon Health AI Advertiser (27:25)

Amazon Health AI presents painful thoughts.

Amazon Health AI User (27:30)

I. I can't stop scratching my downtown.

Aaron Edzinger (27:33)

Mm.

Amazon Health AI User (27:34)

Yeah, but I'm not itching to go downtown and tell a receptionist I'm here to talk about my downtown Some things you'd rather type than say out loud.

Amazon Health AI Advertiser (27:45)

There's no question too embarrassing For Amazon Health AI chat your symptoms and get virtual care 24. 7 Healthcare just got less painful

Aaron Edzinger (27:55)

Being

Chase for Business Advertiser (Alternate) (27:55)

a small business owner isn't just a career, it's a calling. Chase for Business knows how much heart and effort go into building something of your own. That's why they make your business growth their priority. The team at Chase takes the time to understand your mission, where you are now and where you want to go. Their broad range of solutions is designed with you in mind so you can bring your ideas to life. From banking to payment acceptance to credit cards, you can. You can conveniently manage all your business finances all in one place with their digital tools looking for tips and advice. Their online resources are always available to give you the solutions you need to help your business thrive. See how your business can get stronger and go farther with Chase for Business. Learn more@chase.com business chase for business make more of what's yours the Chase Mobile app is available for select mobile devices. Message and data rates. May apply JP Morgan, Chase bank and a member FDIC Copyright 2026 JPMorgan Chase Co.

Gabriel Hunter Chang (29:01)

So I'm curious because you have been working on robots for more than 20 years, right? I mean, so much has happened in software.

Aaron Edzinger (29:10)

Yeah.

Gabriel Hunter Chang (29:11)

Right. And even in other kinds of hardware. Like you were working on robots before there was an iPhone.

Chase for Business Advertiser (29:16)

Right.

Gabriel Hunter Chang (29:17)

And so, so much has changed and it feels like, at least as a, as a layperson, certainly in the home, like robots are not around the way iPhones are around. And even now large language models are around. And so has less happened with robots than you thought?

Aaron Edzinger (29:33)

I would say less in terms of deployed robots. Right. Like, the Roomba is still one of the most successful robots, but there aren't many other examples like it. Right. And so why, why is that?

Gabriel Hunter Chang (29:45)

Yes, like, that's a great way to ask it. And I mean, you worked with Rodney Brooks, right, One of the Roomba guys at mit, and.

Aaron Edzinger (29:52)

Yep.

Gabriel Hunter Chang (29:53)

That's like the only robot anyone can name. And there are lots of failed companies that were Roomba like, right?

Aaron Edzinger (29:59)

Yeah. Yeah. And so it was. Yeah. So I did my PhD with Rod Brooks at MIT and at the time he was spinning up iRobot, inventing the Roomba. And around the time I finished, he had just sold his millionth Roomba and he inspired me to go off and do my own company.

Gabriel Hunter Chang (30:16)

I mean, it must have seemed like they're here. Home robots are here. You would think at that moment it's like, okay, here we go.

Aaron Edzinger (30:23)

Yeah, yeah, well. But things seemed very hard, right? So it's like, okay, a vacuum cleaner that bounces around off walls. Like, we can do that. Right?

Gabriel Hunter Chang (30:33)

Yeah.

Aaron Edzinger (30:34)

And I think the insight from Roomba that I still carry is, you know, they thought really carefully about the price, the complexity of the device. They weren't trying to. There was a competitor called the Electrolux at the time that was about $3,000, had all sorts of cool features in computer and sensing. But with a Roomba, what they did is they went to the mall and they asked people, how much can you spend on something without having to ask your significant other for permission?

Jacob Goldstein (31:03)

Oh.

Gabriel Hunter Chang (31:03)

Huh.

Aaron Edzinger (31:04)

And the answer was about $200.

Gabriel Hunter Chang (31:06)

I was just talking to a friend who basically does a B2B business that's business to business. And apparently that's a sales strategy in business as well. Like, the person who you're selling to, they have some spending authority that they can spend on their own. Of course, it's about a thousand times greater. Right. It's like they can spend $100,000 or whatever without asking somebody else. So you really want to price it like that, otherwise it's much harder. So anyways.

Aaron Edzinger (31:31)

Yeah, exactly. But, you know, I think the Roomba success in part was they made a minimal product within that spending authority and kind of made it all work. And I think the reason we haven't seen a lot beyond that is it's been hard traditionally to find the things that match all those constraints. Technology can be there, the use case can be there. Right. Clearly, in home care, there's so much need for help in the home, but can we do it at a price? And is there technology there? You know, $5,000 would be a lot of money for most families.

Gabriel Hunter Chang (32:07)

Yes.

Jacob Goldstein (32:07)

Although for a caregiver.

Gabriel Hunter Chang (32:09)

Right. Like, if you narrow it not to, like doing the dishes, but, I mean, you know, many people have been bankrupted by just needing to hire a simple caregiver.

Jacob Goldstein (32:22)

Right.

Gabriel Hunter Chang (32:22)

Like, people could definitely figure out how to spend $5,000 on a robot that could significantly replace a caregiver. One would think. I guess that's your bet. That's your business.

Aaron Edzinger (32:33)

That's partially our bet. Yeah. And I think it's not replacing, it's augmenting.

Gabriel Hunter Chang (32:39)

Yes.

Aaron Edzinger (32:39)

Right.

Gabriel Hunter Chang (32:40)

That makes sense.

Aaron Edzinger (32:40)

Yes. Because we're not going to do everything, and nor should we, but there just aren't enough people. Hours, Caregiving hours in the world today to take care of everyone who needs care. That's just the truth of it.

Gabriel Hunter Chang (32:52)

Yeah. Zooming out from. From. Hello, robot. From your company to the field more generally. Like, what's the landscape look to you? How does the landscape look to you right now?

Aaron Edzinger (33:02)

Yeah, it's fascinating. There's a lot of, first off, a lot of capital flowing into it, into two areas, really. One is robot foundation models. This is sort of deep learning the LLMs of robots. And then the other would be humanoids. Right. And sometimes it's both. Right. But there's a lot of excitement around that. And I would say generally they both have a little ways to go before they get deployed and really applied.

Gabriel Hunter Chang (33:28)

I mean, the foundation model, one, you know, an LLM for robots, sounds good to me. It seems interesting to me. Tell me about what's happening there.

Aaron Edzinger (33:38)

Yeah, well, it's. First off, it's remarkable what you see the robots doing with these models, Right. So peeling an egg or tying shoelaces or folding laundry. And they're getting better, right? They're having longer.

Gabriel Hunter Chang (33:54)

I think the peeling an egg was teleoperated by.

Aaron Edzinger (33:56)

Required by. Fair enough. But they are doing things like that that you think, wow, I just could not imagine this. And they're getting better in that they're having longer memory, so they can do stuff over a longer period of time. They're getting better at recovering from issues, but also they haven't yet learned to generalize. Right. So if you think about what they're learning, as we were saying, such richness, diversity in the environments they're meant to operate in, they're learning on one kind of little island and a whole sea of data that they need access to to really be generally useful. And so that's kind of a real challenge that that field is trying to work through is how do you get to that generality and that robustness to that diversity? And it's very much an unsolved problem. I think they're making progress. But it also means to me that this is not something that we're going to be deploying into a robot product anytime soon.

Gabriel Hunter Chang (34:55)

You know, it's interesting because it seems like one of the, frankly, big surprises about large language models is how good they are at generalizing.

Jacob Goldstein (35:04)

Right.

Gabriel Hunter Chang (35:05)

How, how, how they often are much better than specialist models. Right. Just take a frontier model and throw anything at it and it'll be really good. Not quite, but kind of.

Aaron Edzinger (35:15)

Yeah.

Gabriel Hunter Chang (35:15)

And. And so it's interesting to hear that that's not at all the case with robot foundation models. Why do you suppose that is?

Aaron Edzinger (35:24)

Well, you know, I think one thing is just the amount of data that's available. Right. So if you think of a LLM, one simple way to think of it, it's going to predict the next word that you say. And it has the whole corpus of the Internet to make that prediction. Right. For a robot to predict the next thing that's going to happen, when it's peeling the egg, there's not a lot of data out there for it to understand that. And some people are saying, well, I can do that in simulation, but it's very hard to imagine simulating a lot of that kind of complexity that's actually in the world.

Gabriel Hunter Chang (36:00)

Just because the world is so diverse. What do you even simulate or how

Aaron Edzinger (36:03)

do you even simulate an egg being peeled? And so I think it's a challenge because robots are in the physical world.

Gabriel Hunter Chang (36:13)

Is there something, as I understand it, the basic model, the transformer model that powers all of the frontier LLMs today was developed based on the nature of language itself. Right. Like, is there a thought that that's maybe just not the best fundamental model for understanding the physical world, for training robots? Might it be the case that some very different approach ends up working much better?

Aaron Edzinger (36:41)

Yeah, I know less about that. I do know there is a large interest in world models today that can kind of infer the physics of the world. And I think one thing that is beneficial is there is so much structure in the world, both from physics.

Gabriel Hunter Chang (36:58)

Like laws of physics.

Aaron Edzinger (36:59)

Like laws of physics, yep. But also in the way that we design our homes, our appliances, our. You know, there is a lot of structure to be gained and learned. And so you can imagine these models become very good at predicting what your home is going to look like because it's seen a lot of homes and they don't vary that much fundamentally. So there is enough structure to learn from what that model and how that transpires, I don't know, but people are exploring that.

Gabriel Hunter Chang (37:29)

I mean, one of the. One of the obvious challenges is we're just making the robots come live in our world. Right. And I could imagine for a case like someone with quadriplegia, it might. I could imagine a universe where it's like, well, if you set up your room this way, not yet you have to remodel your house, but you have to do some set of things, then our robot can do a lot. And I can imagine that being worthwhile in many settings. Is that something you are doing or thinking about, sort of a compromise solution like that?

Aaron Edzinger (37:57)

Absolutely. And you know, as an example, Stretch three, as it is today, has a hard time opening some refrigerators. Right. It's just the magnetic force is too high. But you can attach a little, you know, grabbing thing that breaks the seal and makes it easy to open it. Right. And so it's a $3 part. You stick it onto your fridge. Now Stretch can open their fridge for you. And for, particularly for these users, you know, those accommodations make a lot of sense. And you know, I do think, you know, maybe over time you'll have special plates that the robot knows it cannot, you know, grab and not break. And you know, there will be accommodations that people make for it. But you know, for now we are using and counting on those types of accommodations.

Gabriel Hunter Chang (38:46)

So I want to do two kind of forward looking things. First, the sad one and the happy one. Basically the first is how might it not work out? And then the second is what does it look like if it does work out for hello robot. So like what might go wrong for, for your company in a, you know, significant way?

Aaron Edzinger (39:07)

Right. Well, there's always the risk of just the business doesn't work out. I think, I think from, from the technology and product side, it's really about acceptance, right. So I think implicit to bringing this technology into your life, into your home, there has to be trust, right? And people have to trust and want this technology around them. And so I think, you know, there, there's always a risk that people are like, you know what, we're gonna, we're gonna find different ways to address the caregiving problem, you know, which, you know, there's lots of, gonna be lots of solutions to that. But robots may, may not turn out to be part of that story. We're bullish that it will, but there, it's a real kind of complex dynamic between the people and the technology. So that's sort of the bear case. The bull case would be, you know, 10 years out. Call it a $5,000 device that's in your home that can provide caregiving support in the times when there's not care present. So basically it's just kind of facilitating quality of life at home, agency at home as you want to kind of age independently without having. So when you're, when your grandkids come over, you're not, they're not doing chores for you, you're interacting. Right. Because the robots helping fill that gap.

Jacob Goldstein (40:30)

We'll be back in a minute with the lightning round. I'm not like a coffee connoisseur, but recently I tried coffee from a company called Perc and I loved it. And I'm not just saying that because this is an ad for Perc Though this is an ad for Perc, I really did think the coffee was delicious. The bag I'm drinking at the moment was grown in Peru. Perc sources coffee from all over the world. They have lots of different kinds of coffee to choose from and they color code their bags. Blue bags are more mild coffee and pink bags are more wild coffee. That Peruvian coffee I'm drinking now is wild and if you're on the fence, I recommend trying wild. The other morning I had my first sip and I thought of that Will Ferrell line in old school where he hits the beer bong and then he says once it hits your lips, it's so good. Find the coffee that matches your vibe and get 15% off your next order with promo code problem@perkcoffee.com that's P R C coffee.com promo code problem Amazon Health

Amazon Health AI Advertiser (41:43)

AI presents painful thoughts why did I

Podcast Announcer (41:48)

search the Internet for answers to my cold sore problem? Now I'm stuck down a rabbit hole filled with images of alarmingly graphic sores in various stages of ooze. I can clear my search history, but I could never unsee that.

Amazon Health AI Advertiser (42:05)

Don't go down the rabbit hole. Amazon Health AI gets you the right care fast. Healthcare just got less painful.

Chase for Business Advertiser (42:13)

Being a small business owner isn't just a career, it's a calling. Chase for Business knows how much heart and effort go into building something of your own. That's why they make your business growth their priority. The team at Chase takes the time to understand your mission, where you are now, and where you want to go. Their broad range of solutions is designed with you in mind so you can bring your ideas to life. From banking to payment acceptance to credit cards, you can conveniently manage all your business finances all in one place with their digital tools looking for tips and advice, their online resources are always available to give you the solutions you need to help your business thrive. See how your business can get stronger and go farther with Chase for Business. Learn more@chase.com business chase for business make more of what's yours the Chase mobile app is available for select mobile devices. Message and data rates may apply. JP Morgan Chase Bank NA Member FDIC Copyright 2026 JP Morgan Chase Co.

Gabriel Hunter Chang (43:13)

Let's

Jacob Goldstein (43:14)

finish with a lightning round, okay do you still make art?

Aaron Edzinger (43:18)

Oh, not for a while, but it is. How I got into robotics was doing robotic sculpture and then that led. One thing led to another. But yeah, really fascinated by love art but also love kind of robotics and how it relates to art.

Gabriel Hunter Chang (43:38)

What's the least practical robot you've ever built.

Aaron Edzinger (43:42)

Wow. So let's see. In the 1990s in San Francisco, we had a robotic art performance troupe and it was a theatrical multimedia performance of robots on stage, acting, performing. So this was a kind of robotic gallery of kind of characters from the local neighborhood. So these were street musician and someone who was kind of causing trouble. And so they're kind of robotic denizens of the tenderloin of 1990s San Francisco put onto stage.

Gabriel Hunter Chang (44:23)

What was the first robot you ever built?

Aaron Edzinger (44:26)

It was actually the robotic street preacher. So it was a little robot that would go around and big spinning head of a speaker and it would be on stage and perform and we would sit there and tele operate it. And it had, you know, no real autonomy. This was, you know, everything was hand soldered, hand built and their computers were super tiny or super small capability. And we put that on stage and performed it.

Gabriel Hunter Chang (44:55)

Did you ever try and build robots when you were a kid?

Aaron Edzinger (44:59)

No, you know, strangely, never had a fascination with robots as a kid. I came out of, grew up in Washington state on a farm. And you know, I think my interest in robots came from having to do farm work and learning about automation.

Gabriel Hunter Chang (45:16)

Like Henry Ford, you know, Henry Ford was the same story. He grew up on a farm and thought there was too much labor.

Aaron Edzinger (45:21)

Exactly. And I just thought, you know, there's gotta be a better way than this. My, my mother's family, which is Mexican, had a history in kind of, kind of working in the fields and kind of migrant labor. And so kind of the relationship to labor and, and work kind of inspired my interest in, in automation and then robotics.

Gabriel Hunter Chang (45:42)

Did you ever try and build a machine to do your chores for you?

Aaron Edzinger (45:45)

I absolutely did. It never worked, but it was fun.

Gabriel Hunter Chang (45:50)

I guess that's still what you're working on at some level.

Aaron Edzinger (45:53)

Exactly.

Gabriel Hunter Chang (45:54)

What's one thing you learned working with Frank Gehry?

Aaron Edzinger (45:57)

Oh, right. So we had a project looking at conceptual, a large robotic kind of botanical garden and this whole thing in Singapore. So we were looking at developing that with him. And you know, I'd say actually the thing that was most interesting is just he really put a lot of trust into the young people in the studio and really was generous in that way. And you know, it's been an inspiration for me.

Gabriel Hunter Chang (46:30)

That's a lovely surprise. 1. This is, I guess, some bias that I have. I imagine a famous architect as not being like that.

Aaron Edzinger (46:37)

And so to hear that he was

Gabriel Hunter Chang (46:38)

like that is so lovely.

Aaron Edzinger (46:39)

Very, very generous in a way that can be inspiring to people on their career, so I would try to carry that forward in my own work.

Gabriel Hunter Chang (46:49)

Thank you for your time.

Aaron Edzinger (46:50)

Good luck. It's been a pleasure. Thank you.

Jacob Goldstein (47:00)

Aaron Edzinger is the co founder and CEO of hello Robot. Today's show was produced by Gabriel Hunter Chang and edited by Lydia Jean Cott. Our engineer this week was Hansdale Shee. We're always looking for ideas for who to talk to and what to cover on the show. You can email us at Problemushkin FM. You can find me on X JacobGoldstein. You can find me on LinkedIn. I'm Jacob Goldstein. Thank you very much for listening to the show and we'll be back next week with another episode. I'm not like a coffee connoisseur, but recently I tried coffee from a company called Perc and I loved it. And I'm not just saying that because this is an ad for Perk though. This is an ad for Perc. I really did think the coffee was delicious. The bag I'm drinking at the moment was grown in Peru. Perc sources coffee from all over the world. They have lots of different kinds of coffee to choose from and they color code their bags. Blue bags are more mild coffee and pink bags are more wild coffee. That Peruvian coffee I'm drinking now is wild. And if you're on the fence, I recommend trying wild. The other morning I had my first sip and I thought of that Will Ferrell line in old School where he hits the beer bong and then he says once it hits your lips, it's so good. Find the coffee that matches your vibe and get 15% off your next order with promo code problem@percccoffee.com that's P E R C coffee.com promo code problem service

American Military University Advertiser (48:40)

opens doors and at American Military University it can open doors for the whole family. If you have a loved one who served in the military, you may qualify for reduced tuition. AMU offers flexible online programs designed to fit your schedule so you can keep moving forward wherever life takes you. Learn more at amu.apus edu military Open doors to the future for you and your family with the help of the American Military University.

Ryan Reynolds (49:10)

Ryan Reynolds here from IT Mobile. I don't know if you knew this, but anyone can get the same Premium Wireless for $15 a month plan that I've been enjoying. It's not just for celebrities, so do like I did and have one of your assistants assist assistance switch you to Mint Mobile today. I'm told it's super easy to do@mintmobile.com

Podcast Announcer (49:29)

Switch upfront payment of $45 for 3 month plan equivalent to $15 per month required intro rate first 3 months only, then full price plan options available, taxes and fees, extra fee, full terms@mintmobile.com this

Aaron Edzinger (49:40)

is an I heart podcast.

Podcast Announcer (49:43)

Guaranteed human.