# Jeff Burnstein & Devinder Sodhi: Physical AI | Intelligence at the Frontier 2026

- Channel: [Funding the Commons](https://streameth.org/funding_the_commons)
- Date: 2026-05-09
- Duration: 31:05
- Topics: Jeff Burnstein, Devinder Sodhi, Association for Advancing Automation, physical AI, robotics industry, embodied AI, automation and jobs, US robotics policy, national robotics commission, humanoid robots, AI and manufacturing, workforce training AI, US China robotics, Intelligence at the Frontier, Funding the Commons, tech for good, AI policy 2026, industrial AI, AI and labor, robot safety standards
- Watch: https://streameth.org/watch/yt-A6RpqrfHbwE
- YouTube: https://www.youtube.com/watch?v=A6RpqrfHbwE

## Description

The US invented industrial robotics. Then it didn't invest in it. Japan took the lead. Then China. Now China is the largest user of robots in the world and pushing to become the largest supplier. Jeff Burnstein has watched all of this from the inside for over 40 years.

Jeff, president of the Association for Advancing Automation, sits down with Devinder Sodhi of Frontier Tower for a conversation covering the full arc of physical AI — from why the word "robot" still scares American voters, to the case for a national robotics commission, to why Detroit's collapse was never really about the machines. Done right, he argues, automation creates more jobs than it eliminates. But only if governments invest in workforce training rather than waiting for the market to sort it out.

Speakers: Jeff Burnstein (Association for Advancing Automation), Devinder Sodhi (Frontier Tower / Tess Ventures)

This session was part of Intelligence at the Frontier, held March 14–15, 2026, at Frontier Tower in San Francisco.

Learn more at https://www.fundingthecommons.io

## Transcript

Welcome, Jeff, and thank you for uh actually coming to our last minute thing today. Sunday's a little bit lighter, but I like it that way. Uh just because I don't feel the pressure as much, but uh we still have like I think 500 people in the building, which is crazy. Um thank you so much. Um your association, uh I first heard from you actually at CES and that was fantastic. Uh we got to get him on too >> maybe now. Um if you don't mind just for the folks who might not wait who knows about Jeff >> nobody in this room >> nobody this is what an opportunity to like spread your brand uh if you wouldn't mind giving us just uh a little bit more about your association of automation and how the changes uh both to the naming uh have have gone through in like the past few years and like where the focus is today. >> Yeah. So the association uh was founded uh 1974 as a robotics association. Over the last four plus five plus decades we've added associations on machine vision, motion control and artificial intelligence. 1450 companies in 35 different countries. We're focused on educating companies on how to successfully apply these technologies. That's really the key for our association and um things have changed quite a bit in terms of the interest in robotics. The for those of you who aren't aware the robot industry started in the United States and um US was the clear leader but it didn't receive a lot of receptivity. A lot of the unions were worried about losing their jobs. Um, and the government really didn't support the robotics industry in the US. The Japanese decided they really wanted to go all in on robotics. They had government support and they became the leaders in robotics for a long time until most recently the Chinese surpassed them. And now China is the largest user of robots. they'd like to become the largest supplier. That's one of the reasons they're all in on humanoids. And so, it's been a really interesting ride for me. Um, I was an English major in college, intended to go into public relations or maybe sports writing when I graduated college. Found my way to a engineering association, society of manufacturing engineers, and that led to another job marketing um in robotics. And for me, my skill is in translating what the technical people have to say to the public and to um the policy makers and to the press. So that's how I spend my time. That's how I built my career. You learned a little bit about these technologies along the way. That's awesome. How many people know about Jeff now? Just trying to see who's paying attention. Thank you. Okay. Uh it was kind of a trick question because I think the naming change is important. Uh we've gone from just pure robotics to physical AI and embodied AI is sort of the word of on the street right now that is really popular and companies certainly put a lot of time into rebranding themselves as this. Do you think that moves the needle? Do you think that's an important thing? Um whether it comes to messaging or whether as it comes to like just getting more attention where does it help? Yeah, I think when we talk about robotics, there's a lot of public interest in it anyway, but when we talk about physical AI and that gets to be interpreted as robotics, that really raises the visibility of the industry and it gets the attention of policy makers on robotics as well. I spend a lot of time in Washington DC. I'm heading there again uh later this week. Policy makers are really interested now. They've been supportive of AI to some extent, but now they're starting to support robotics as well. And it's because of those big companies saying physical AI basically equal robots. >> Yeah, I I I followed some of your policy work to sort of study for this interview. What do you think has been one of the most impactful things that you've been able to do with respect to policy to get like a lot more attention around this for the US? Because certainly like you've mentioned, China is rising. Japan has always been great at it. >> Yeah. >> And in my mind, certainly in the US, we do get hurt if we're not at the forefront of these technologies. >> Two things. Um, the US is really worried about China and the leadership that they have in robotics. They see this as a national security threat as well as a competitive threat. The other thing is we started advocating for a national strategy and what we told the policy makers in the US is look this is how Japan did it this is how China has done it we don't have a strategy we talk about it's important to us we want to win but what is the strategy to win and and we recommended six things and and some of these are starting to be um you know perceived is important and legislation has recently been introduced to um take the first step which is creating a national robotics commission. We have no central office who talks about robots. So if this legislation were to pass which it might that would be the first step. >> That's huge. I actually had no idea we didn't have a central body governing important things like robotics. Um this is I'll keep doing this. This is another trap question in the sense that now that we've established that there's some kind of threat, there's a person inside of me that wants to work with everybody, how can we make it so that the entire world can benefit, that we can have partnerships that are extremely beneficial for both countries cuz people don't want to see conflict as much as, you know, it is very much a part of our lives. Is there a way that we can work with other countries, form coalitions to like advance the whole world, not just us? Well, I think we're going to have to do that on certain technologies. I I think, you know, there's a real risk of AI, AI safety is really a hot button issue. And there could be a real risk u when you talk about humanoids that if we don't come up with a way to safeguard them and make sure they're used for good purposes instead of evil, um I think we are going to have to work together. And that's one thing that our association keeps trying to stress. Not all of this is competition. A lot of this is cooperation. So hopefully we'll be able to um continue to cooperate in areas that require it. >> That's great. Do you have like a concrete example of like a movement that we're working with like the rest of the world leaders? Well, we we're very active in robot safety standards and our original industrial robot safety standard was published in 1986 in the US. It became the basis of the international uh standards organization robot safety standard. So, we're active in ISO. You work with people from all over the world on these standards and right now we're working on humanoid standards. >> Okay. Um carrying the same thread. you have 40 years plus of experience in this field. >> Um, is there a specific thing that you think like carries forward? Is there a mistake that we keep making as like a group or like a a coalition of groups that you'd like to see fixed? >> I don't know about the coalition of groups so much that I'd like to see fixed. The mistake we keep making in the United States is to worry about this issue of robots and jobs. And it's continuing to this day. I told you about this legislation that was just introduced into Congress. It has bipartisan support. However, it hasn't been voted on. And when it is voted on, it may not have bipartisan support. Here's why. When we met with Senator Schumer's office, a Democrat, he said, his office said, "We like it. Can you change the name of it? Can you call it manufacturing policy instead of a robotics policy or strategy?" I'm sorry. And we I said, "Well, you could, but it's not just manufacturing that we're talking about." And he said, "Well, the optics of this is that our voters hear the word robot and think it's coming to take their jobs." And the other people who see it the same way are the MAGA Republicans. So, how are we going to have a bipartisan support to pass your legislation recommendation? >> Yeah. >> This issue has not gone away, and it should. And what we present is here's the history of the impact of robots on jobs in America. Whenever robot sales go up, unemployment goes down. When robot sales go down, unemployment goes up. That is not a job killer. Now, you can say whatever you want. Oh, there's just uh you know a correlation, but it's not causation. Okay? But if what the other people are saying is true, you would see when robot sales go up, the jobs go down. But they don't. And the reason they don't is companies become more competitive. They win more business. They invest in more people as well as the automation. That's the winning formula. >> This is great. And I'm going to ask you for citation resources because my whole business is around reducing tech anxiety. Uh that's kind of like the core of what I do. And I want to collect more data around this. >> Yeah. I'll send you the two charts we just sent to the government after uh the meeting I was in this week. >> Yeah. So, moving into like robots will take my job. One of the things that people are crazy about right now are humanoids. >> There's the G2s, the G1s, all of these things. They're like potentially going to do everything for us. For me, the most interesting part about that is how fast do they break down break down cuz we don't have that much data on like whether these things are going to be economically viable over time or will we have to keep replacing parts. What's your feel for humanoids? Are they going to be useful or a trash sheep? >> Who knows? I mean, the reality is that customers are going to decide this if this is the right form factor. There's a lot of solutions out there in robotics that don't require something to walk on two legs. Um, or you could put a a mobile manipulator, like a human torso on a mobile base. But the reality is if humanoids are going to be successful, they're going to have to be safe. They're not. So safety standards are going to be required. And they're going to have to be better at a given task than an existing robot or anything else. People don't even care if it's a robot. They just want a solution. And then when you talk about our homes, that's a whole another area of safety issues. Robot falls on a child. the robot um breaks down, who repairs it? Um there's so many unknowns in a home. Every home is different. Can a robot really be successful in our homes as many people say I read today, if you want to follow me on LinkedIn, I posted a story about by60, this study came out. By60 there'll be more people will own more humanoids than cars. A lot of people don't believe that's true. By60 a lot of people think that is true. But we the point is we have no idea if humanoids are going to be successful yet or not. There aren't enough of them out there. >> Yeah, that's fair. And in fact I think twice we've talked about something that it's at a personal level. Right. The people here especially with the conference today care about the human component. One is that there is a very real feeling of something will take my job away once something that looks like a human, walks like a human, acts like a human, is walking around the, you know, and can can do all these things. I can train it on. >> And then on top of that, you have this idea of ownership of humanoids, which just like as soon as you hear it, you almost want to cringe because like it it feels wrong just viscerally, like owning something that's so close to being human. this imagery of robots being so humanlike and then the ownership of them and them taking away real human jobs. How do you create messaging that helps people with that? This one's complicated because there are some people that don't see it that way. That think that something that is like a human in our homes might be really good for our aging population to have a companion and it's almost like it's a person because we don't have enough people to take care of our elderly population which one day soon I may be a part of. So I I don't know that I would care what it looked like if it was taking care of me. So I don't know. Um that remains to be seen it whether or not people are going to be accepting of the humanoid form factor in our homes or in our daily lives in other ways. So we'll just have to see how that plays out. >> Okay. And you mentioned one really good use case helping the elderly. What are some other use cases where you think robots can actually help people? Well, robots, I think, if you again, if we're not just talking about humanoids, what are the big problems that the society faces, we want to make the planet more sustainable, we want to clean the oceans, we want to um build affordable housing, we want to eliminate food shortages. What role can robots play, traditional robots or humanoids, in helping us achieve those things? I think there's a role in all of them or particularly in the health care field, maybe surgery, but we have some surgical robots now that aren't really robots. They're guided by the doctors. But the real value there would be in helping find new drugs that would treat um diseases that we can't treat yet or find cures for cancers. Um that would be the real benefit. >> Okay. Yeah, I I I am all for that. I helped build the Monarch platform at Johnson and Johnson with like hair surgical platform >> and the mission was very much helping people and and kind of trying to work with all the different components, right? Like there's insurance where like if you get the price down enough with a coupon, then real people can actually access these benefits. Um and I think it's always a holistic approach. The robot in and of itself can't do a thing. You have to build a system around it so it seems like a usable thing. >> Right. Exactly. >> Cool. Okay. So automation gains are super fast, right? Uh, is there a realistic path where embodied AI flows broadly? Like there's there's like less inequality in terms of like where people at the very bottom can benefit in terms of just purely socioeconomics or do you think it's always going to be sort of like top down? >> Another great question that we don't really know the answer to. I mean, if you look at what people are saying right now when it comes to AI, the real risks are for white collar workers and you know, maybe we've trained people to go into the wrong fields. Um, traditionally the fear has been that blue collar workers were going to lose their jobs due to robots. Now, I take a a different view of the job loss issue for both AI and robotics. I think we've always adapted as a nation, as people, to new technologies and we found a way to use them to our benefit. I just can't imagine that that's going to change just because it's called AI or robotics. I just I don't see that. I could be wrong, but I don't see that. >> Okay. Um I hope that that's the case. I I I really do. And I I kind of like this like from someone who's so close to policy, I really like this sort of positive point of view because like you said, sometimes we get stuck on those little things like what is the title of this thing? It enrages me to no end that that's a thing that we actually spend time talking about when there's problems to solve. Um what advice do you have for someone like me who just has no patience for that kind of stuff? You know, it's it's hard to give advice other than you have to take the view that these technologies are going to be helpful to society on balance and not harmful. And if if if it's not going to happen that way, if these technologies end up being harmful, that's really not the fault of the technology. That's on us as people. we should stay in control of these technologies, make sure safeguards are in place and um use them to our benefit. >> Okay, I totally agree with this and I keep saying okay but that seems to be like my word for the day. >> Okay, but yeah, there you go. So that's that's one real really really interesting area for me is security cuz as you know we don't have like a lensure for software development. we do have for certain mechanical fields, for certain like engineering fields and as software and robotics just going to squeeze like the hardware world and the software world together. I personally believe that we should have something for software that holds us to a greater standard if especially if we're going to build stuff that affects safety, affects humans, affects people in like a meaningful way. So what guard rails do you know that the physical world, the robotics world has in place that the software world isn't even thinking about? Well, again, the traditional ones um with traditional robots that you know they they've got to be protected in a way so that they can't hurt you. When you put AI in the loop, that shouldn't change. We still need to safeguard the people that are working around these machines. And again, the form factor doesn't matter. Can be a mobile robot. We have safety standards for that. It can be a so-called collaborative robot. We have safety standards for that. Traditional industrial robot, we have safety standards for that. We need them for humanoids. We need them for every form factor. Now, the AI itself, we may not know exactly what it's doing, but when it's combined with a robot, we need to make sure it's adhering to to standards. >> Yeah, I think that's completely meaningful. It's a huge part of like whether you're working in medicine or you're working in factories, like worker safety, human safety should be at the forefront. It can also be the thing where we get completely clogged down, right? Um you can create regulation that'll keep you from making any kind of progress and certainly the US is uh sometimes guilty of this. >> We talked to policy makers about that and um I think more and more are understanding we don't want to overregulate this but we want to be assured that people are going to be safe. >> Yeah. >> You know there was some talk about maybe we need a robotics regulatory agency in the U. No, you don't you don't need that. Um there are really good standards. Companies follow those standards. We should be safe. We have been the industrial robot industry has a a very good safety record over the years. And the reason is in the early days whenever anybody got hurt because of a robot whether it was the robot's fault or the person's fault didn't matter. That had a chilling effect on the industry. Yeah, >> we've seen that with autonomous vehicles. >> Yeah. >> One one autonomous vehicle does something and somebody gets hurt or god forbid dies, it shuts down for a while. And we need to make sure that safety is job one. That's always been our mission. >> Okay. And and I see K. Okay. Um the there's this really interesting piece here. We hold robots to a greater standard. A human getting into a car accident is an everyday occurrence and we have statistics around that. Robots do not get held to the same standard. It's a much higher one. Much higher. What do you what are your feelings about this? Is this the right way to approach this or is it getting in the way of innovation or is it the right thing to do for humanity to give us all of the uh the the good feelings about building a new technology? >> I think it's the latter there. I think we need to make it a priority. I understand that autonomous vehicles on the whole would probably be safer than a human driver who's texting or on the phone or, you know, has had too much to drink. Um, but we don't want machines ending up killing us. >> Yeah. >> And we need to make sure that they're capable of doing the job that will keep us safe. It's not going to be 100% but we want it to be as close to 100% as possible. I don't think we should back away from that. >> I agree. Uh 100%. Before I keep going, uh many of you have been listening to us. Do you have some questions? Uh yes, please. Uh yeah, of course. Just yell it out with all your strength. >> I don't have such great hearing though. >> Sharing. Um I want to push back a little bit on the optimism you have for your policy making because it's not everything has been having like great track records for example like social media right like because a lot of the effects of certain new technologies they're not just you know immediately seeing like we we didn't know what social media were going to have like such a polarized um society and misinformation and uh basically targeted information four different profiles and different geographies of people. They're just in that information bubble because advertisements want them to be. And also like girls killing themselves because they have to have this 247 competition of beauty contacts online and people are addicted to their phones that they're now socializing. There's supposed to be a social network for the well-being of people, but now people are just not socializing at all like in person. These are all like really like emerging you know sort of like phenomenon that we didn't really anticipate when say oh just a bunch of people you know talking to each other online great that's great discovery but but I'm just pushing back on the optimism do we actually really know like different scenarios of like how this can change social uh society for example one concern I have with AI and robotic is that humans will have to compete with machines what does it do to the human sense of value and dignity So if you're not someone who's so brilliant to manage a bunch of AIs and robotics now you have to really just compete and does that mean that category of humans will just be so deolished in their intrinsic value? So that that'll be my question. Thank you. >> A really really uh wellthoughtout question. Thank you. Obviously, I can't answer what that future is going to look like in terms of people competing with robots and AI. I can only tell you what history has shown. People don't want to do dull, dirty, and dangerous jobs. There's a shortage of, I think, four or 500,000 welders in the United States. Robots do those dull, dirty, dangerous jobs and have for decades. This is not a new technology. When humanoids finally find a use case, they don't have one really yet. People say, "Oh, they can do this. They can carry a tote in a in a in a warehouse." Um, people aren't signing up to do those jobs either. Amazon has over a million robots in their warehouses, um, distribution centers. They've hired a lot of people, a lot more people because they were able to open up more of these uh distribution centers. And we've actually created jobs because of robots. The real threat to somebody's job is when their company goes out of business. It's it's not that it's competing with a robot. You're comp your company is competing with some other company somewhere else that might be implementing robots. If our companies don't take advantage of the most effective tools out there, all the jobs are at risk. We saw this. We saw what happened in the United States. Um, in the last two decades or more in Detroit, where I live, a lot of the automotive manufacturing jobs were shipped to China to chase lowcost labor. They could have automated and kept these facilities, these factories in the United States. I grew up in Detroit. I still live in the Detroit area when those factories closed because the we were chasing lowcost labor. It wasn't just the people who work in those factories that lost their jobs. It was the people in those neighborhoods where those factories were. There were restaurants and bars and bicycle shops and and dry cleaners. They all went out of business. these neighborhoods hollowed out. The tax base, the value of homes, the quality of the schools, they all were suffering because those factories went away. And so I don't feel like that the enemy is the robot. The enemy are the companies who aren't taking advantage and staying competitive. Yes. >> Yeah. So uh at the state s um my background is actually a factory automation. I'm an industrial engineer. So I used to work with Rob Automation Radley. So pretty well known I guess in the Detroit area >> and uh and uh you know moved over to the software world but uh you know I've spent time working on very complex safety systems like oil and gas where and then moving into software and data analytics uh side of things where it was just this attitude of [ __ ] it ship it and uh over there in oil and gas you have like a triplex system that is on like a large compress a large gas turbo bine that's controlling a compressor and all that equipment is getting retrofitted nowadays to be more and more advanced. Um, so the interesting thing is there are going to be more jobs. It's as people have to look at how to find that opportunity and right now there's an opportunity to turn the rust belt back on. uh with AI and you know because these systems emerged initially from like MRP material resource plat ERP was just kind of trying to figure out where these systems go and now we have the ability with AI to connect at a greater level and also translate things because uh when I was doing factory automation I have to go and talk to multiple departments to get something done you know a process engineer is not going to understand the mechanical department's complete needs and translate it over uh the wallet over to the the finance department and the glue that kind of held that together was like the sales engineers who had a deep understanding of the application and how to apply it. So now with that same opportunity we could actually apply that across the board and it is comprehensibility is the main issue that AI can solve and if you put that between yourself and the systems you could also control what information comes to you because you don't necessarily need a screen. So the war for attention can also change dramatically. So >> Oh no, I just wanted to ask what was the question? Sorry. >> Oh, I I just had a comment uh comment on that uh on on the AI it kind of where where things are going with that. I I just have a different perspective if it a way to different apply it differently. That's what I build systems for. >> I kind of want you up here next time. >> Sure. >> Um we have time for one more question if anyone has one. Oh, >> question from the crew. >> You mentioned doing marketing in this space. I've been in other situles that translate technical stuff to help people understand. >> Um, and so, uh, I think that's really fun being able to be there to learn that and then help other people understand that. Um, and I was curious what sort of terminology that you came across in the industry because people in industry, you know, they have these things, they love it. They love using it. They think it's use it like it's the the best way to describe something, but no one else wants to hear that or understands that. And any fun ones that sent out to you where there was the technical term and then it had to be this totally different thing for other people to get it. >> Oh, people would tell me about the number of axes the robot had. And people now um you know they're talking about agentic AI and tokens and um look I don't know what this stuff is. How am I going to explain it to some person who who's even has less idea than I do about this stuff and there there are a million of these things but those are the ones that pop immediately to mind. I I have no idea what a token is when it comes to AI and yet I hear it talked about all the time. How do I explain that? >> You should come to my class. >> That's what I don't want to do. I don't want to have to know that. >> That's fair. Degrees of freedom is a big one, right? Like that that degrees of freedom. >> Oh, degrees of freedom. Yeah, that's another one. Yeah. >> Yeah. Yeah. Yeah. It's it's it's so much there's so many interesting terms to go by. I'm I'm glad you have a positive outlook. I'm glad people are questioning the positive outlook because when we have that kind of discourse, meaningful output comes from that. And maybe somewhere along the lines, we can help society in a meaningful way. And I think the whole point of this conference is like how do we make sure a lot of that happens from the bottom up where we help the most amount of people rather than just the very top firms that can afford this technology. And it's becoming more of a truth. So, uh yeah. Any last comments you have? No, but I would say um if we do this right, there will be more jobs created and they will be available for everyone in society. But one of the things that we have to do is retrain the workforce for what those jobs are going to be. And that's part of our recommendation to the US government in our national strategy. We have to invest in workforce training and educational programs. If we don't do that, then we won't be able to take advantage of these tools. Yeah. Uh, perfect. Thanks very much. Uh, and thank you for coming today. Yeah. >> [music]