Soonish - Soonish 5.17 David Mindell on What it Takes to Power an Industrial Revolution

5.17 | 10.24.25

David Mindell is a historian, an engineer, a startup founder, a venture investor—and now the author of The New Lunar Society: An Englightenment Guide to the Next Industrial Revolution. The 2025 MIT Press volume is all about James Watt, Matthew Boulton, and the other inventors and entrepreneurs who kickstarted the first industrial revolution in Great Britain back in the late eighteenth century, and what they got right and what they missed about how technology can transform work and how to translate invention into social progress. But it’s also about engineers innovate (or fail to innovate) today, and what they might learn or relearn if they took a look back at that founding generation of industrialists.

Mindell, who's been a friend ever since we were both doctoral students in MIT's Program in Science, Technology, and Society in the early 1990s, is the perfect guest for this 60th and final episode of Soonish. The show has always been motivated by a set of big questions: How is computing changing the nature of work, play, artistic expression, and communication? How can we design our cities, our transportation systems, and even our political systems to be more resilient? In an economy dominated by strife-fueled social media and rising technofeudalist empires, what's the future of democracy? How much of our techological future is predetermined, and how much of it can we shape proactively?

David brings to bear the tools of historical scholarship—along with his experience in engineering, academia, and the entrepreneurial world—to explore the same kinds of questions. This new book, in particular, asks how Watt, Boulton, and their colleagues distilled Enlightenment scientific values, hands-on experimentation, and collaboration into a set of founding principles for industrial society—and how can we rethink those principles for a world of labor scarcity, climate change, pandemics and other global disruptions, and burgeoning new technologies like artificial intelligence.

Notes

The Soonish opening theme is by Graham Gordon Ramsay.

If you enjoy Soonish, please rate and review the show on Apple Podcasts. Every additional rating makes it easier for other listeners to find the show.

Hear about everything going on about Hub & Spoke in the collective’s free bimonthly Substack newsletter, Penny Farthings. Subscribers also get access to our quarterly anthology show, The Hub & Spoke Radio Hour.

Full Transcript

[Music: Hub & Spoke Sonic ID]

David Mindell: Hello, Wade. Whoops. Hold on.

Wade Roush: Yeah. Hey, David.

David Mindell: How are you doing?

Wade Roush: I'm doing great. How about you?

David Mindell: Well, I'm got a cold, but otherwise I'm okay. So if I'm too sniffly for you or whatever, we can do this a different time.

Wade Roush: But no, you sound great. So I have a bunch of questions. I wrote them out in sort of the shape of a conversation that I'd like to walk you through.

David Mindell: Is this for Soonish, or are you calling it something different?

Wade Roush: Yeah. It's. This might be like the last episode of the show in its current form.

David Mindell: Okay.

Wade Roush: Being out here in New Mexico gives me, has given me a lot of new ideas about where to take the show. So I think the next episode after this will probably be very different. But this is a great episode to end on in a way. As we'll discuss, we'll discuss.

Wade Roush: You're listening to soonish. I'm Wade Roush.

Wade Roush: My friend David Mindell is one of the most brilliant and accomplished people I've ever met. He's an author, a historian, an engineer, a startup founder, a venture investor, a husband and father, and even a general aviation pilot. I've known him ever since we were both graduate students in the history of technology back in the early 90s, getting our PhDs from an outfit at MIT called the Program in Science, Technology and Society, or STS. After he got his degree, David went on to become a full professor at STS in the history of engineering and manufacturing, with a joint appointment at MIT's Department of Aeronautics and Astronautics. And that's still where you'll find him today, focusing on research areas like autonomous systems and the ways humans and machines collaborate. Now, if you're an Apollo fanboy like me, your favorite David Mendel book might be Digital Apollo, which was all about the guidance computers on board the Apollo command module and the Lunar Excursion Module, and the constant give and take between those machines and the astronauts themselves.

Wade Roush: The Apollo astronauts could not have gotten to the moon without automatic controls, but they were also very reluctant to give up control over their spacecraft. And Digital Apollo is a gripping account of that tension. But last year, David came out with his seventh and perhaps most prescient oriented book. It's called The New Lunar Society An Enlightenment Guide to the Next Industrial Revolution. It's all about the scientists and engineers who kickstarted the first industrial revolution in Great Britain back in the 18th century, guys like James Watt and Matthew Boulton and what they got right and what they got wrong about the process of innovation. But it's also about how engineers innovate or fail to innovate today. And what the inventors and policymakers of the 21st century could learn or relearn, if they would just take a long look back at that founding generation of industrialists. Like all good history books, it has a lot to teach us about the changes we're living through today, including the rise of the internet and artificial intelligence, and the way those technologies are changing the nature of work. So naturally, I wanted to ask David to come on the show to talk about all of it.

Wade Roush: But I also had an ulterior motive, which was that I knew chatting with David would be a great way to return to the themes that inspired me to start this podcast back in 2017 and bring some of them full circle. Because here's the deal. I've put out 60 episodes of soonish over the last eight and a half years, and I've had a sense for a little while now that it's time for me to wrap it up and embark on a new and completely different audio project. My talk with David took place back in May of 2025, and I could have shared this interview a lot sooner, but I didn't want to publish the final episode of the podcast until I knew more about what was coming next. Now I think I finally figured that out. To hear more about what that is exactly, just stay subscribed to this podcast feed where I'll drop some news about the next podcast as soon as it's ready. Meanwhile, thank you so much to everyone who's been listening to soonish over the years and to everyone who still believes that the future is shaped by technology, but technology is shaped by us.

Wade Roush: David Mindell, thanks for joining me.

David Mindell: My pleasure. Always happy to talk. We've been doing this for 30 years.

Wade Roush: We have been, if not longer.

David Mindell: Longer, actually.

Wade Roush: I've been looking forward to this for a while. So. Okay, so, David, this is your fourth or fifth or sixth book, right?

David Mindell: Seventh.

Wade Roush: Seventh. Okay. So your passbooks have all looked at how technology is intertwined with human practices in various ways, right? So, uh, you did a book on the development of the first ironclad warships during the US Civil War. Uh, you wrote about the rise of feedback in computing and control. You've written extensively about the US space program, uh, robotics. So, but this is a very different book, and it's kind of made up of a series of essays and arguments more than it is like a focused history of any particular thing. That said, I want to start with the history. So for folks who don't know what the Lunar Society is, who was in it, what its significance was. Can you start off with a bit of a, uh, an outline of, of this group and what was going on in those decades, in the late 18th century when England was beginning to industrialize? Who were these guys? How did they come together and what was their significance?

David Mindell: So the Lunar Society was a group of really friends, um, who met on a regular basis during the 1760s through the 1780s, 1790s. And, um, they were composed of scientists and doctors and other kinds of thinkers, and they would meet for dinner on the full moons. That's why they called it the Lunar Society. So they would have some light to walk home by in Birmingham, England. So very much not London, kind of out in the industrial periphery of the of, of Britain at the time. And um, they talked about the big ideas of the scientific enlightenment they started maybe this is 50 years after Newton's death. So his revolutionary ideas, in a way, are way, are just beginning to be absorbed into the broader world. And, um, they were all interested in, you know, the transfer of heat and, um, botany and geology. And although many of those sciences hadn't really existed yet, and it was really started by Matthew Boulton, who was a manufacturer of small, cheap metal objects which were called toys in Birmingham in the middle of the 18th century. And he was a very, um, kind of gregarious, outgoing, um, intellectually curious manufacturer who also was a great innovator in the manufacturing processes of the day. And his friend Erasmus Darwin, um, the, the, the grandfather of the great naturalist Charles Darwin and who was a physician and, um, they became kind of fanboys of Benjamin Franklin reading about Franklin's electricity experiments, Franklin came to visit them and taught them how to do experiments in electricity, and they were so inspired by Franklin that they decided they wanted to gather as many people as they could to have these kinds of conversations.

David Mindell: And, um, the Lunar Society never had any formal membership list. It wasn't a real organization in any formal sense. One of the things I like about it, and they didn't keep records, they just hung out and talked. In the end, it was the application of those enlightenment ideas to the industries that many of the members were a part of that really changed the world. So, um, many familiar names were part of this group, including Joseph Priestley, the great chemist who really laid a lot of the foundations for modern chemistry. Um, uh, James Hutton, who founded the The modern Science of Geology, uh, Josiah Wedgwood, the pottery manufacturer who's a big character in the book, who was actually Charles Darwin's other grandfather. So in some literal way, among other things, the Lunar Society spawned the most important scientific idea of the 19th century, and the Wedgwoods and Darwin's were intertwined as families for many generations. And it was just this remarkable group. But it was also the group that brought in a young instrument maker from Glasgow. James Watt and Watt and Boulton formed a partnership to build steam engines. And that really was the founding moment of what we now call the industrial revolution. And so just a really remarkable group of people who interacted in just lots of interesting ways. And I think I and many others who have studied them just became kind of fascinated and a bit addicted to the stories.

Wade Roush: So this relationship between Boulton and James Watt is kind of like the central relationship in the book or one of them. So could you talk a little bit more about what and who he was and how his partnership or friendship? Both, I guess, with Bolton came about?

David Mindell: Sure. What um, was from Scotland, um, and was living in Glasgow, and he came from a family of sort of very practical makers who worked in the shipping industry, um, making everything from blocks which are pulleys for ships, um, but also instruments. And this is a period when scientific instruments and navigational instruments are very similar, and the skills of making navigational instruments are very important to the shipping trade, including the slave trade and, um, what is trained in that discipline? He actually goes to London and he trains in London for that. And then he returns to Glasgow and he's given an appointment as the kind of instrument maker at the University of Glasgow, and he has a little workshop right there on campus. In fact, it's kind of amazing how I work on a campus and everybody's all excited about maker spaces. And that's exactly what what had. And there was something about him that was just he was very skilled. People loved to watch him work. And he also would have conversations and talk. And so there was a bit of a kind of mechanical salon going on. Adam Smith would drop in and visit us. A number of other instrument makers, chemists, professors were sort of part of this little circle around James Watt.

David Mindell: He was a very kind of diffident character. He didn't believe in himself. He was always depressed. He always struggled economically. He had a very tough life. Um, his first wife died at a very young age, leaving him with two small children. But he was a brilliant draftsman, a brilliant designer, a brilliant thinker, and in many ways the prototype of what later became known as an engineer. Um, he was interested in science, skilled in science, but also skilled in hands on mechanical arts, and was able to fuse the two of those things in his work. But as with many startups today, he always struggled to find backers, and when he came up with this great idea, the the the idea not he didn't invent the steam engine. He invented the separate condenser, which was a critical improvement on existing steam engines that made them practical and much more efficient. And even then, he struggled for years to find time and money to sort of the same thing for him to develop those things, and was at the end of his rope and really unable to do it on his own in Glasgow.

Wade Roush: And this is when, roughly around the time that he gets to know Boulton and he gets gradually drawn into the Lunar Society circle, and Boulton winds up helping with that problem of time and money. Right?

David Mindell: Right. Well, actually, he first visits Birmingham and Bolton isn't there. But he's toured around the factory, and he immediately sees that there's something here that he doesn't have. And actually, he uses this great phrase. He says, what I need is engineering and great because he's able to make his steam engine model work at the level of a laboratory tabletop device. In fact, that device still exists. But in order to make it practical, you need to build it into something that's the size of a building and to extend the the precision that the instrument maker would use at his small lathe to this amazingly large scale, which was the province of of a profession called millwrights at the time, the people who made mills, who were also kind of early engineers. But what wasn't a good manager, he didn't relate well to the millwrights, and they didn't realize it at a time. They needed a whole new industry of machine tools to make steam engines actually work. So he was very much struggling with scaling up this invention of his to prove that it worked in a real engine. Bolton didn't necessarily have all those things, but Bolton had the scale of the operation, the workmen, the, the tools, um, and the capital, um, to be able to do that.

David Mindell: And actually when, when Bolton, on his second visit to Birmingham, Bolton and what met and they spent two full weeks together and they really hit it off and, and that's at some level where Bolton really had a kind of picture of what the Industrial Revolution would look like, that what himself did not have, what imagined building better engines to pump out mines, which is what steam engines were used for at the time, Bolton was himself operating a factory. He was very frustrated with his water power because it froze in the winter and it dried up in the summer, and he didn't have regular power. So he knew what a customer would need from a steam engine. He had been experimenting with engines on his own already, and much to his credit, I think when he met what he realized what had a great deal that Bolton didn't have either. You know, I always say he had the humility to know when excellence had come to him and recognized that in James Watt, and immediately stopped his own experiments with steam engines and began negotiating with a license to get a license to the Watt engine.

Wade Roush: Right. And the two of them do negotiate a license. And basically, you know, if you fast forward, that was the sort of moment that enabled Watt to really kind of finish scaling up the idea. And then, you know, within 10 or 20 years, there are hundreds of these steam engines with separate condensers, like running mills and factories all over that part of England.

David Mindell: Right, exactly. And so what moves to Birmingham is actually right around when his wife dies, he's at the very bottom of his life personally. He's out of money. Um, again, he was a really depressed, not an easy person to work with. And, and Bolton was a kind of cheery, outgoing personality who tolerated what for many years. And, um, at one point, what's his second wife actually writes a letter to Bolton and says, I worry my husband is suicidal. Can you help me? He's really in the bottom. But Bolton is able to push him in a constructive way. He makes four additional inventions that are absolutely critical to making the steam engine work practically. And he's not. Bolton is not willing to stop with the mining engines, insist that they build a rotary engine that will drive a factory, as one of the historians says, drags what kicking and screaming across the threshold of his own mediocrity and really makes what into the great person that he's later kind of lionized as really the first modern engineer?

Wade Roush: Yeah, yeah. So I love this part of the story. Um, so you're a historian of technology by training, as I am. Um, you're also an engineer and an entrepreneur and an investor and a teacher. So I'm curious about, like, how you come at these two characters. Bolton. And what? So what is it about them? What qualities do they have that speak to you, and what do you think today's industrialists could stand to learn from them? And we're going to we're going to expand on that question for the whole rest of the conversation. Right.

David Mindell: Well, again, just the nature of the partnership is fascinating because they bring such different skills to the table. And again, these men are lionized as these great founders. So a lot of modern historians will reject these kind of heroic mythologies. Rightfully so in many ways. You know what is not buried in Westminster Abbey? But he is the first engineer to have his bust memorialized there. He's a great kind of founding hero in in English mythology. I was very interested in, like, what that means to be heroic as an engineer. And what are those kind of personal qualities? What's interesting about this period of time is, and for many years after is that heroes are sea captains and admirals and generals and great warriors and even priests and royalty. But the notion that these kind of nerdy, you know, mechanics, which was the condescending term used at the time, not nerdy, but mechanics could actually be heroic, is a very modern idea, and I think a very hopeful idea in a lot of ways that they're not out there fighting wars, they're not using physical force to conquer anybody. They're building things and they're improving systems and networks and doing so in a kind of relentless way that is not really captured by the phrase the pursuit of profit. They're very clear that they're trying to improve the world as they see it, and it takes another 50 or 75 years. But the idea that that kind of character is a modern character, to me, is something that we should be re-examining and maybe even celebrating. Um, again, I think a lot of tech titans today have believed their own mythological press and return to this kind of Homeric, conquering hero sort of thing. And, um, I think there's a lot of value in re-examining these characters who are, frankly, we're a lot more humble, even in their days, and really focused on building the world, providing things that people needed and wanted, and that was their way of changing the world. That, to me, is a form of human action that that we need more of right now.

Wade Roush: So the format of the book is a little unusual, David, in that, um, you kind of interleave chapters about the Lunar Society with essays that really reflect on today's challenges, especially regarding manufacturing, industrial innovation, reindustrialization the future of work and what jobs will be like. Talk a little bit about how you decided to organize the book that way. It's it's a it's a very personal take on history in a way. And very unlike, as I was saying earlier, it's not at all like your earlier books.

David Mindell: Great question and big topic. I when I have written history before, I always kind of try to stick to chronology and there are always multiple chronologies, but, um, it's the easiest way to tell a story when you depart from chronology, you can do it, but you impose a kind of cognitive cost on the reader that you risk losing their attention. And so you have to work hard to make sure you keep a thread going. This is the first book I've written that doesn't do that. Um, no. Partly that was because I've done those kind of chronological histories and used historical narrative enough times that this was an interesting writing challenge for me, that I know how to write the historical narratives, but to do this unusual format made the writing challenge interesting and new. Um, my publisher actually, in all truth, said, you know, don't just put all the history first, you're going to lose people who care about the current day stuff, which I think was good advice, because it really is aimed at a present day audience. And so I took it upon myself to do this context switching every other chapter, certainly for the first two thirds of the book and a little bit of trust me, these two things are going to converge. But as the book goes on, they do refer to each other more and more.

David Mindell: And then the last few chapters really are talking in a combined way about where we are and where we're going. And you can't write about the present in the same way you write about the past. So it's tempting to say, I'm just going to write about the Matthew Boulton and James Watt of today, but I don't have their letters to their wives that I'm reading. I don't have the benefit of hindsight to know who is successful and who is not successful. I don't. There's a lot of and all of those intimate details you couldn't write about a present day company anyway. They're confidential as well. They should be. And the personal details too. Whereas with these folks, 250 years ago, you know, those letters have all been archived. You have access to a level of intimacy that you just aren't going to get in the present day. And so, um, at the same time, I think the historical stuff is both strange enough because it's so far away and so different. The world they lived in, but also familiar enough that we can reflect on. It helps us see the world we're in today as a little bit strange. And what's strange about it, and that's helpful because it's hard to do when we're so immersed in it every day.

Wade Roush: Yeah. So there is the novelty of the challenge of writing history in a different way. But also this is, like you said before, this is one of the most deeply covered parts of the history of technology. There are hundreds of books about this era and about steam engines and about Newcomen and that engine, and then Watt and Boulton. And so you wouldn't have just set out to write another book about Boulton and Watt anyway, right? I mean, I'm assuming that, like in the very conception of this project, you were thinking there's something about this collaboration and, and this group of people around the Lunar Society that speaks to us today. And you wanted to make sure to to bring that out by, like, literally telling the two stories simultaneously.

David Mindell: Exactly. And, you know, really comes out of this, I think first, coming out of the pandemic, where we collectively realized that there was a way to invent work that was totally different from the way we had been working. And as you know, for historian of technology, work is such a fundamental idea. And here we are, thrown into this remote world, and now we're in some weird hybrid version of it. But people realize that work could be different from 9 to 5 in the office every day, which it always has been for a lot of people. But for the vast majority of kind of modern workers have worked in that kind of 9 to 5 office setting. And, um, so suddenly there's a possibility of a change in work that's more than a generational change. It's a change in a pattern that got established at the beginning of the Industrial Revolution. It took 100 years to teach people to come to work at the same time every day, and to go to work for the same number of hours every week. Um, and factory owners were constantly complaining. Look at these crazy people.

David Mindell: All they do is come to work long enough to feed their families, and then they go home and enjoy their lives. That's not industrial work. Um, and so what are the possibilities that are open today, beginning with work. But obviously all of the latest developments in artificial intelligence, the changes in the global supply chains, obviously, the political moment we're in and of course climate change. We need to be reinventing our industrial systems. And they don't have to look the way that they've looked before. And that just kind of naturally led me to these people who invented these industrial systems and looked at the world they lived in and said, it doesn't have to look the way it looks before, but we need new ways to transport things. We need new ways to work. We need new ways to make things. We need new ways to study things. And we're kind of at a moment like that again, where we have to reinvent this industrial system and we should keep what's good about what we've done and not keep what hasn't been working.

Wade Roush: I'd love to hear you say more about how and when you came to understand that we are in one of those moments when we really do need to rethink how we build things, how we ship things, where we get the energy to do all that. I imagine that it was a combination of your your, your academic studies and your practical hands on experience running companies. And I'm asking you to recount your own personal journey to this realization. Like that. We we really are kind of in trouble. We were pretty good at building things, and now we've outsourced a lot of that knowledge. And if we really if we care about kind of figuring out a humane and equitable and climate friendly way to to move forward, we're going to have to rediscover or relearn a lot of those skills.

David Mindell: Yeah. So there's a chapter in the book called The Great Toilet Paper Crisis, which most people know what I'm talking about, just by saying that, um, was in those first weeks of the pandemic when, um, the world felt like it was shutting down and Americans somehow came to the conclusion that they were going to need more, that they were worried about having access to toilet paper in a pandemic world. And, um, there were runs on it in various places, supply shortages and so on. And that was a sort of funny in retrospect, but revealing in the sense that, um, people knew that in some visceral way that things come from somewhere else, and if they couldn't come from somewhere else, they were going to be limited. Which is also funny, because if you think about the amount of toilet paper the country needs, if you're working from home or you're working at work, you need the same amount of toilet paper. And I go into that a little bit in the book. But the pandemic had this way of exposing these global supply chains that we all depend on your computers, your sneakers, your, you know, little trinkets that you buy on Amazon, whatever. It all has to come from somewhere. And that's something that, as a historian and a kind of systems oriented engineering thinker, I've thought about for a long time.

David Mindell: But here it was in 2020 on the public consciousness in a way that it hadn't been before. And even this term supply chain, which was really a kind of, you know, jargon for academics and other logistics professionals, suddenly becomes on everybody's lips and exposed the fact that we live in these global systems that are actually quite fragile, very dependent on fossil fuels, very fragile, very dependent on labor systems that we're all part of, but also are kind of invisible to us. And then there are various, you know, climate induced crises going on around the globe and going to 2022. The Russian invasion of Ukraine threatens all that stability even more. That really got me thinking about the foundations of industrial civilization, which we all live in. And as I quote a journalist in the book, it became clear that to decarbonize, we could not de-industrialized we would have to reindustrialize. So, um, if if you think that there's an urgency to addressing climate change, there's no way you're going to get two solutions to the climate crisis without making a whole lot of new stuff. And whether that's wind turbines or electric cars or electric airplanes or whatever else. Um, it's also going to include a lot of really boring stuff like wire and connectors and electrical plugs, and that's all stuff that needs to be manufactured.

David Mindell: And I think among the the academic certainly among academic humanists and to some degree among academic engineers, there's this idea like, oh, you know, somebody else imposed all this stuff on us and, and we just have to throw off the shackles, right. When actually, like, industrial systems, they're part of us. We are the systems that that we've produced that support us. In fact, there's a chapter called Supply Chains Are Us, right. Every time you click something on Amazon, you are supporting a global supply chain. It's not something that's out there imposed upon us by other people. It's something that all of us benefit from and contribute to the to the harms from. And we have to kind of own that in order to improve it and transform it. I mean, even in the the field of the history of technology, which you've been a part of. There's this sort of like distanced sense that, you know, oh, all these fossil fuels are are terrible. And yet, like, how did you get to work today? Um, and where does your cell phone come from? And even the strawberries on your plate are there because of fossil fuels. And one has to begin to see that everywhere in order to understand the work of transformation that we all have to take on.

Wade Roush: Which is going to be an incredibly interdisciplinary challenge already is. And, um, it takes me back to a question about the Lunar Society, I guess. So you say toward the end of the book that you've been convening your own dinners where you bring together friends from the technology world? I'm assuming from the academic world, maybe from the entrepreneurial world. And you've called it the Lunar Society in homage to the originals. Um, why start such a thing. What kinds of stuff do you talk about at these dinners? And do you feel like you're you're trying to get at the modern day equivalents of the, of the technical and scientific sort of webbing that you had to that the original Lunar Society had to work out. They were building like they were not only like trying to integrate the latest science into engineering, but they were also like literally forming the business partnerships that would enable these engineering ideas to to scale up and be commercialized. So a whole bunch of important stuff was happening. Do you have similar ambitions for your new Lunar Society or what are you doing there?

David Mindell: I mean, I do have similar ambitions, but only in the sense of like the most important things that came out of the Lunar Society were people working with each other. And that's true of the the dinners that we hold today in that if one person meets somebody else and they have a new idea or even just talk to each other and begin a relationship. Um, in, in furtherance of their work. Then then it's a success. Like the original Lunar Society, the group that I convened has no status. There's no notes, there's no, um, membership lists. Um, it's just the pure value of it is the conversations people have with each other and what they take from it. You know, there are people who have said the the original Lunar Society played the role of a industrial research establishment. And, um, today we have universities and universities kind of play that role. And that's true to some degree. Although universities are very highly professionalized and very highly disciplined, and it's much harder for people to walk across the various disciplinary aisles. Plus, the kind of entrepreneurial nature of modern research has meant that everyone's always just kind of hawking their own interests. And, um, at least we try with the current Lunar Society to a not fill it with academics, because that's easy to do when you work at MIT. But I don't do that. Um, and bring people into contact with each other who wouldn't meet each other otherwise. Um, and really, um, have some conversations about where some of these things are going. Ultimately, the goal of the book and the goal of the Lunar Society is to be able to find a way to talk about the future of industry. That's not like Silicon Valley utopianism, that's realistic and grounded in the real world, but also optimistic and hopeful.

Wade Roush: Speaking of optimism and hope, nine years ago, maybe, maybe it's been ten years now. You, um, took on the challenge of starting a company called thematics around some of your ideas in, uh, radar and automation and control. And so you went through that whole process of trying to scale up an idea firsthand. And then after that kind of as an act two in the world of technology and business, you went on to co-found a venture capital firm called unless where you're investing in some of the entrepreneurs who are in some ways pushing forward on this vision of a revitalized American industrial scene. And you can't do any of those things. You can't be a company founder, and you can't be a venture capitalist unless you're sort of, like, innately optimistic and maybe even insanely optimistic. Like, um, so I wondered if you could share some stories, maybe from Humanichs or from unless that speak to this, this larger issue we've been talking about of, around how we need to kind of reinfuse, um, the, the system of, uh, of creating engineers and putting them to work and funding them in the United States or maybe around the world in order to deal with these challenges like climate change.

David Mindell: Mhm. Yeah. I mean, you're certainly right to point out that this book is heavily informed by my own experience as an entrepreneur, company founder and then more recently as an investor. Um, and I think one of the major things I take from all of those experiences, which are all still ongoing, is, um, um, we're actually quite good at inventing things in this country. Um, and the R&D system, maybe until the last three months, has been very healthy and strong and, um, but we're much less good at adopting it and deploying it into really changed the way that we do things, especially in the more industrially oriented sectors. There are some good reasons for that. If I'm flying airplanes around and I need to keep people safe, I have a very high bar for what I'm willing to do differently. Um, and there's some not so good reasons for that. And the, you know, the kind of people who work at these scrappy startups, um, like to push fast and want to really create things and apply them. But the way the economy is structured right now, very many of the decisions about adopting those technologies are in the hands of people in very large, slow, lumbering organizations, whether they're public sector or private sector and very different kinds of people who work in those organizations. And we really are um, I feel like the transformational work we need to be doing is kind of grinding to a halt in these moments of adoption and how those decisions get made. Um, I think part of that is actually we've hollowed out a good part of the workforce that is really the folks who can do that well, and those are people who have kind of risen up from the shop floor.

David Mindell: I use the analogy of master chiefs in the Navy, who are the senior enlisted people who really know how the ship runs, as opposed to the junior officers who are sort of like clueless college boys. Every factory has senior, um, tradesman, senior, um, blue collar workers, um, who have been through the ranks and, and are well positioned to adopt technology and and transform it. But we've hollowed out that segment of the economy over the last 30 years through globalization and it and a lot of other things. And, um, I really increasingly find that it's that missing middle, that that's where the action happens for adoption. And that's a key part of the workforce that needs to be revitalized. And part of that is just by kind of, um, starting to see it again and not making it invisible and, um, elevate and recognize the work and the intelligence that's, that happens in, in that part of the work spectrum. And, you know, I feel like the moats around MIT are much deeper than they've ever been, and we're much further from actually getting our stuff out there at scale than we have been before. And there's a lot of reasons for that, but this is one of the key implications. You know, we've become great at product innovation and very poor at process innovation. That's a big theme in the book.

Wade Roush: You just hit on three things I want to try and come back to in turn. So well, one of them is the the missing middle that you talked about the the master chief personality being selected out of the labor force. One of the things you've worked on in the past decade has been this very large project at MIT called The Work of the future. And you co-authored a book by that name, and the book was the end result of a long internal study at MIT about labor and technology and economics. And, um, in the context of everything we've been talking about. I just wondered if you could extract 1 or 2 of the themes of that project that are maybe the most relevant to you, or the most critical to you?

David Mindell: Sure. So from 2018 to 2022, a bunch of us at MIT did this work of the future project. It was originally addressed at this question of somebody literally came to the president of MIT and said, I need MIT to tell me if a robot is going to take my job. And that was 2018. And of course, between 2018 and 2022, the pandemic happened, and we ended up looking a lot at these kind of technical questions around workforce and, um, industrial structure in the United States. And, and to some degree, this Lunar Society book is my personal sequel to that project. I learned a lot from lots of different folks, um, who we collaborated with there, um, about how to think about work, about about the current state of the workforce in the US, about some of the economics that surround them and other technological issues, from driverless cars to other things. And as a historian of technology, again, you're taught you can't see technology without seeing labor. And yet labor is really made invisible in most of our daily lives. And most people don't see the labor around them. And the labor that goes into either the products or even more so, the services that they consume. And, um, how do things look different when you can see that labor and understand a little bit about the various economic and social forces that is creating that labor. And again, that missing middle idea very much comes from that work. And um, but then I kind of started looking at it and say, you know, we also have this thing we talk about called the Valley of Death, which is when technologies are having trouble that once they've been developed, getting deployed, all the stuff early in the technology cycle are done by the coastal elites living in cities with advanced degrees and very nice career prospects.

David Mindell: And, um, the technologies that those folks, including me and many of the people I work with are generating are having a lot of trouble finding their way across this kind of workforce Valley of death, to be really taken up and adopted by the people, um, who need it, who are the people who own and manage and run the process? And they are conservative by nature, as they should be, but also they really need anything that is going to help them do their jobs. The Lunar Society book is in some way an opportunity to explain a lot of those ideas, from the work of the future project to a broader audience. Update it for 2025 instead of 22, and, um, use the history as a way to tell that story, and as the historian on the work of the future project, people are often asking me, what is the deep history here? And I thought about that for a long time, as a lot of historians do. I kept getting drawn further and further back into the more fundamental moments, and landed on the Lunar Society well after the work of the future project was done. But that's where it felt like a lot of these decisions were made. And then you say, these guys who invented the Industrial revolution, they got a lot of things right. They changed the world and transformed society. It, you know, fed the masses. Um, and they missed two key things they got wrong environment and labor. Um, how do we fix that?

Wade Roush: Good questions. I mean, yeah, this is the paradox of modern existence, right? I mean, if they hadn't figured out that they could use the enormous sort of energy potential in fossil fuels to run steam engines and then locomotives and then everything else, we wouldn't have 8 billion people on this planet.

David Mindell: Exactly.

Wade Roush: You wouldn't want to reverse the clock because, like, here we are with all of these people having pretty interesting, comparatively long, you know, healthy lives. Maybe, uh, you know, maybe you can say they should have figured out faster that, uh, these fossil fuels were releasing carbon dioxide that was eventually going to raise the global temperature. Right. But interesting. Sort of like almost science fictional story there of, uh, society having to kind of pay the piper eventually for its exponential growth over two centuries.

David Mindell: Exactly. And I do try to kind of recast the steam engine as an invention that way, where, I mean, it's not a coincidence that it starts in mines, right? And particularly the early steam engines were used in coal mines because they consume so much coal that the only place you could actually use them economically was at a coal mine. And, um, a steam engine is not just a source of power. It is literally a a kind of match between the power embodied in the earth and, um, and human society. It is the moment that starting with the fossils that that that fossil fuels sort of were were harnessed and very much used to amplify the power available to human society. And those two things are very much related. And there's also a kind of, I call them chthonian machines, which is a great word that the Greeks used for the earth gods, like the nymphs and the other gods and goddesses who wandered around in the forest, as opposed to the Olympian gods who lived on the top of the mountains. And there's a sort of dark and mysterious aspect to that history as well, which is drawn out from Birmingham and a bunch of other ways, including one of its most famous sons, J.R.R. Tolkien.

[musical interlude]

Wade Roush: I'll continue my conversation with David Mundell in just a minute, but I want to pause the conversation here to share a very exciting update from the Hub and Spoke Audio Collective, of which soonish is a proud founding member. The news is that Hub and Spoke has a new member show called The Audio Podcast. It's the podcast counterpart to Audio Flux, a project founded by audio greats John Delore and Julie Shapiro to encourage bold storytelling through short form audio. Amy Pearl is the host of the podcast, and the trailer just dropped this week.

[Audio Flux Trailer]

Wade Roush: Here at Hub and Spoke. We're huge advocates of curiosity, compassion and excellent craft in independent audio, and so we couldn't be more thrilled to welcome John and Julie and Amy into our merry little band, and to do everything we can to help promote their show. You can learn more about Audio Flux at Audio Fluxes, and you can subscribe to the audio podcast wherever you get your podcasts. Okay, now back to my interview with MIT's David Mindell.

Wade Roush: I don't think we can have a conversation, you know, right now in May of 2025 and talk about the future of jobs and labor and who's making up the workforce without talking about artificial intelligence and and where it's going. And I know, of course, this was a theme in the work of the future project, but it feels like we're almost in a different world today than the one we were in when you started that in 2018, and maybe even from when you finished it in 2022. So I wonder, sort of like how closely you've been following the developments of the last couple of years In large language models and generative AI, and how you think about the real pace of innovation, because it seems like the first question everybody always asks is, how soon is this particular technology going to take my job? And people have been asking that question about every technology for the last 200 years, right. So in a way, are we just like falling into the same rut, or should we be asking a different question? Is this a familiar kind of catastrophism or a apocalyptic thinking that we need, that we need to see beyond?

David Mindell: So I do think there are aspects of these conversations that have recurring features about jobs and automation and mechanization, and and we can see those again throughout this history. Um, I also say in the book, like as a historian, the last person ever to say that there's nothing new under the sun. Um, but but thinking historically helps you understand what is new and what is fundamental. So these questions about jobs and our relationships to technology are fundamental. Um, AI and remote work and generative and language models, those are all new. And we still have to grapple with what they mean. And we don't know because we don't have too many precedents. Um, I also think a lot about the physical world and technology in the physical world. And, um, we're not yet at the point where AI has mastered that, those kinds of environments. And in fact, I just gave my last lecture. I taught real time in safety critical software in Aeroastro at MIT this, this spring. And the last lecture was about like, you know, where are we with AI in actually flying planes? Um, and the answer is very early there. Um, because we don't yet really know how to validate them, um, for safety and other things that people care about with flying airplanes and, um, a lot of people working on that, a lot of smart people working on that.

David Mindell: There's a lot of interest in being able to get there. Um, but, uh, the FAA still has, you know, really not certified any autonomous aircraft, uh, you know, any larger than a drone. And we're a long ways from flying on an unmanned airliner as a passenger. Not to say any of those things are impossible, but it's a good little measure of what is our confidence in the current AI vis a vis things that cost human lives. And many people who operate large systems like an automobile factory or almost any factory, um, have a similar concern about their own use of capital that we do about human lives in the sense of, I don't want to turn over my factory management to something if it could likely screw up even a day's worth of production. For me. None of that is to say that the AI that people are using today, they aren't Profound and important. And we are even as recently as last week, seeing the white collar and the coding workforces getting reduced, um, at companies, although those companies may also have been bloated already and they're just taking a good opportunity to, to, to get focused. I actually was in Las Vegas recently, and I was overwhelmed at the amount of labor that it takes to run that place. There are just people everywhere.

David Mindell: There's just thousands and thousands of human beings doing all these kinds of work. And, um, that work may all be automated one day, but it's not at the moment. None of that is a definitive answer to your question. But people are very good at finding new kinds of work, and one of the lessons from the work of the future project was, yes, new technologies displace people. Yes, we do more work with less labor. That's the definition of productivity, and it's almost the definition of technology. And that's very painful for the people whose work is displaced that way. Socially painful. Personally painful. And therefore, we just have to double down on making sure that we're constantly inventing new kinds of jobs and new kinds of industries that will take advantage of those people. So on the one hand, you say, you know, um, Microsoft just laid off 7000 people last week. And in the same breath, literally on the same page of the newspaper, you read, TSMC is trying to build semiconductors in Phoenix, and they can't find enough people to help them build semiconductors. So those are shifts. And we should encourage those shifts. And unfortunately, they're often not the same individual people who go from one to the other. But we want to make sure that overall, the the labor market is able to shift in those ways.

Wade Roush: I want to go at this one more time. I feel like people like you and me, who spend so much time thinking about the past and reading about the emergence of new technologies like steam power, for example. We we we see these long drawn out processes of adoption. And sometimes it's painful. And there's a lot of barriers to adoption as, as you know, and in the book you cite this 30 year rule, you say basically, you know, it's not a rule, it's just kind of an observation. But it often will take multiple decades for, uh, important industrial innovation to be fully integrated into the, the industry. It has the potential to transform. And and that was definitely true with with steam. It took probably more than 30 years, maybe 60 years depending on when you start the clock. Um, I'm just I guess I'm wondering whether you can allow for the possibility that, um, AI might be different. There might be a fundamental discontinuity here that breaks the 30 year rule, uh, because of the, um, just unexpected sort of power of the, um, of the statistics of these giant spreadsheets that really generative AI is not. Obviously, it's not intelligent, it's not aware, it's not conscious. But it is really good at really fast at doing math in ways that are unexpectedly powerful. That seems like a potential step change. So I'm just wondering whether, um, whether you can imagine a future that is very, very different from today in a, in a relatively short amount of time?

David Mindell: Well, I certainly think and I kind of say this in the book, like the the transformations in the software ecosystems have always been so much more rapid. Right. Um, and um, and those software ecosystems are very important in the world today. And they attract a lot of eyeballs and they, you know, sell a lot of products. And, and they're integrated into the economy. I, you know, I would love it to be able to say how the food arrives in your doorstep is going to be transformed by AI ten times more quickly, and it has transformed a little bit in the last ten years. Now it's a guy on a scooter instead of a guy driving a car. Um, very much enabled by digital technology. Um, uh, and I think the driverless car Odyssey, which is now, depending on how you count 15 years into its development cycle, much less deployment cycle, um, shows you that it's it takes it's slower when it comes to stuff that can kill you. And again, there may well be breakthroughs in understanding how to validate safety critical systems using AI. I hope there are. It's actually quite painful to do it now. Um, and I think that, uh, innovations in how we think about safety and risk are a really untapped area of research in a lot of ways.

David Mindell: And the electric power networks are undergoing some change, um, in response to the AI boom. And I'm involved with a little bit of that in various places. And, you know, wouldn't it all surprise me either, if I remember this kind of feeling from the late 90s when the internet was first taking off and it felt like every week something was suddenly done digitally on the internet. That hadn't been the week before, even though the internet had been around for a while. It all kind of congealed quickly. And, you know, that could happen in robotics. I've been in robotics for 40 years, and there's a lot of excitement over humanoid robots. I think we're only even beginning to understand what data needs to be collected to train those robots to do anything useful. I know we are only at the beginning of that. Collecting that data is not so straightforward because you have labor surveillance, privacy issues, and a lot of other kinds of issues around it. It's not as easy as what I bought on the internet last week, and a lot of smart people, a lot of capital going into that. But it's going to take a little time.

Wade Roush: You mentioned that we are actually pretty good here in the United States at R&D until maybe the last three months. I think you finished your book in sort of mid 2024. So before the election. There's a chapter in there about the Inflation Reduction Act and how it was starting to kind of lay the groundwork for a fundamental sort of government reinvestment in new styles of manufacturing and infrastructure and all that. Then we got Trump. So can I ask you to expand a little bit what what things are you most worried about? What things are you maybe less worried about when it comes to industrial policy? And do you see sort of headwinds and tailwinds for the kinds of transformation that you advocate?

David Mindell: Yeah. Um, I mean, that's a big question. And, um, a really important one. On the one hand, you know, I'm troubled by the the way that a lot of these cuts have been done and there's a lot of arbitrary mistakes and thoughtlessness behind it. On the other hand, I'm also troubled by the university R&D system. Just call it that, for for the moment has not done a good job at of of articulating really what it contributes to the country. They've sort of fallen back on the Vannevar Bush 1945 science, the endless frontier arguments, which we know. I mean, I actually believe those arguments, but they failed from a public policy politics point of view. Most Americans don't agree with that. You have to argue it in a different way. And you also can't and shouldn't claim that the system we have is the only possible one, and it's working at peak efficiency. And again, I wonder, gee, you know, um, I'm very sorry to see 500 people laid off from JPL last year. This was even before these cuts. But those people can make a huge difference in the in the space industry, um, by going into private industry. And it is entirely appropriate conversation for democracy to have about. Yes, we believe that we need to be funding both fundamental science and more applied science, but at what scale and toward what ends? Um, is is a legitimate conversation to have.

David Mindell: And again, I see around me all these very smart people going into R&D careers like and very advanced, sort of high powered PhD stuff. And, and yet I see the people at these big organizations who aren't able to do the adoption. And maybe we need a little rebalancing of the prestige reward incentive system for our smartest young people to share that there's, you know, a great deal of societal value, and hence there should be prestige to go into these positions where you are actually transforming the systems today, tomorrow, and not just 15, 20 years from now. And and that's where I think the moats around the universities are deeper than they should be. Nobody's really having that conversation right now, and I think it's one that needs to be had. We can't just go, oh, any cut is terrible. We can't we can't do things any different way. Right. We have to be willing to say, as a university, as an R&D system, like, how are we able to adapt to what's needed today and contribute the best way we can.

Wade Roush: That's great, but it feels like what's going on is just a wholesale and, like you said, kind of thoughtless turning down of the thermostat, which overall is a horrible idea. I'm sure you agree with me, right?

David Mindell: Yeah. I mean, I wouldn't call it wholesale in that I think there are things that have been exempted and they don't talk about it much. But if you look around, there are certain places getting attacked and other places getting attacked less. You know, the universities have very broad political footprint in this country. And they, um, they're not all being treated the same because not all of them make the same ideological points today. And there, again, I don't think the American public is as aware as it should be of the ways that the R&D system contributes to our daily lives. And that's our fault. You know, we've not communicated that either enough or more likely, the right way. But I'm not in charge.

Wade Roush: Let me wrap up with a question that I like, because it, in a way helps to wrap up a lot of the themes I've been exploring on this show for the last like eight years. So you say over and over again in the book, one of the parts I like the best is that technology is not some external force that just advances under its own steam, so to speak. It's that's an argument that goes by the jargon word technological determinism. It's just the idea that humans could stand aside and the technology would keep going. And it's just not true. You tell stories over and over in the book about personalities and people like Boulton and Watt and Priestley and Franklin and all of these people who had their own quirks and passions and interests and biases, and how those all help to shape the work they did. You know, I think you and I both believe that technological change is just the sum total of all the decisions we make about who to vote for, what to buy, how much time are we going to spend on on Netflix versus going out for a hike? You know, that kind of stuff. And I guess I wanted to close by maybe asking you whether you see that kind of thinking spreading in any way. Are you are you optimistic that this vibe, that this understanding that we can shape technology through our day to day decisions is filtering down and maybe helping people be a little less fatalistic about things like advancing artificial intelligence and the prospect of mass joblessness. Like we actually can do something about that. We could decide to regulate it, for example.

David Mindell: Yeah, well, I think you rightly tied it back to the AI conversation. Um, actually, when we started the work of the future project, the economists were saying, what is this technology going to do to us? Right. And, and over the course of the project, they flipped around and they began to see that actually we can shape technology in ways not I wouldn't say, you know, control or design because there are plenty of unexpected. But shaping is a good term because it suggests a more gentle, uh, influence. And now there are projects at MIT on shaping the future of AI from even economists. Right? They don't actually know very much about what levers to pull, but at least they're talking. Daron Acemoglu is talking a lot about that. Right. We can make AI one way. We can make it another way. Let's try to make it in ways that are consistent with our values. And to do that, we need to articulate our values. And I think that's actually very true at the company level. And I think many a lot of those decisions are not made in boardrooms or in Silicon Valley. A lot of those decisions are made by individual managers who choose to deploy technology in their organization, you know, thoughtfully or thoughtlessly. Right. And the like you say, the decisions that affect the outcomes are a lot lower in the system than most people think.

David Mindell: Ai is still a very human technology, right? It's all based on observations and data from human behaviors. And so we're influencing it already by those behaviors that we enact and the way those behaviors are deployed and how they're weighted. There's a lot of human decision making that goes into all of those models as they're created and deployed. So there's some work to do to just make that human decision making visible. Um, there's a ten year old example, but one of the first driverless cars, like, I looked at the configuration file for it and there were like 1200 parameters that some human had set. You know, how far from the next car do we consider to close to warm from a decision? Is it 1.2m or is it 0.9m? And that's still true in a lot of these systems. So I actually think at some level the the conspiracy of technological determinism is a bit of a ceding the battle to people who want it to look that way. And, and just as in political conversation, like the pushing back and the, the voices are important and they actually will shape the outcome, whether it's more regulation or more people, you know, creating different models of their own or even just being explicit about the values, those are all important outcomes.

Wade Roush: Well, David, thank you for continuing to highlight these voices, tell the stories of these people who are involved at the level of the gears and the code and the configuration files. It's fascinating stuff. And that's the that's really the joy of history of technology to me. So that's why I love I continue to love reading your books, and it's been such a pleasure knowing you all these years. And so thanks for joining me today.

David Mindell: Likewise, as you always know, I think you and other journalists play a critical role in these systems of public understandings of technology.

Wade Roush: That's what we're trying to do.

[musical interlude]

Wade Roush: That's it for this episode. And that's it for the podcast!

Soonish was written and produced by me, Wade Roush.

Our theme was composed by Graham Gordon Ramsay.

All the other music in this episode was created by Title Card Music and Sound in Boston.

You can find David Mindell's book, The New Lunar Society, at any major bookseller, and I'll include links to David's other books and projects in the show notes at soonishpodcast.org.

Before I go, I want to say thank you to all of the other producers at Hub & Spoke for being such amazing friends and colleagues over the years. You guys are really the best.

And I'd like to remind you to check out Hub & Spoke’ss newest member show, the Audio Flux Podcast.

And I also want to extend a deep thank you to everyone who has supported Soonish over the years on Patreon. Your donations to the show have really meant the world to me. I still support a bunch of cool projects as a contributor on Patreon, and it's been wonderful to see how many people agree with me that it's important to vote with your wallet to support your favorite independent creators. Thanks again.

Goodbye for now, and please stay subscribed to this feed to hear all about my new podcast which is coming...you guessed it...

Computer Voice: Soonish.

Soonish 5.17 David Mindell on What it Takes to Power an Industrial Revolution

5.17 | 10.24.25

Recommended Reading and Browsing

Notes

Full Transcript﻿

Full Transcript