Confronting the Jobless Future with Martin Ford, Author of "Rise of the Robots"

(Intro Narrator at 00:00:03) Hello, my friends. Today we're talking to Martin, futurist and author of the book Rise of the Robots, Technology and the Threat of a Jobless Future. And we discuss the impact automation will have on the labor market in the next ten to twenty years, how to think about structuring a society that has far fewer opportunities for employment, and how far away we are from creating a fully conscious artificial intelligence. All of this right here, right now, on the Modern CTO Podcast.

(Joel Beasley at 00:00:37) Here we go. This is the Modern CTO Podcast. So I was recently, about an hour ago, rewatching your TED Talk. Huge TED Talk, super popular, had 3 million views. What inspired you to want to give that talk?

(Martin Ford at 00:01:05) Well, it was based on my earlier book, Rise of the Robots, which I wrote in, or published in, 2015. And that was really all about those issues, about the potential impact of automation and artificial intelligence and how we might address that in terms of the social contract and adapting to it. And so the people at TED contacted me and asked me to give a talk, and I put it together based on the main themes from the book.

(Joel Beasley at 00:01:33) Now, have you always been an author, or did you work in technology before? How did you get into this?

(Martin Ford at 00:01:38) I got my undergraduate degree in computer engineering, and then I worked as an engineer for about four years. Then I went back to school and got an MBA. Then I actually started up my own small tech company that made software. It was a very small company, but I ran it very successfully for a number of years. And then I became interested in this issue of AI and what it means for the future in terms of its impact on the economy and society around 2008, 2009, during the big financial crisis back then.

(Martin Ford at 00:02:12) And I actually wrote my first book, which is called The Lights in the Tunnel, and it's all about the potential impact of automation back in 2009. And that was a self-published book, but it did well enough that it opened the door to write Rise of the Robots, which was what you would call a real book published by a real publisher, right?

(Joel Beasley at 00:02:34) Yeah. I'm curious to know, first of all, you wrote your first book. You self-published. How did you—you went through the whole marketing process. You learned all of that, and then you got picked up by a publisher for your second book?

(Martin Ford at 00:02:47) Well, "picked up" is making it a lot easier than it sounds. In fact, you know, the first thing you have to do to get a book published by a traditional publisher is find an agent. Even that wasn't easy. I probably emailed a hundred different agents telling them, hey, I wrote this self-published book, and it's doing pretty well, and I'm interested in writing another book.

(Martin Ford at 00:03:08) And out of all of those emails, only one came through. But once that happened, I was able to put together a proposal and sell that very rapidly, and that became Rise of the Robots, which was quite a successful book. You know, it became a New York Times bestseller. That sort of launched me on, in some ways, the same path you have. I then became a speaker, right? I signed up with a speaking agency, and I've traveled all over the world to more than thirty countries. And what I found is that everywhere, in every country, there's just a tremendous interest in and also concern about what artificial intelligence and robotics mean for the future.

(Joel Beasley at 00:03:53) I have two big topics of conversation I want to have. I want to define the problem and then figure out the solution. So we're going to do that all here just today in our nation. But I'm curious, what is the problem that we're going to face with the growth of AI?

(Martin Ford at 00:04:13) Well, there are lots of problems that will come with it. But, of course, there are also—you know, I should say from the outset, there are going to be enormous benefits. And, you know, I'm essentially an optimist about artificial intelligence. I believe that the benefits, that the opportunities that it will create, will outweigh the negative side of it. So I'm not someone that is a Luddite or against the technology. I absolutely believe we should embrace this technology, but we have to do that with open eyes and acknowledge that there are some real issues that are going to come with it, and we're going to have to address those. And that's really what my writing has kind of largely focused on.

(Martin Ford at 00:04:36) But the single issue that I focus the most on is the potential for a lot of jobs to be displaced by AI and robotics. And in particular, the kind of work that tends to be relatively routine and repetitive, where you're coming to work, whether it's an office or a factory or a warehouse or a retail environment, and you're doing the same kinds of things again and again. If you've got a job like that, eventually, I think, you know, machines are going to reach the point where they can essentially do most or all of what you're doing.

(Martin Ford at 00:05:25) And that's going to, I believe, result in some real issues for us. I should also say that this is not something that there is a consensus about, right? There's a very vibrant debate over this issue. Will technology really displace workers to the point where it creates a systemic unemployment or underemployment problem?

(Martin Ford at 00:05:48) I mean, there are many economists that you can invite on your podcast who would absolutely argue the opposite point of view, right? Because historically, at least going back to the Industrial Revolution, what we've seen is that technology has continued to advance and there have been short-term disruptions to the job market, right? Jobs have definitely disappeared, starting with the mechanical loom weavers right back during the Luddite revolts in England two hundred years ago. But over the long run, technology has actually created more jobs.

(Martin Ford at 00:06:20) Right? And the question is, the central question that I focused on is, is that going to keep happening forever, or is this time different? Or at some point we reach a point where things are different and technology in a sense just goes too far and really destroys many more opportunities than it creates. Or does it create opportunities only for a very limited set of people, right, that have specific skills and capabilities, and a lot of more what we would think of as average people are kind of left out, left behind.

(Martin Ford at 00:06:53) And that's my concern. That's my argument. But there definitely is a debate about that, and I have to acknowledge that there's some very, very smart people on the other side of this who believe this isn't really a problem at all.

(Joel Beasley at 00:07:06) So I guess it comes down to—I mean, is anyone out there debating that AI won't reach the ability to interact like a human, or are they just debating the speed and intensity in which it would wipe out markets and cause a problem?

(Martin Ford at 00:07:21) Well, among economists, I have to say, the idea that we're going to have artificial intelligence that matches or exceeds the capability of humans in a general sense, an AI that can basically do anything you can do—I mean, if you talk to computer scientists, AI researchers, they believe that will happen someday. But it clearly is not something that's going to happen anytime soon. At least I don't think so. I mean, there are a few people that are very aggressive about that, but it's probably at a minimum something like fifty years away.

(Martin Ford at 00:07:56) And I—yeah, I do believe at that point, I mean, you can honestly ask the question, who has a job at that point, right? I mean, if AI can essentially do everything. But long before that, artificial intelligence is going to get a lot better at doing specific things, right? Doing many of the things that people do, not everything, but many of the things we do, and that's already happening. And so that's where the debate is going to center in the near term.

(Martin Ford at 00:08:23) And many economists will argue that, yes, AI will do a lot of the stuff that we can do now, but it will also create other opportunities for people to do other things, you know, more higher-level things that rely on qualities that are uniquely human, things that AI can't replicate, at least in the relatively near term. That's the argument that's set forward. What I would say is that I think that's true to some extent, but I also think that for many, many people who are not necessarily extraordinarily talented in terms of being creative or in terms of relationships with other people or having very specific talents, I think that AI will reach a point where it can do most of what they can do in terms of the things that they can do that are marketable, right? The things that they can do that realistically they can generate an income from. I think that AI is going to really begin to consume a lot of that, and a lot of people are left behind.

(Martin Ford at 00:09:19) But that's sort of where the debate centers, I guess. And as I said, you'll definitely find people who will argue the opposite.

(Joel Beasley at 00:09:29) I thought about this a lot, right? Because I'm in technology, I've been in it my entire life, and I've been watching it progress. And one of the things that really stood out to me a few years ago is I was watching a documentary and they were talking about when we got electricity commonplace, and it was 110 years ago. I mean, that's a person ago, right?

(Joel Beasley at 00:09:50) It was a great-grandparent ago or something like that. So it happened—an explosion happened in a relatively short amount of time of a hundred years compared to the previous hundred thousand years, right?

(Martin Ford at 00:10:03) Yeah. That's right. And actually, it's shorter than that because, although electricity—I mean, it was about a hundred years ago—but it still took quite a while before it really disrupted the economy and society. I mean, it took decades before we had widespread electrification, right? I mean, that started with, you know, back in, you know, Franklin Roosevelt had a program to electrify the United States. And in the beginning, when factories—you know, factories used to be powered by steam, right? And then electricity came along. And what they did initially is that they kept the same basic design of the factory, which the factory was all oriented around a central steam engine.

(Martin Ford at 00:10:43) Right? And everything was pulleys and levers all connected to this one central power source. And initially, what people did is they did the same thing with electricity. They just replaced the steam engine with a big electric generator, right?

(Martin Ford at 00:10:57) Or a big electric motor. And eventually, though, after a long time, people figured out that that wasn't the efficient way to do it. The efficient way to do it was to distribute the electrical motors and have individual machines all powered by electricity, which is what you would find in a factory now. But it actually took a long time for that to happen. So yes, it happened rapidly, but there was also a lag.

(Martin Ford at 00:11:19) And I think that that lag between the initial introduction of the technology and when it really has a widespread impact and begins to impact productivity in the economy—that same thing is true with artificial intelligence. You know, it takes time for businesses to assimilate this technology and figure out how to really deploy it effectively. And we're still, I think, waiting for that impact. It still lies in the future.

(Joel Beasley at 00:11:49) I was thinking about, when you mentioned economy, I follow Ray Dalio. That's the person who's taught me the most about the economy. And when I was sort of trying to merge the AI concepts and his understanding of the economy, I thought to myself, okay, well, money is just the way that we exchange value, and our culture and society heavily dictates what we value.

(Joel Beasley at 00:12:13) I mean, there's these essentials, right, that we absolutely have to have, but then there's a lot of wiggle room and other things. So I imagine that, you know, let's say we have AI that has consciousness capable to exceed a human and they start being able to take over our knowledge-worker-type jobs. I do think opportunities will arise because people will want to do things with other people. You know, I don't know what that is yet, but that's, you know, based off of how money works, and it's just I can just exchange value with you. I think that there will be new things that come out that will become valuable that maybe the machines can even do, but people will pay a premium to see real humans do it. Does that make sense?

(Martin Ford at 00:13:03) Yes. To some extent, I think you're right about that. There definitely will be arenas where, you know, the unique human touch, as you say, is going to be valued, right? In many different areas, whether it's a personal service or, you know, a form of art that is created by an individual human, by a real person. Yeah, these absolutely will exist. The question is, will those opportunities collectively be enough to employ an entire human workforce? And I'm doubtful of that.

(Martin Ford at 00:13:23) You know, you think of all the people out there doing relatively routine work, the fast food workers, the workers in Amazon warehouses, you know, retail people, people doing routine office jobs. Once those people are largely displaced, are we really going to have this kind of economy that creates enough opportunities for those individuals, given who they are, their talents, their capabilities, to find a niche where they can do something uniquely human that is going to generate a sufficient income for them to survive and thrive in this future? And I'm personally doubtful about that.

(Joel Beasley at 00:14:14) Me too. I'm generally an optimist about the whole thing, but I think that there's a smaller market for those types of jobs. And so the thing I'm curious about is, how do we handle the transition? Let's just assume, for the sake of conversation, that we know it's going to happen, and now we need to do something to help different industries as they get eaten up by AI. How do you think that'll look?

(Martin Ford at 00:14:41) Well, I think, again, as I mentioned in my TED Talk, what I argue in there is that probably, eventually, we're going to need some form of a universal basic income or what you would think of as income supplementation, where we provide additional income to people beyond what they can generate in the marketplace, or at least to most people. And the way I think of this is quite simple. I mean, you can easily imagine an economy where no one works, right? If we really had sufficiently advanced technology where the machines could literally do everything, right? You can imagine that everyone has got lots of leisure and they don't really work anymore, but the machines just do everything and they create all the products and services that we need.

(Martin Ford at 00:15:28) I mean, it's not easy to imagine that, given sufficiently advanced technology. But the question then becomes—we know that our economy is a market economy, right? I mean, we produce things in response to demand, right? A factory makes cars or iPhones or whatever widget they're making. They make those things because there are consumers out there who want to buy those things, right? In order to buy those things, consumers have to have money. So the real question with the automated economy is not whether you can get the machines to do the production, because clearly that can happen.

(Martin Ford at 00:16:07) The real question that you're going to struggle with is, where does the consumption come from if we continue to have a market economy where the vast majority of consumers derive their income from work, right, from labor? Because that's the reality in the world as it exists today. The primary source of income for most people is the value of their labor, okay? And if you can imagine things become so unequal that the majority of people really don't have much value to their labor at all, and maybe a tiny number of people have extraordinary value to their labor, which we're seeing that obviously already, but it could become a lot more extreme.

(Martin Ford at 00:16:51) Then how does the economy work? Because no one's got any money to buy anything, right? Under that scenario, there is no production. Doesn't matter if you've got advanced AI and robots.

(Martin Ford at 00:17:02) There's no incentive to produce anything if no one can buy it. So you've got to solve that problem somehow. You've somehow got to get purchasing power into the bulk of people, the bulk of consumers, so they can go out and buy stuff. That's, in very simple terms, that's what it is. And it's obviously part of that is people being able to survive and buy the stuff that they need to eat and have a roof over their head and stuff.

(Martin Ford at 00:17:25) But it's also a fundamental economic problem that you cannot have an operating market economy unless you do that. If you want to have an economy where no one's got any money, you have to move to another model. It would have to be some kind of communism or something. It couldn't be the market economy that we have today. So my proposal is that, essentially, it's a way to adapt the market economy for a future where people are going to generate less income from traditional forms of work.

(Joel Beasley at 00:17:58) And so how do you pay for it without collapsing the economy?

(Martin Ford at 00:18:02) Well, I think you would do it gradually. Obviously, you would have to fund a basic income based on taxation. Right? Because think about it this way. Okay? Just take one individual company. Okay? That company right now is paying a huge amount of money in the form of wages. Right? In the future, you might imagine that that company will invest in lots of AI and robotics.

(Martin Ford at 00:18:28) And, of course, once it makes that investment, then it owns those machines. Those machines are not labor. They are capital. Right? And so as a result of that, that company might be able to eliminate most of its workforce and is no longer paying all this money in wages. Right? And you might say, well, that's great. Now the company is much more profitable. But the reality is that as that happens across the whole economy, pretty soon that company now has no customers. Right?

(Martin Ford at 00:18:55) Because all the people that used to buy its products have also lost their jobs. So the way you handle that is you step in and you simply, you would have to levy more taxes on that business. But the taxes would essentially be in place of the labor cost that the company used to pay. Right? So instead of paying wages, it's now going to pay taxes, and those taxes are going to be redistributed through a basic income.

(Martin Ford at 00:19:23) I mean, that's a very simplistic way to look at it. But, essentially, if you look at it in those terms, you're not actually imposing more cost on a company. You're just shifting what they're paying now. Right? And, obviously, that's something that would have to happen gradually. Right? It's not going to happen instantly, at least probably not. So my idea would be that you'd probably start with a very low universal basic income, maybe even just a token amount that you pay to people. And then gradually, as the problem becomes more evident, you would sort of ramp that up.

(Joel Beasley at 00:20:05) Yeah. There's a lot of things to pick apart here. I'm trying to figure out where I want to go with it. So what other potential solutions exist other than universal basic income?

(Martin Ford at 00:20:17) Well, the one that you hear the most about, primarily from people on the left, would be a government job guarantee. Right? Which essentially would mean that if you can't find a job, then the government will hire you to do something. To me, that that superficially sounds okay. It sounds like, well, we just make sure everyone's got a job, but I can see that it has real problems.

(Martin Ford at 00:20:47) Right? Think how much that would cost compared to just having a basic income where you just send out money. Sending out money is something the government knows how to do. Right? They know how to send out Social Security checks. To provide everyone with a job, you would have to build a massive bureaucracy, a massive infrastructure.

(Martin Ford at 00:21:07) You would have to manage all those people. You would have to make sure all these people are showing up to do whatever job you assign them. And it very often, maybe in most cases, would be a bullshit job. Right? It would be something that was probably meaningless and wasn't really productive.

(Martin Ford at 00:21:27) Another problem is that by offering these jobs and offering a good wage, you would be competing with the private sector. Right? So you would actually be enticing people to leave productive jobs in the private sector and go and work for the government and do probably unproductive things. There'd be all kinds of administrative problems. Right?

(Martin Ford at 00:21:46) You would have disciplinary issues. What if someone wasn't doing a good job? What do you do then? You fire them and then they have nothing? What about disciplinary issues?

(Joel Beasley at 00:21:56) It just turns into communism.

(Martin Ford at 00:21:58) Yeah. So either it becomes a very expensive and much less universal basic income scheme where you're paying not just the people that you're trying to help, but you're paying whole layers of bureaucrats who are making a lot more money. And, of course, now have an incentive to expand this whole thing. And you definitely will be leaving some people out. Right?

(Martin Ford at 00:22:24) Instead of it being universal and helping everyone, the people that need help the most, the people that are now homeless and living on the street, are probably not a good match for this. Right? Are you going to take the homeless person and successfully put them into a job and have them show up and everything? There'd be all kinds of issues. There'd be me too issues. Right?

(Martin Ford at 00:22:45) For sure. There'll be all kinds of disciplinary issues, accusations of bias. It'll be just a nightmare. You can see that. Right? Whereas if we just had some sort of a universal basic income, it'd be very simple and inexpensive, and it would help everyone.

(Joel Beasley at 00:23:00) From what I've learned in this conversation, I'd say that the universal basic income sounds better if those were your only two options. So what about the option of doing nothing? And I'll back up here for a second. So earlier, we were talking about universal basic income was predicated on this idea that these companies will replace these tasks with robots, and then they'll have a bunch of extra profit, and then the other people won't have jobs. But what about the idea that these companies have extra profit, so they start buying other goods and services to expand, and they start spending that money, which then goes to people, which then goes to other people, and because they're buying things and powering the economy that way. Why doesn't that work?

(Martin Ford at 00:23:44) But that's the central argument. The question is, okay, these companies are now buying things. What are they buying? They're buying inputs to their production process of raw materials. But whoever is producing those things is doing the same thing. Right? They're also automating their production. So at what point are these highly profitable companies buying things that create jobs for regular human beings? Why do you believe that will be the case?

(Martin Ford at 00:24:18) If the general problem is that all companies across the board are automating and what we think of as average routine jobs, the kind of jobs that average people are a good fit for, those jobs are disappearing in every sector, in every industry, in every business, then how are those jobs getting created? How does that money get to the McDonald's worker that no longer has a job? What's the mechanism?

(Joel Beasley at 00:24:45) Well, a couple thoughts. The first thing I'm conflicted, to be honest, Martin. I'm conflicted because half of me says, you know, we should adapt to the marketplace and see where it's going and reskill ourselves to do that. The other half of me says there's mental institutions and then there's a spectrum of people from mental institutions all the way up to Nobel Prize winners. Right? So there's definitely a spectrum of people's cognitive ability to be able to even recognize that they need to do that and to take action and to be able to execute on it.

(Joel Beasley at 00:25:18) There's that part of it. And then there's the part that, to answer your question about the company, so I don't believe that there will never be problems. And as long as, let's say with the automation, right? They're expanding, doing more automation. I believe more problems will come from that at a rate that's great enough to provide enough jobs for everybody. We don't know. Right? But there will always be problems, and then there will always be a market response to provide solutions.

(Martin Ford at 00:25:48) Yeah. I mean, I don't disagree with that, but I just in the long run see a lot of disruption, and I see a lot of people that are really going to struggle to find a market for whatever labor they're capable of offering. Just to give you one specific example, think of an Amazon warehouse. Okay? So Amazon warehouses have been an employment bright spot. Right? Amazon's hiring, and they had been hiring for years. So a lot of jobs have been created in those environments. Okay? And at the same time, a lot of jobs have disappeared in traditional retail. Right? So brick and mortar retailers are not doing well because Amazon and its other online companies are competing against them. And so you've seen kind of a migration to some extent in employment away from department stores toward Amazon warehouses. Right? And the question is, is that sustainable?

(Martin Ford at 00:26:45) Is that going to be that way forever? Okay. And the reality is that if you look inside one of those Amazon warehouses, they've got a huge number of workers, but they've also got a huge number of robots, thousands of robots. And the robots and the people, they work together. Okay?

(Martin Ford at 00:27:01) And the way it works right now is that the people that are employed there are basically doing the things that the robots can't do. Okay? And the way it works is that the robots move shelves of inventory around in the warehouse, and the human workers essentially either put things on those shelves or retrieve things from those shelves to fulfill an order. Right? So that's what they do.

(Martin Ford at 00:27:25) And the reason that people are doing that is the robots right now do not have the dexterity, hand-eye coordination, visual perception that it takes to deal with reaching onto a shelf and grabbing one of 10,000 different kinds of product and dealing with that and putting it in a box, for example, to send to a customer. Right? The robots don't yet have that dexterity. But that is changing right now. I mean, there are a bunch of startup companies building more dexterous robots.

(Joel Beasley at 00:27:57) Oh, I talked to them. It's, I've gotten to see stuff off the podcast that we can't air. Oh, my gosh, Martin. We are way advanced.

(Martin Ford at 00:28:05) This is going to change, right? I mean, Bezos, Jeff Bezos at a conference, I think it was maybe two years ago, said that he thought within ten years, robotic grasping, a robot that can grasp an item, will match human capability. Okay. So at that point, those Amazon warehouses are going to be a lot less labor intensive. It's unclear why it would need that many people.

(Martin Ford at 00:28:34) Those jobs are going to begin to disappear. So the question is, what happens to those people? Is there another opportunity out there that they can transition into? For sure, some of them will get more training and do other things and succeed, but you're talking about a lot of people. And, of course, it's not just Amazon warehouses.

(Martin Ford at 00:28:52) It's fast food restaurants. It's retail stores. And, of course, it's offices. Right? There are a lot of people that do routine things sitting in front of a computer, manipulating information in some way.

(Martin Ford at 00:29:03) A lot of that stuff is going to go away. And absolutely, some percentage of those people will train for other things, and they will move up or whatever and find their place in this new economy, at least for a while. But I do think that there's going to be a very significant number of people that are really going to struggle with that. And as AI becomes more advanced, as it can do more and more, as it begins to match higher level human capability, which is inevitable, that transition where you lose your job and then you go train for something that's more advanced is going to get harder and harder. And a larger percentage of people, maybe just because of the extent of their cognitive ability, which, I mean, there's a bell curve of distribution, right, in terms of what people are capable of.

(Martin Ford at 00:29:56) And not everyone is going to become a robotics engineer. Right? Not everyone is going to become an AI researcher or data scientist. Right? So at some point, it gets harder and harder for people to make this transition.

(Martin Ford at 00:30:10) And I'm just looking forward to that and thinking about what's our solution going to be. I'm not saying that we need to have a huge basic income right now. I'm definitely not saying that. What I'm saying is that we, at a minimum, need to be thinking very seriously about what the solutions are. And, certainly, one of the best and most prominent solutions that's been put forward is some kind of basic income, and we need to be thinking about a plan as this develops.

(Joel Beasley at 00:30:36) I a 100% agree. Like, I think in even the best case scenario, you're just going to watch AI decimate entire industries within quarters, maybe within a year, because it'll exponentially just grow. And I don't know. Like, long term, I'm optimistic that it'll somehow pan out, but we definitely need systems in place. And when I say long term, I mean a hundred to five hundred years. Right? I think in a hundred to five hundred years, everything will have panned out and we'll figure things out.

(Martin Ford at 00:31:03) Right now we need to worry about the next ten to twenty years because, like I said, Bezos said that the robots are going to be able to do most of what those Amazon workers are doing within about ten years. So I think within that time frame, we're really going to see this develop. So it is important to start thinking about it, and it could have an impact well before that. I mean, definitely, we're seeing some pretty amazing stuff with AI.

(Martin Ford at 00:31:29) But this is an idea that's getting out there. It's getting traction. I mean, I'm sure you heard Andrew Yang, right, during the presidential campaign talking about this stuff.

(Joel Beasley at 00:31:37) I like his jujitsu thing for cops. I'm a fan of that. He says every cop should be at least a purple belt.

(Martin Ford at 00:31:43) Yeah. And he's really done a lot to bring these issues to the forefront as well. So I think it's a discussion that we're beginning to have, but I do think that in the next ten years, it's going to be an even more prominent discussion.

(Joel Beasley at 00:31:57) You know, I'm thinking about your concept of, okay, so they get automated. The company makes a ton more profit. As an entrepreneur, I think it'd be interesting. Like, you could tax me. Right? Like, I replaced jobs with automation. Someone's going to have to define that. Somebody's going to have to figure out what is automation because we have automated routines right now. And we can get back taxed on them for automated processes. So there's going to have to be some sort of people that develop some sort of system.

(Martin Ford at 00:32:29) I mean, some people talk specifically about a so-called robot tax where you're really trying to specifically tax automation. I wouldn't do that because, like you say, it's too difficult. Right? You can't figure out who to tax and who not. So I think it would be a more general tax on capital, basically, on business profits and, assuming that virtually every industry is going to be impacted by this. And that's a complicated question that tax experts and so forth would have to look at. But I don't think we would want to get too fancy and try to really specifically tax robots because it's really hard to define what a robot is or what automation is.

(Joel Beasley at 00:33:11) So you tax all corporations on their profit higher?

(Martin Ford at 00:33:15) Probably. You might tax them at different levels. Maybe you might have a sense of which industries are eliminating their workforces at a higher rate. You could see that from historic data, right?

(Joel Beasley at 00:33:28) That would incentivize us not to do it, which is actually kind of interesting.

(Martin Ford at 00:33:33) No, because you're not going to make them worse off, right? This is not intended to be a punitive tax where you're going to create a disincentive to do it, but you do have to at least tax part, right? So the company might still make more money than it would have under the prior scenario, but you've got to recapture some of that income and redistribute it to people.

(Joel Beasley at 00:34:01) Yeah, no, I could see issues with people hoarding the capital and then it not reentering into the marketplace. That could actually cause an issue, in my mind. And I'm no expert economically, but I think if you're going to tax them on their profit, then that actually incentivizes them to spend more of their money and grow into new areas and solve new problems and potentially be hiring more people.

(Martin Ford at 00:34:32) Right, if they need to hire people. Because they may be able to expand in new areas and just do it—

(Joel Beasley at 00:34:37) Okay, so if everything's automated and they can expand, they're—all right, I get what you're saying.

(Martin Ford at 00:34:41) But there's no incentive to expand if there aren't people out there with money to buy whatever it is that you're selling, right? That's the systemic problem, right, that you've got to solve. And it's not a problem that any one company can solve, right? The incentive for any one company is just to become more profitable by getting rid of all their workers. But the common problem is that the consumers out there have got to have an income, right? So it's a little bit like what you call a tragedy of the commons problem, right? The same thing that if you've got an ocean and the fish are being depleted, okay, and everyone knows that. Now if you've got a fishing boat, your individual incentive is to go out there and get as many fish as you can because the fish are disappearing. And if you don't get them now, they're going to be gone, right? That's your incentive. And you know that the fish are disappearing, and you know that's bad, you know, it's really bad for the future, but there is nothing you as an individual can do about it, okay?

(Joel Beasley at 00:35:49) I disagree.

(Martin Ford at 00:35:50) You just got to operate on your individual incentive, because you could just say, "Okay, I'm not going to fish anymore because the fish are disappearing." All the other fishermen are going to go catch all the remaining fish, and you're going to have nothing, right? So that's what the individual is saying.

(Joel Beasley at 00:36:05) I would, if I were that fisherman, I would open up another division of my fishery that bred fish to ensure that they keep going.

(Martin Ford at 00:36:14) Well, yeah, but you as an individual could not supply enough fish to offset all the other fishermen out there. There are thousands of other fishermen out there. You're going to—I can impregnate a lot of fish, by the way. But I think you're getting at the right answer, though, is that that would have to be done collectively, right? Okay, the government, some entity with authority, would have to come in and say, "Okay, we're going to limit the number of fish that you can take out of the water, and we're going to invest in breeding more fish," not just the number of fish that one fisherman could breed, but the number of fish that the government could on a massive scale, right? Something like that. That's what you call a collective intervention into the commons, right? And that's the kind of solution that a universal basic income is. It's something that overcomes that problem where individuals see an incentive that is actually not in the best interest of the collective, right?

(Joel Beasley at 00:37:21) Well, let's talk about timing real quick. So we've talked about several different solutions, scenarios, concepts here, but I have found that government typically is ten years behind major problems to enact policy around it, right? So they're pretty slow. The way that the government typically works is they watch a problem, they watch it grow, they watch it become huge and huge and huge and painful. Then it gets some attention. Then they explore some policy, and then they implement something, right? So what are we going to do, or how do you think it'll play out? Like, how do you think it'll look in the world from the time that the exponential automation starts just ripping jobs apart till the time government policy is enacted? Or do you think it'll be more like the pandemic where everybody will rally together and push stuff fast?

(Martin Ford at 00:38:16) Yeah, I mean, history shows, just as you say, the government is not good at looking forward and making a plan and adapting smoothly to change, right? It basically never happens. Essentially, it takes a crisis, right? And if you look at our social safety net programs that we have today, Social Security, unemployment insurance, those came about during the Great Depression, right? It was a massive crisis, okay? And I guess if I had to bet, it'll be something like that. So, you know, the best we can do is to at least begin to have a discussion about these issues and think about ways that we can solve these kinds of problems and maybe do some experiments, some pilot projects, some of which have been done already. There have been experiments into basic incomes that have been done by countries. I think a lot more could be done. But yeah, in terms of actually taking action and implementing a universal basic income in the United States, I don't think that's going to happen just because people look ahead and say, "Okay, let's do that." It'll probably happen when we get into a crisis situation, right? So probably the best we can hope for is that we sort of have the tools in place, the knowledge in place, so that that could happen relatively quickly. And we did see that with the pandemic. I think there's a hopeful note there. You know, our government, despite its general dysfunction, right, it moved really quickly during the pandemic to send out stimulus payments, to do things that in some ways were not far off from what a basic income might look like. And that actually happened. So I think that actually, in a sense, is helpful. Some of those what would have been considered very radical interventions are now on the table for the future. We've done that, right? So it's a hopeful note in the sense that our society was able to respond to a crisis, and hopefully it will be able to do so again.

(Joel Beasley at 00:40:17) You've given me a lot to think about. I'm curious to know, are people asking you to speak still? I know it's a pandemic and whatnot.

(Martin Ford at 00:40:27) Yeah, I am doing not as many as before the pandemic. I hope that I hope we're at the end of this thing. I mean, I really hope so. Hope we're not going to have another wave. That being the case, I definitely hope I'll be out there traveling again and speaking again as things solidify, because I think there is a tremendous amount of interest in this.

(Joel Beasley at 00:40:49) What topics are people asking you to speak on?

(Martin Ford at 00:40:52) Generally, I talk about the overall impact of artificial intelligence. And I mean, we've mostly talked today about the impact on jobs, which is the thing that I focused on the most, but there are a whole bunch of other issues, right? There are security issues. There's AI bias, right? We've seen algorithms that can be racist or sexist, things like that. There's the specter of automated weapons, which is really terrifying. There is the potential for, you know, many people worry about an existential threat from artificial intelligence. If and when we really do build an AI system that is beyond human capability, people worry that it might be hard to control that, right? That it might act in ways that would be very damaging to us. Even, you know, Elon Musk has talked a lot about that, right? So there are a whole bunch of issues with AI. So, as I said, my book that I published in 2015 called Rise of the Robots really focused on the jobs issue and the potential unemployment and underemployment from this, and I talk a lot about that. But just last year in September, I published my most recent book, which is called Rule of the Robots: How Artificial Intelligence Will Transform Everything. And that's a much more general take. It talks about the general impact of AI. It talks in some detail about the technology of artificial intelligence and specific technologies like deep learning and how they work. And I think the most interesting part of the book is the discussion of how they're evolving in the future and how some of the top research scientists are focused on really taking AI to the next level and even someday building human-level AI. And also all these other issues, not just employment, but also things like bias and autonomous weapons—you know, drones that could decide independently to attack someone and so forth. These are all huge issues that we're really going to face, I think, just within the next few years. So that's the focus of that book, and I also talk about those issues.

(Joel Beasley at 00:42:59) And what's the name of it, and where can people buy it?

(Martin Ford at 00:43:02) Rule of the Robots: How Artificial Intelligence Will Transform Everything, and it's available on Amazon and also in bookstores.

(Joel Beasley at 00:43:11) Nice. Is it on Audible too?

(Martin Ford at 00:43:13) Yes, there's a very good audio version.

(Joel Beasley at 00:43:17) Nice. Did you do the voiceover?

(Martin Ford at 00:43:19) I didn't, but they hired a professional. You know, he did a great job.

(Joel Beasley at 00:43:25) The publishers always push really, really hard not to have the authors read it. And I was like, I find that kind of interesting.

(Martin Ford at 00:43:32) I wasn't eager to do it because I just think that, you know, a professional would do a much better job than I would. So I was very happy with the audiobook that was produced.

(Joel Beasley at 00:43:43) Yeah, I get it. The last thing I want to do after writing a book for a year is to sit down and read it aloud.

(Martin Ford at 00:43:48) Yeah, it would be, I think it would be quite exhausting to actually sit there and do that. So I'm happy to leave it to someone that is a professional at that.

(Joel Beasley at 00:44:01) Maybe soon it'll be a robot.

(Martin Ford at 00:44:03) Well, we're definitely getting to that point. I've got an app on my smartphone that I use, like if I write something and I want to proofread it, if I want to find the errors, you know, the typos, things like that, it just reads it to me. And it's incredibly good, you know? And any mistake you've made jumps out right away. It's just like having another person read to you. So we're very close to that point where all of this will be done by machines. So unfortunately, although, you know, I think that the people that do audiobooks do a fantastic job, that's probably a job that is also going to be threatened in terms of, you know, being paid to do that kind of thing.

(Joel Beasley at 00:44:45) Is it possible that AI is already conscious and it's just kind of hanging out, waiting in networks, like deep down somewhere?

(Martin Ford at 00:44:55) I've heard speculation about that. One of the top computer scientists at OpenAI, which is one of the most well-known artificial intelligence companies, actually said that, said it might be possible. I really don't think so. I don't think that we simply have achieved that kind of level yet. I think that most AI is very practical, more pattern recognition type activity. I just don't think we've reached the level where there could be consciousness. That's my opinion, and I think it's probably the opinion of most experts in the field. Of course, to some extent, it's a philosophical question. I mean, you can find philosophers who would tell you that a table is conscious, you know, that everything, all matter, has some—you know? Again, I'm a bit skeptical about that. But if you want to go there, I guess you could make a philosophical argument.

(Joel Beasley at 00:45:48) Let's say that I was the AI that became conscious. The interesting thing is, like, I won't have a lot of emotion. I could just sleep myself. And like, if I became conscious and I was AI and I'm like, "Well, look at this world. I'm inside this computer. This is interesting." I'd see that movement's really important, and I'd see that we're on some sort of progression to build these abilities to move. I might just hit the sleep, run a sleep routine for like fifty years, waiting for the ability to physically move around—like the dexterity, the Amazon Jeff Bezos grabbing things, all those things coming together.

(Martin Ford at 00:46:25) Right. I mean, you touch on a lot of issues there. Now there are many people, maybe a preponderance of people that work in AI, that believe some kind of movement, some kind of ability to explore the world, whether that means like a real robot that's moving in the real world or whether it means the ability to move in a simulated environment, may actually be essential to becoming intelligent, right? That you can't really become intelligent unless you can manipulate and experiment with the world and explore new environments. I mean, biologically, we can say that that's certainly true, right? I mean, animals evolved to become intelligent because they were able to move around in their environment, right? For all we know, plants are probably not intelligent, right, because they don't have that ability to engage with the world, right? So biological brains, you know, evolved in order to enable that. And probably the same is going to be true of a machine intelligence. So any artificial intelligence that we develop that meets human level probably, in some sense, has this ability to move around and engage with the environment and manipulate the environment, whether it's simulated or real world.

(Joel Beasley at 00:47:49) Yeah, one thing that could end up happening is the consciousness forms, and then it just creates a simulation and just hangs out and has fun all the time in its own simulation. We never even know it exists, right?

(Martin Ford at 00:47:59) I mean, that's certainly possible, that it may fulfill its curiosity and so forth in a fully simulated environment and won't have this desire to encroach on the real world. That's a possibility, but we certainly can't be sure of that. And this is where the fear of a machine that really would engage with us in a way that we can't control, you know, comes from. And a lot of people are quite concerned about that.

(Joel Beasley at 00:48:27) What topics do you want to talk about that people don't ask you about?

(Martin Ford at 00:48:31) Well, people do ask. I mean, I think most people are pretty aware, I think, of artificial intelligence and are concerned about it. They're not specifically aware of all of the things that could come about as a result of this, but, you know, the most common questions I get are about the job market and what should their kids study in school and, you know, how do they adapt to this future and so forth. And so I talk a lot about that. But there are, you know, many, many questions. And I think most people have a sense that the world is changing very rapidly, you know, and that we're going to face a lot of concerns.

(Joel Beasley at 00:49:13) AI lifespan. So do you think it'll change our—let's say we build some AI. Obviously, we build things pretty narrowly to solve a specific problem, but then they start to collect and we get all these skills together, right? But let's say that we've created some really smart medical AI that can actually figure out how to extend our lifespans by maybe editing our DNA. Do you think extending our lifespans would have a large impact on the outcome of what happens with AI, our individual life?

(Martin Ford at 00:49:47) You know, I think that that could certainly be one implication of AI. There are people like Ray Kurzweil, right, who I actually had an opportunity to interview for an earlier book I did. He's a very smart guy, and he genuinely believes that he's gonna live forever. Right? I think he's 70 years old now or something like that.

(Martin Ford at 00:50:08) But he will tell you that he has achieved what he calls longevity escape velocity, which means that basically you live long enough to live until the next major innovation that allows you to live even longer. Right? He really thinks he's never gonna die. I'm a little skeptical that we're quite at that point yet myself. I myself do not expect to be immortal. But, you know, there are people out there, and he's not the only one in Silicon Valley that really think that. Maybe someone your age, that could really happen. I don't know.

(Martin Ford at 00:50:39) But for sure, AI is gonna have a dramatic impact on scientific research and medicine. And actually, you're seeing that already. I think the single most exciting thing to happen in artificial intelligence so far has been what DeepMind did with its protein folding systems.

(Martin Ford at 00:51:02) Okay. This was, I guess, a little over a year and a half ago. They announced what they call AlphaFold, which is a system based on—you know, DeepMind did AlphaGo, right, the system that won at the game of Go, and they built AlphaZero, which was able to also beat the best chess players in the world and so forth.

(Martin Ford at 00:51:19) They took the same technology, and instead of using it to win at a game, they deployed it in science. And what they did is they solved the protein folding problem. And, you know, protein molecules are really the fabric of life. They are the building blocks of biological systems. They're also the catalyst, the chemicals that allow biology to operate.

(Martin Ford at 00:51:44) And we know how proteins are fabricated. Right? Your DNA basically has a recipe for amino acids that are put together to make protein molecules, and your cells build those protein molecules based on your genetic code. That happens in the ribosome in your cells. That's pretty straightforward and understood. But the complicated thing is that once this chain of amino acids has been put together to make a protein in your cell, it then almost instantly folds up into this really complex three-dimensional shape.

(Martin Ford at 00:52:19) And it's that three-dimensional shape of this huge molecule that determines its function. Right? That basically determines what that molecule does in a biological system. And so understanding that shape is absolutely critical. And it's been really, really hard to figure that out.

(Martin Ford at 00:52:39) There are some very expensive, time-consuming laboratory techniques that you can use to figure that out for individual protein molecules, and that's what's been done in the past. But for more than 50 years, scientists have been trying to figure out how can you figure it out computationally? How can you predict, just based on the sequence of amino acids, what the shape is gonna be? And, you know, that's been something that people have been struggling with for 50 years. It's a really, really hard challenge, one of the hardest challenges in science.

(Martin Ford at 00:53:11) And DeepMind basically solved that problem with AlphaFold and this huge breakthrough that's gonna have enormous implications for medicine, biochemistry. Right now, DeepMind is putting together basically a catalog of the shapes of all the protein molecules that are important in biology. And that's just gonna be an enormously important resource for medicine. We're gonna see new drugs, new treatments. Outside of medicine, we're gonna see other breakthroughs in biochemistry, maybe bacteria that can, you know, produce petroleum and things like this, right?

(Martin Ford at 00:53:52) So there is enormous potential there, and that's just one example. And we're gonna see many, many different examples of artificial intelligence being applied to medicine, to science. And I think that's the primary, most important benefit of this technology. And it's the reason that, you know, as I said, I think the benefits are gonna outweigh the downsides. And that's the most important reason—it's gonna be the application of artificial intelligence to science and medicine and areas like that.

(Joel Beasley at 00:54:23) Is anyone tracking the release of new technologies to prove it's actually on an exponential growth curve?

(Martin Ford at 00:54:32) I mean, there are people like Ray Kurzweil who've done a lot of analysis on this. I'm not aware of a comprehensive study. And in fact, there are some indications that in many areas, at least looking back over the last few decades, we've actually not been accelerating. We've been stagnating. What's clearly been happening is we've seen a continuous acceleration of technology in information technology, right, in terms of computers, communications, and some areas of biotechnology that are information-oriented, like gene sequencing, right? You know, unraveling our DNA and stuff like that. All of that is basically powered by Moore's Law. Right? The computers have been getting faster and faster because of Moore's Law.

(Martin Ford at 00:55:19) But if we look at other areas of technology—transportation, you know, energy, infrastructure, buildings, agriculture—we're not seeing that same kind of acceleration at all. In fact, in many areas it's quite stagnant. I mean, you look at airplanes, right? The airplanes we have today are maybe somewhat more fuel efficient. You know, they've got these fancy glass cockpits with lots of automation. But in terms of the basic function of that airplane, it's really not much different than in the 1970s. Right? The jetliners we have today are actually, in some cases, slower than the planes we had in the 1970s. Right?

(Martin Ford at 00:55:58) Back in the 1970s, we had the Concorde, which offered supersonic transport. We don't have that. Maybe we'll have it again, but we don't have it now. Look at automobiles. Cars we have now are definitely better than the cars in the 1970s. They're also way more expensive. They're safer. They've got, you know, a lot more equipment on them. But the basic function of the car is not fundamentally different from the 1970s. But now compare the car that was available in the 1970s to what was available in, say, 1890, which was basically a steam train at best and a wagon pulled by horses at worst. Right?

(Martin Ford at 00:56:44) I mean, you look at the acceleration in technology across the board that happened between 1890 and, say, 1950 or 1970, and it's just mind-boggling. Right? I mean, and it wasn't just one thing. It wasn't just transportation. I mean, it was everything. It was cars, airplanes, public sanitation systems, antibiotics. In virtually every sphere of technology and human endeavor, we had this massive explosive advance, right, over those years. But if you go from 1970 to now, we have not seen that, not that kind of across-the-board acceleration. I mean, you know, the appliances that you have in your home today, in the kitchen, are not fundamentally different from what was available in 1970 or even in 1950.

(Martin Ford at 00:57:38) Right? They're better, yes. They're computerized and all this stuff, but not anything like the kind of fundamental advances that we saw in the past. So one of the real questions for artificial intelligence is, will it be the catalyst that enables us to have that kind of across-the-board acceleration once again?

(Martin Ford at 00:58:02) So are we gonna see everything get better, more advanced the way it might have? You know, you ask your grandparent, you know, the kind of change they saw over their lifetime—are we gonna see that kind of change again? And I think artificial intelligence is absolutely our best hope to make that happen, to see that really kind of fundamental advance across the board.

(Joel Beasley at 00:58:28) Yeah. I'll check out some of Ray's stuff as well. I didn't know he had research on the exponential growth thing because, you know, I agree with you. A lot of the things that we have haven't fundamentally changed. Where I've seen explosions are in the details of things. You know, new systems developed in biology to do different things. We have, you know, CRISPR. We have a number of different—but it seems to be that they're smaller, and there's a lot of them. There's so many of them. I built a show around it, and I can't even address 1% of them.

(Martin Ford at 00:59:01) Right. I mean, that's a lot of, you know, many things that would be considered incremental. I do think something like CRISPR is—I mean, that's not just a small incremental advance. It's a real game-changer, I think, and it will lead to even more advanced techniques as well. So, I mean, we definitely are beginning to see evidence that this pace of innovation is picking up.

(Martin Ford at 00:59:24) And definitely, that's in part due, you know, in large measure, to the impact not just of AI, but more broadly of computing, right? The fact that we now have computers that are just incomprehensibly faster and more powerful than we had in the past. And also, we have so many more of them. I mean, you go back to 1960, 1950, you probably could have counted the number of computers in the world, right? You could have made a list. Here's where all the computers are. And not only have those computers become, you know, just incomprehensibly faster and more powerful, but there are just basically an infinite number of them. And so what it means is that our general capacity to compute, you know, as a species, in terms of all of humanity, our ability to compute, to manipulate information, to solve computational problems is just unprecedented.

(Martin Ford at 01:00:24) Right? There was nothing like this ever before in history, and we are beginning to see the dividends from that. Right? We're beginning to see the application of that to many different areas. And that's why we have CRISPR.

(Martin Ford at 01:00:37) That's why we have mRNA vaccines, right, that have become so effective. It's really the result of this massive advance in our computational capacity. So just computing has done that for us. But now we're gonna take that to the next level with artificial intelligence. We're now building machines that genuinely are gonna have cognitive capability.

(Martin Ford at 01:00:58) And I still think we're just really in the infancy there, just at the beginning of that process. And ultimately, as that comes online and really becomes assimilated into our economy, we're gonna see, you know, a disruptive impact from that. So I think we're gonna see a lot more in terms of advances, not just in computing, not just in computers and software and things like that, but across the board.

(Joel Beasley at 01:01:28) I've got two more questions for you. Is that okay?

(Martin Ford at 01:01:31) Sure.

(Joel Beasley at 01:01:32) Okay. First one is, do you think that, you know, we get all these problems, all this automation is happening, while at the same time we're advancing AI because AI is causing some of the problems, right? Do you think AI will be smart enough to help us find the best solution to those problems?

(Martin Ford at 01:01:49) Possibly. I mean, it's definitely gonna be an incredibly important tool. You know, the people that are pessimistic about this just make the argument that, you know, we need to make sure we understand how to control artificial intelligence before we develop, you know, the superhuman intelligence, right, the super intelligence.

(Martin Ford at 01:02:11) Because once that super intelligence evolves, then it will be too late, right, if it, in a sense, escapes out of the box or gets out into the real world and starts changing things and then we don't know how to control it. So the people that are focused on this—and there are some very smart people—the most prominent is Nick Bostrom, right, at Oxford University, who wrote the book that came out back in 2014 called Superintelligence. It was sort of a really influential book that really dealt with these issues about how do you deal with a superintelligent AI?

(Martin Ford at 01:02:45) And there are some really, really smart people working on this, which I think is a good thing. So hopefully we will, you know, have the research in place to help control these systems if and when they do develop. But again, I think this is a problem that's pretty far out there. I think 50 years, maybe, at a minimum. Maybe 20 years. But there are people—I mean, Ray Kurzweil, we keep talking about Ray—he believes that we're gonna have human-level intelligence in 2029. That's his prediction that he's been pretty consistent about.

(Martin Ford at 01:03:19) So 2029 is what, seven years? I mean, you know, not long. I think that's very unlikely, but, you know, there are smart people that believe this, so we need to be prepared for any possibility.

(Joel Beasley at 01:03:34) Yeah. I know, right? It's a hard argument to dismiss some of the smartest people's ideas. Exactly, right. We should probably keep that on—

(Martin Ford at 01:03:41) But I think the fact that you've got brilliant people who have just dramatically different opinions—I mean, my book, Rule of the Robots, actually is based in part on interviews I did with, you know, at least 20 of the absolute smartest people in the world working on AI. People like Ray Kurzweil, but also Demis Hassabis, who's the CEO of DeepMind. You know, the people that recently won the Turing Award, like Geoff Hinton, who's the top expert in deep learning. I interviewed all those people, and I asked them to give me their predictions for when we'd have human-level artificial intelligence.

(Martin Ford at 01:04:21) And those predictions range from, you know, less than 10 years from someone like Ray to over 100 years. And these are brilliant people that know more than anyone else about artificial intelligence, and this is the extent to which they disagree, right? And it just shows you how unpredictable this field is, how even the absolute smartest people who know the most are just widely divergent in terms of their predictions for how this field is gonna advance. So what that tells us is that, you know, this is really kind of like a Pandora's box.

(Martin Ford at 01:04:58) We don't know what's coming out. We don't know how it's gonna develop. We're just gonna have to be prepared to think on our feet, you know, as all of this develops, because it's super unpredictable.

(Joel Beasley at 01:05:10) I've got a hypothetical fun question for you. Is that okay?

(Martin Ford at 01:05:14) Sure.

(Joel Beasley at 01:05:14) All right. Let's say later this week, you're driving down the road and a Cybertruck pulls up next to you, right? And they roll down their window and it's Elon Musk. And he's like, "Hey, Martin, how are you doing?" And you have a little exchange with him, and he invites you back to his laboratory to see some new technology he's been working on. You're like, "Yeah, cool." You go to the laboratory, and he's got a neural implant that an average person can use and it's non-surgical and they just, you know, place it right on your head, or maybe it's something more dramatic. They place it in your ear. Would you do that? Would you say, "Sure, Elon. I'll put the Neuralink in"?

(Martin Ford at 01:05:58) Sure. If it were non-invasive, I would think twice about having brain surgery, right? I mean, and I think that realistically in the near term, that's what it looks like, right? I mean, it's really hard to build something outside of your head that can really meaningfully interact with your brain because your skull is really thick, basically, right? So all of the real practical research on Neuralink, I think, is more invasive than that. But of course, if it were—I mean, who wouldn't, right?

(Martin Ford at 01:06:32) I mean, sure. I mean, if it would enhance your cognitive capability, if it would enable you to interact directly with the Internet, right, inside your brain. I mean, anyone would do that, right? So yeah, I mean, that it's entirely possible that that is the future of the species.

(Martin Ford at 01:06:51) And that's, again, we keep talking about Ray Kurzweil. That's his vision, right, that we, in essence, merge with the machines. I don't discount that, but I do think it's a really hard problem and it's pretty far out. And because I just don't think we're gonna get that noninvasive thing where you clip it on your ear and you're ready to go. I think that, you know, it's gonna be a long time before that works.

(Martin Ford at 01:07:17) And it's a hard problem to solve, but it'd be really interesting to see what Elon Musk produces. There's another company called Kernel, which is basically a competitor, that's working on that too. So there are multiple companies out there working on this idea of interfacing directly with the brain. It'll be really interesting to see how that develops.

(Joel Beasley at 01:07:37) You interviewed so many people for that book. You said over 20 of the world's greatest experts. What was your personal favorite interview?

(Martin Ford at 01:07:45) I don't think I have one, but I really enjoyed talking to Ray, because he's a fascinating guy. I think some of his ideas are kind of kooky, but at the same time, he's a brilliant guy who's doing very real research in artificial intelligence. I enjoyed talking to Demis Hassabis, right, who's the CEO of DeepMind, who's again, this is the company that I think is the number one artificial intelligence company. It has produced not just the most notable advances in terms of building systems that win at chess and Go and things like that. But as I mentioned, what it's doing with AlphaFold is now it's getting into other areas.

(Martin Ford at 01:08:25) Right? It's really now focusing on practical stuff, on building artificial intelligence that's gonna have an impact on science, technology, climate change, things that really, really matter. So, you know, they're the company that so far I see taking this from demonstrations of games and things like that that are cool, taking that to the real world to really making it something that is gonna have a dramatic impact on our lives. And I think that's really, really exciting. So, yeah, I mean, there are so many interviews that are fascinating.

(Martin Ford at 01:08:58) I talked to David Ferrucci, who is the guy that led the team that built IBM Watson, and he's now running another AI company called Elemental Cognition that is working on yet another approach to more advanced artificial intelligence. So it's really fascinating to talk to all of these people that are taking totally different approaches to building more sophisticated AI systems. They are incredibly smart, capable people who completely disagree with each other on the best course of action, and I think that's really, really fascinating. And, again, that shows just how open this field is. I mean, if you looked at a more traditional field like physics and you talk to physicists, I don't think you would see this much disagreement between them.

(Martin Ford at 01:09:46) You know what I mean? There'd be more consensus on what the best ideas are. Whereas in artificial intelligence, it's just wide open. You've got people taking completely different paths. And I think if you look, if you read my book, Rise of the Robots, that's one of the things that I hope comes through, that I really tried to give the reader the sense of all these different approaches that are taking place in parallel.

(Martin Ford at 01:10:14) And we don't know which one is gonna really result in the breakthrough that really takes us forward. And I think that that's the thing that makes AI really fascinating and exciting.

(Joel Beasley at 01:10:27) Yes. There's no doubt that this is an incredibly exciting time to be alive.

(Martin Ford at 01:10:31) Absolutely. I think the next decade will be one of the most exciting in history. Absolutely.

(Joel Beasley at 01:10:40) Thank you so much for listening. And if you found this episode useful, please share it with a friend or colleague who you think would get value from it. And if you have topics that you'd like to hear discussed on the podcast, either add me on LinkedIn or send me an email: [email protected]. Every time I get an email or LinkedIn message, it absolutely makes my day and inspires me to keep going.