„Rise of the Robots: Technology and the Threat of a Jobless Future”
We talked with Martin Ford about artificial intelligence – what threats does it pose and to which sectors of the job market, what chances does it create for us and how should we react to its fast development.
Martin Ford – he is an author whose work focuses on artificial intelligence, and its impact on the job market, economy and our daily lives. His work includes „The Lights in the Tunnel: Automation, Accelerating Technology and the Economy of the Future” (2009), „Rise of the Robots: Technology and the Threat of a Jobless Future” (2015) and „Rule of the Robots: How Artificial Intelligence Will Transform Everything” (2021).
Agata Zadrożna, Rafał Górski: In 2017, your book „Rise of the Robots: Technology and the Threat of a Jobless Future”, was published in Poland under title „The Dawn of Robots: Will Artificial Intelligence Take Our Jobs?”. We believe the original title, „Rise of the Robots: Technology and the Threat of a Jobless Future,” is much more fitting. Seven years have passed since its release. Could you list the three most important issues you predicted accurately and three issues that did not come to pass? Why did the latter not happen?
Martin Ford: In terms of the predictions that are accurate – the most important thing I said in the book is that artificial intelligence and robotics are going to continue to advance. And I would say that maybe the technology has moved even faster than I expected. Certainly, in the last year or so, we’ve seen all these developments around ChatGPT and some other things. These are already at a level that I probably would not have expected, and I think we’re already beginning to see, definitely impact on work. For example, we hear stories about graphic artists that have lost their jobs because of AI systems that were able to generate images. I think that we should look at the capabilities of these chat systems with regard to writing material – there’s an obvious threat to many white collar jobs. Now, one thing that we’ve learned since I published the book, is that primarily, the concern is with particular tasks within a job, rather than a job itself – so it’s not going to be the case that an AI system or a robot is going to come along and do exactly what a person is now doing, and so that job disappears. What happens is that these systems are able to take on more and more of the tasks, than the person. Which means that more time is freed up and it’s kind of an open question whether there will be enough more creative or other work for that person to do. But I think in many cases what we’re going to see is kind of a consolidation – because a lot of what each person is doing can be automated. Then what was once two jobs becomes one job. So the boundaries between jobs, the definitions of jobs and the job titles – all of that changes in response to this technology, and I think that, that is largely happening now. In terms of what has not happened – here [in the United States] we certainly have not seen high unemployment, unemployment rates are certainly low. However, at the same time, unemployment really only measures people that are actively looking for a job, and those numbers are, according to standards, very low. There’s still many other people that have completely given up and left the workforce. Those numbers are getting historically much higher. As I said in my most recent book: if you go back to like 1960 about 97% of working age men, meaning men between the age of 25 and 54, were either working or looking for a job. And now that number is less than 90%. Maybe 88/89% in the United States. So there’s a huge number of men, that at one time would have been working, but are now completely out of the workforce, and it’s not because they’re old, and they retire, because this is the fine age working group. So something is happening there, and I think the technology has to be part of that.
The truth is that, back when I wrote the book, I didn’t make really precise predictions because I know that that’s kind of pointless. So I think in general, the way things have played out continues to pretty much be in line with what I said. We have not yet seen mass unemployment or systemic unemployment. But I do think that the technology is moving in such a way that I’m still really concerned about, especially with the developments with AI in the last year or so.
I’m a sociology student, so predicting the future based on observing the present and learning from the past is very interesting to me. I wanted to ask what was your take on it.
What I try to do with my books is to explain the technology and the trends that are happening and the things that we should worry about rather than really making precise predictions because it’s really impossible. I didn’t expect ChatGPT – I mean, I was surprised by that. It’s a breakthrough that came out of the blue. And there will be more of those. So I think it’s kind of a foolish thing to make precise predictions. But if you pay a lot of money to a consulting company, they will try to do that because that’s how they make their money, right? But usually those kinds of precise predictions almost always turn out to be wrong. So I think what you have to do is to understand these trends and the technology so that you can react in real time as these things develop. That’s the best approach.
What do the latest scientific studies say about the question, „Will artificial intelligence take our jobs?”
I know that one of the people who work on this is Eric Brynjolfsson, who is a pretty famous economist, is one of the people that work on these issues, and he actually wrote a book called Race Against the Machine. His research shows that these latest innovations in AI – matter of fact generative AI like ChatGPT and so forth – have not eliminated jobs or replaced jobs, but the impact that they’re having is that they are beginning to equalize different groups of workers. By utilizing these technologies, what we’re seeing is that more average workers – in other words the workers who are not really the best, the most talented – are able to up their game and compete more effectively with the “top-notch” workers. Whereas the workers that are already very proficient – so far these systems are not helping them as much. So this is kind of equalizing things. However, my guess is that it’s not necessarily going to continue. I think that generative AI has really only been with us for about a year or so, so it’s very, very early in the story and the technology is progressing very fast – I don’t know if you saw, but there was just an announcement a couple of days ago from open AI that they’ve got a totally new approach now to ChatGPT so that it’s now doing more reasoning, beginning to think like a person. So these systems are getting a lot more capable, and I do think that whatever the research shows, for the first year that they’ve been around is not going to be the same as what we’re going to see after they’ve been around for five years. It’s going to be a big difference.
What is your current response to so-called experts who argue that we are not facing a „jobless future” because people will simply find work in other sectors?
That’s the way it has worked in the past, and people who say that are basing it on history. But if you look back, for example, at one time in every country most people were farmers, right? Most people worked on farms. They had very primitive tools. And then we had a revolution in agricultural technology – tractors and combines and all this equipment. So, in more advanced countries, only a tiny number of people now work on farms, right? And what happened is that all those people got left. They lost their jobs on farms. But then there was this other sector rising – which was manufacturing. So, they all went to work in factories. That’s what happened around the late 1800s and early 1900s in the United States and other countries. And then later, factories also began to automate. So, there’s automation and later there were robots, and AI in factories now. And also, in countries like America, there was offshoring, right? Where a lot of the factories went to China and other places like that. And so what happened in countries like America and most of our countries is – people moved into the service sector. Most people are working in offices, they are working in fast food restaurants, grocery stores, and most of the jobs are in the service sector. But what’s happening now with artificial intelligence is that this technology is finally coming for everything. AI impacts every sector: agriculture, manufacturing, service sector, healthcare. There is no part of the economy that is not going to be impacted by this. So what I said happened in the past – people just moved to jobs in another sector, in another area – I don’t think there’s going to be another sector. There certainly will be new industries that didn’t exist before, but those industries are right from their inception going to use AI and robotics, right? So they’re not going to be labor-intensive. They are not going to employ a lot of people, so that’s why I believe that this time is different. That’s where I would differ with the people that, based on the historical record, think that there’s not going to be a problem, I think eventually there will be a problem.
Do you think there are jobs that would benefit from the rise of AI and robotics?
There are basically three groups of workers, based on the kind of tasks that they’re doing, that I think will kind of be complementary to AI. In other words, AI is going to work with them and be a tool to help them rather than replace them – at least for the foreseeable future. See, within the next 5 to 10 years beyond that, who knows – I mean, maybe… I think that the groups of workers that are sort of safest right now – that are least likely to be replaced, are for example, workers that are really creative/creating things. I think that AI – for someone that’s really thinking outside the box, creating something new that wasn’t there before – can be a very powerful tool to help you do your job better and kind of amplify what you can do. For now, I don’t think it will, for the most part, replace you. But already there are exceptions, like I mentioned – graphic designers. That’s a creative job, but already there are systems that you can just give a text description and the AI system will generate the photo, generate the image. In general, I think the more creative types of workers are going to be less likely to be replaced and more likely to use AI as a tool to help them accomplish things.
The second area is workers that really need to build deep relationships with people. You know, interpersonal relationships. Think about a nurse – you’ve got to have empathy for your patients, an understanding, really, of human nature. Maybe a business consultant – where you really have to have an understanding of the client in order to help them. So, I think areas like this where we have a uniquely human ability to engage with other people and build relationships with other people – that’s another area where people are going to be relatively safe. So again, AI is also certainly making strides in that area – there are already chatbot systems that can help people with mental health issues, for example depression, which you would think that is something very human, right?
Do you think it’s ethical to have AI talk to people who really need help?
If it’s helpful – and keep in mind that they develop these chat systems, and they have shown them to be helpful to people that have these problems. And the AI system is available 24/7, right? Maybe you have a counselor/a psychologist you go to see once a week or something for an hour, but that person isn’t there at 3:00 AM, if you suddenly feel very anxious or depressed or something. Whereas the technology is always there. So, the technology definitely does have advantages, right? But there are also concerns too – about the ethical nature of this. So there’s a real trade-off.
And then the third area – the kind of job I think is safer – jobs that fall in what we call, skilled trade jobs, like, for example electricians and plumbers. These are people that have to move around in a very unpredictable environment and this requires lots of dexterity, hand eye coordination and also problem-solving skills. So, think about a plumber, coming to your house to fix the pipes. With every situation – the premise is completely different, and it’s very unpredictable. So building a robot that could do that – it would be like a science fiction robot – if you saw the movie Star Wars – C-3PO, like that. We’re not close to that. I think those kinds of jobs are very safe. But, you know, beyond that in particular, any kind of job where you mostly sit in front of the computer, and you write reports, and you play around with spreadsheets and things like that – all of that is definitely going to be, I think, at risk from automation, from artificial intelligence.
The thing is, of course, the manual jobs, the blue-collar jobs – you’ve got to build a very expensive robot. And you need hand eye coordination for a lot of those. That’s really hard to build a robot that… Just now they’re beginning to see some robots, finally, that can begin to get close to human dexterity. They’re going to be able to reach out and grab an item. Your vision has that kind of special recognition to reach for objects, manipulate in the real world – that’s much harder, than just manipulating information, and that that’s the reason that the robotics is in many ways actually a bigger challenge than just automating white collar intellectual work, including, you know, work that is done by very, very smart people, that have gone to school for years and years and years to be trained. Still, what they’re doing is often easier for the AI than what a very “low skilled” person does. Think about the housekeeper that makes up a hotel – like doing all the cleaning, and making the bed. I mean, again, to make the better robot that can do that is almost impossible. And yet that person has very low pay, right?
What would you say to students entering the workforce in this new era of AI-driven job markets? Should they be afraid? How should they navigate this future?
Yeah, the point is not to be afraid, but just to be aware of this and to educate yourself. And I don’t think that all the jobs are going to disappear for a very long time. I think that there may be some impact in the coming years or so. I don’t know how great it will be, but certainly you know, in terms of your own career, if you’re a young person just entering the workforce, the most important thing to do is to be aware of AI and to work with it.
But you know, there are a lot of other things that young people worry about, like climate change, and AI is a part of the solution, right? I mean, AI is a very powerful force for making us wealthier, more prosperous, for solving big problems. In fact, I think we face a lot of problems – like climate change, and coming up with new sources of clean energy and so forth – the problems are so hard that if we don’t have AI, it will be very difficult for us to solve them. I think it’s going to be indispensable as a resource to help us solve a lot of problems that we’re going to face in the future. So overall, I think AI is a good thing. It’s just that it does have a lot of dangers that come along with it, and we need to be very open-eyed and honest with ourselves about those dangers, and we need to have policies that can address those, so that we can leverage the technology and make everyone’s life better, rather than making everyone’s life worse.
Do you think it’s kind of a double-edged sword? Because on one hand we can have innovations but on the other hand, AI also takes a lot of power.
Yeah, that’s true, especially with the latest generative AI. It’s just, the big computing centers, the circle farms, that they have to power this technology, consume enormous amounts of electricity. If that’s not coming from a green source – if it’s coming from a coal plant or something, that’s very bad for the environment. So that’s a real issue. The other side of it, though, is that AI, as I said, is indispensable in terms of innovation to solve those problems. You know, in those big computing facilities, AI is already used to manage the power system, so it uses less electricity and AI is also being used in innovation, like new solar power. There are a lot of companies working on nuclear fusion, right? Which could be the ultimate answer to green energy. And they’re all using AI in that research. And you know, AI is making that possible. So, like you say, I think there are definitely two sides to it, but hopefully AI is going to be ultimately the solution rather than contributing to the problem.
How might artificial intelligence impact the precarization of employment? How is it gonna impact the already unstable work environment that we have?
One thing that happens as a result of technology is that, we have seen kind of a migration away from those full-time, normal jobs to more gig economy type jobs, right? And that happens because the technology makes it possible. So the classic case would be like an Uber driver. You know, that’s possible because of the technology – with the GPS and the algorithms that allocate the rides and so forth, that’s what makes that kind of work possible. Certainly, a lot of the work along those lines is in the gig economy and there are two sides to that: it gives you more flexibility, you can decide when you want to work, but it can also make it harder to work full-time and have a reliable income. We also need policies to address that. That’s a big issue in America. There have been some initiatives to offer more workplace protections and better benefits to the workers that are in that more of a gig economy.
How do you see the role of workers’ unions and workers’ councils in organizing employees in the face of rising automation and AI?
Well, it depends on the country and it depends on the approach they take. In some countries, unions have a lot of influence with government – with policies in governments, which could be a positive thing. But the problem I have with unions is that very often the obvious thing that they’re going to want to do is to restrict the technology. I think it’s the: “we don’t want robots in our factory. We don’t want artificial intelligence to be used in the workplace because it’s going to affect our jobs”. So I think that’s bad because, overall, I’m a proponent of artificial intelligence. I think it’s a good thing. It will help us to become more efficient, it will help us to create new products and services, most importantly, it will help us to have new scientific breakthroughs in medical areas. So I don’t think we want to stop artificial intelligence. But I think we need to have policies to help the workers that are impacted by this, and that’s why I’ve talked about things like universal basic income. The problem is that unions, if they represent a group of workers, they aren’t concerning with what’s happening with all workers everywhere and trying to make a policy for all workers because they’re concerned with what’s happening to them, specifically in their workplace. And so very often what they’re going to do is they’re going to fight against innovation, and I actually think that’s a bad thing. So, I’m kind of ambivalent about unions. If they’re engaged in the big picture, and they have a positive take on technology, then they can have a positive role, but if their objective is just to restrict the technology and hold things back then, and I don’t think that’s good.
As citizens, what should we demand from politicians, expect from journalists and require of ourselves in the context of your books?
My main argument is that I think everyone should, you know, learn about AI, educate themselves about AI. That doesn’t mean understand the technical details, but understand how this technology is developing and how it can impact the world, and the good sides and the bad sides of it. And then we have to demand from our politicians, that we have an honest discussion about this, and that we began to craft policies beginning to address these issues. And you’re definitely beginning to see progress in those areas in Europe – in the EU, there’s something called the AI Act, which most companies are opposed to, like it’s not necessarily great, but it’s a start, right? In California, there’s also an initiative to regulate AI, and also, most of the big companies don’t like that either. You know, they don’t really like regulation. I think ultimately we’re going to have to figure out something to address this, and it’s not just about the regulation of, you know, things like deepfakes and so forth, but it’s also about what do we do about the economic impact, the impact on the job market. That’s a much harder problem than just coming up with a regulation that says you can’t make a deepfake using someone’s face or something like that, obviously it’s illegal, right? The economic progress might be much harder.
There are a lot of ethical issues involved in AI and one of the scariest seem to be automated weapon systems. How can we stop this? Should we organize petitions to politicians? Is there any way to take action?
So, I’ll tell you honestly, I’m not optimistic that it can be stopped because there’s already an initiative in the United Nations to basically ban automated weapons in a way that chemical weapons are banned… Although, even though chemical weapons are banned, they were used in Syria, right? So even banning doesn’t prevent it. But all the militaries – they haven’t agreed to ban this. The top countries like the US and Russia and China – they haven’t gotten together and agreed to ban these because they don’t trust each other, and definitely, automated weapons on the battlefield will have a big advantage. And actually this is already happening right next to you, in Ukraine, they are already beginning to use these weapons, right? And Ukrainians are pursuing that because they’re at a big disadvantage relative to Russia. So they had no choice but to try this technology. It’s hard to tell them, you know, given that Russia is decimating their cities, that they can’t use everything they can to fight back. I think it’s probably inevitable that these weapons are going to be developed. I think the most important thing is to make sure that they don’t fall into the hands of terrorists. That that should be our main focus. I’m kind of skeptical that we can prevent their creation.
Do you have any book recommendations for someone who wants to know more about AI and automation?
There’s a book that came out right before mine did, again, by Eric Brynjolffson, and the other author’s name is Andrew McAfee, named Second Machine Age, and it deals with kind of similar issues to what I talked about, but it’s more optimistic. You know, if you don’t like the idea of what I’m saying – that I think eventually, a lot of the jobs are going to disappear – they have a more optimistic view of it. So, you know, it would make you feel better. There’s another book called AI Superpowers that was written by a Chinese author, [Kai-Fu Lee], that talks more about the competition between China and United States in particular, and it’s pretty good. And I honestly think one good book is my more recent book called Rule of The Robots. It includes like an updated discussion of the impact on employment that I talked about in Rise of the robots. But it also talks about all the other issues, like the ethical issues, you know, privacy and deepfakes and all that other stuff, as well as it explains the technology. That’s a pretty good book for getting an overview of all the issues that you need to understand, not just about the impact on work, but about everything about AI and having this kind of basic understanding of how AI works.
Agata Zadrożna, Rafał Górski
