The Unexplained - Howard Hughes - Edition 592 - Martin Ford

Nov. 22, 2021 - The Unexplained - Howard Hughes

50:58

Edition 592 - Martin Ford

Author and AI expert Martin Ford - on his new book "The Rule of the Robots" - We ask if Artificial Intelligence is getting too controlling - and how can we shape our high-tech future...

Large

Audio Only |

Scroll

Time	Text
	Across the UK, across continental North America, and around the world on the internet, by webcast and by podcast, my name is Howard Hughes, and this is the Unexplained.
	Well, here we are, still stuck in the middle of November, and if you've followed my show for long enough, you'll know that November traditionally has been my least favorite month of all the months in the year.
	I think it just goes back to when I was a teenager.
	Do you have that that there is a particular time of the year that you don't like?
	I think it's just like being a summer person, born in the summertime.
	I think we get to a low point in November, but I have to say, the weather is weird.
	You know, it's almost spring-like at the moment.
	I'm recording this in a t-shirt, no heating on.
	It's during the day.
	It was a little chilly first thing this morning, but nothing that you couldn't handle.
	Very sunny, beautiful, beautiful morning as I record this.
	But it's clouded over a little bit now, and I think it's going to be grey again.
	So very strange weather.
	When I was a kid, which we know is a long time ago in Liverpool, I remember November's as being crispy, cold and icy when I was a kid.
	And then December, you've got snow sometimes.
	Then January, more of the same and pretty cold through February, and then things would magically turn around in March.
	Everything seems to be different in this day and age.
	But then, like I say, maybe you'll appreciate this.
	I do go back a long way.
	Okay, the conversation that you're going to hear on this edition of The Unexplained is a longer version of my radio conversation with Martin Ford.
	He is an IT person and author and viewer of the scene, let's put it that way, to do with artificial intelligence and all the developments that are coming at us thick and fast in that arena.
	He's written one book about this, The Rise of the Robots.
	Now he writes and brings to you The Rule of the Robots.
	Now, a lot of people we've had on this show have been very positive and very upbeat about what artificial intelligence and robotics can bring to us.
	And that's true, you know, elderly care and making better decisions for insurance claims, maybe that kind of stuff AI is all used for.
	But there are downsides to artificial intelligence that you don't often get to hear about.
	So we'll talk about some of those with Martin Ford, author of The Rule of the Robots.
	I think you're going to find this interesting.
	Let me know, of course.
	You can always go to my website, theunexplained.tv, and you can send me a message from there.
	And the website, of course, designed and created by Adam from Creative Hotspot in Liverpool.
	If you want to send me your story, by the way, always pleased to hear those.
	Please send me those by email.
	And if you want to put in the subject line, my story, and if please, you could keep it to two paragraphs, short ones, maximum three.
	Otherwise, it gets very hard to go through them all.
	But you've sent me some remarkable stories.
	And once again, proof of what I always say, ordinary people, I know this in my own life, experience extraordinary things.
	So your story is always welcome.
	All right, let's hear the conversation now then from my radio show, Martin Ford, author of The Rule of the Robots.
	Martin Ford is online to us from California.
	He's a futurist and a man very much concerned with these things.
	His first book about this, I think, five years ago now, The Rise of the Robots, gave us a few warnings and a few indications of what was to come.
	The latest book might have a chilling title, depending on how you look at it.
	The Rule of the Robots is out now.
	And Martin Ford is online to us now from California.
	Martin, thank you very much for joining us.
	How are you?
	I'm good.
	Thank you for having me.
	So, Martin, let's start with you and your interest in this.
	You know, why are you concerned about this?
	Well, I've been involved in technology work for a long time.
	I actually ran a small software company for many years out here in California.
	And it became obvious to me that technology was going to have a big impact on the nature of work, that a lot of jobs in particular would be automated as artificial intelligence advanced.
	And that's what really motivated me to write my earlier book, Rise of the Robots.
	And since then, that process has really continued.
	And artificial intelligence is now reaching literally into everything.
	It is becoming almost like electricity.
	It's everywhere.
	And it will become even more so in the future.
	It's going to impact virtually every aspect of our lives.
	It's going to impact every industry, every business, really every aspect of our society.
	So that's what motivated me to write this second book, Rule of the Robots, How Artificial Intelligence Will Transform Everything, because I truly believe that this is going to be a technology that does transform just about everything.
	Give me a little thumbnail sketch, a summary of what you said in The Rise of the Robots, because I'm guessing that that is the work that set the stage for this.
	Right.
	Rise of the robots was primarily focused on the impact of AI and robotics on the job market.
	And it focused on the fact that a whole lot of jobs, in particular, any kind of job that tends to be more routine, repetitive, and predictable, the kind of job where someone is coming to work and doing the same kinds of things again and again, that that would ultimately be susceptible to be automating.
	So we could see, you know, in the coming decades, literally tens of millions of jobs disappearing in both in the United States and in the United Kingdom.
	So a huge impact on the workforce potentially.
	And that's something that I think we're really going to have to adapt to.
	We're going to have to find solutions to that, or we're going to end up with a society that is much more, even much more unequal than it already is.
	And many, many people are going to be left behind.
	They're going to be struggling to have an adequate livelihood if we don't take on this problem.
	Okay, well, we'll deal with all of that, which is, I think, really, as I said, hot button stuff a little later in this conversation.
	Just give me an idea of the history here, though, because I think an awful lot of us, our heads are spinning with the pace of it all, Martin.
	You know, I love technology.
	It's absolutely empowered my life.
	I'm recording these words with you.
	On Friday, I'm able to, you know, do recordings at home for broadcast.
	That is something that 20 years ago I couldn't do.
	So technology has empowered me, but the pace of it and the speed of it has increased enormously.
	I don't know whether it's exponential.
	It seems to be.
	You know, how did we get to the point that we're at?
	Well, most of this is really driven by what's called Moore's Law, right?
	The fact that computers are getting faster and faster.
	They roughly double in speed every two years or so.
	And that's been going on for decades.
	And that's really the primary force that is behind all of this.
	And that's what's led to all the changes we see around us.
	But within all that technology, the particular technology of artificial intelligence is, I think, the one thing that is going to be the most consequential of all.
	And we're really only beginning to see evidence of that.
	You see it all around you already.
	You know, you see that Amazon's Alexa can begin to have a rudimentary conversation with you.
	You see that you can translate between languages on Google.
	You see, for example, Tesla's self-driving function built into their cars already.
	So we're beginning to see this already, but over the coming decades, it's going to become vastly more important.
	So this is really becoming the driving force, you know, that's really going to completely transform our whole society and our way of life.
	Well, you know, I tend to agree, Martin.
	I don't think there's any way that you really could disagree with that sentiment because the evidence is all around you.
	But it's hard for people, isn't it?
	It's hard for me getting on with our daily lives as we do to actually step back and appreciate the pace of what's happening to us.
	If I tell you that I got my first computer, I was working for a company in London called Capital Radio and I bought it through the company.
	It was a Dell.
	And they said to me, you're probably not going to need all the hard drive capacity because you just won't be able to fill that.
	It was three gigs.
	And that was 1997.
	The computer cost me an awful lot of money and I spent one year paying it back off my credit card.
	That is how expensive it was.
	But it absolutely, even though by today's standards it's laughable, it completely and utterly transformed my life.
	I think the thing that I'm getting to is that most of us don't actually have the time to stop for a second and think back what life was like before all this stuff began to happen for us.
	Do you agree?
	Exactly, right.
	It's moving at a pace that we really is sometimes difficult to comprehend.
	And you don't realize how fast the technology has advanced, really.
	I mean, most of us walk around with a computer in our smartphone that is dramatically more powerful than a supercomputer would have been, you know, something owned by the government, the most advanced computer in the world back in 1997, as you said.
	If you go back to the moon landing, sorry to jump in, you know, we've got more, I've got more, here's my little phone here in my hand.
	You can't see it, but, you know, I've got more computing power here than my first computer at home and more computing power by a long way than the lunar module that landed on the moon.
	Oh, exactly.
	Yeah.
	I mean, dramatically so.
	I mean, I mean, you know, orders of magnitude more.
	So yes, it's, it's, you know, this technology has just moved at an extraordinary rate and it is accelerating and continuing to accelerate.
	And that is really what is making technologies like artificial intelligence possible.
	You know, it's the fact that computers have gotten so much faster, so much more powerful, that is really bringing to life technologies that just a few years ago would have been science fiction, that you would have only seen in science fiction movies, right?
	And this is being made reality by these advances.
	Before we get into the specifics of the book, which I want to get into in the next segment of our conversation, you know, I'm a great fan of Star Trek, The Next Generation, where, of course, as you know, for a while, Jean-Luc Picard, the captain, was assimilated by a mechanical person, the Borg.
	Are we getting to a stage where we may be assimilated by technology?
	It's a crazy question and, you know, maybe not the best phrased question I've ever put.
	But could we get to a situation, bearing in mind that neurotechnology is upon us, that era is here?
	Do you think that we might get to a situation where we will become part of the technology, not just be made redundant by it, will actually be subsumed into it?
	Yeah, I mean, that's certainly a possibility.
	And I think that we're on sort of on that path and that's happening already.
	We don't yet have implants into our brains connecting us to computers.
	Although, as you mentioned, there are initiatives in that area.
	There are a couple of companies, one founded by Elon Musk that is actually trying to work in that area.
	But even setting that aside, we are now completely tied to our smartphones, right?
	I mean, how many times during the day do you look at that little screen that you carry with you?
	You become almost attached to that.
	You become integrated with your phone and the technology around you.
	So in that sense, at least, this is already happening.
	And it definitely is true that in the future, maybe decades from now, we may see even a more direct integration between these technologies and people.
	How has technology, if it has, empowered you?
	Well, it's enormously empowered me just in terms of writing a book, for example, the way that I do that today compared to the way I might have done that 20 years ago.
	I mean, I would have been going to the library and trying to find paper magazines and books and references now that's literally all at my fingertips.
	Actually, another technology I used when I wrote this latest book, Rule of the Robots, is that I actually dictated a lot of it.
	Rather than sitting down and actually writing the text, what I did, at least initially, is I actually dictated it into my phone, my mobile phone, and an application there was able to recognize that speech and turn that into text.
	And then I began working with that text and essentially edited it to get it into a format to write the book.
	So these technologies are really transforming the way we all do our work on a daily basis.
	And there is a lot more to come.
	in the future, the way we interact with computers will, I think, become much more conversational.
	It will become almost as though you're literally speaking to another person.
	So it's going to be just enormously transformative.
	And like I say, I want to get into the deep specifics of the book in the next segment.
	But you said it's almost as if it's like talking to a person.
	I agree.
	Sometimes it is.
	You know, sometimes I can be on the phone to an automated system and make the mistake of thinking that that is a person on the other end, making decisions based on things that I say to it.
	And of course, it isn't.
	But you can be forgiven for thinking that.
	What are we as human beings to do then about that?
	Do we need to educate kids in schools about where we came from and what really matters and what it's like to interact with a system and not to forget that the system is there for your benefit?
	You are not there for its?
	Right.
	I mean, this is a problem that we're going to have to figure out as a society as this kind of line between what is human and what is machine begins to blur.
	And in fact, you already have people developing unhealthy relationships with social media, with video games.
	You've got already technology stepping into roles that we would have thought were uniquely human.
	For example, there are chatbots that can actually provide basic mental health assistance to people.
	People with depression, for example, can interact with one of these chatbots, and that's been shown to be quite helpful, which is a good thing.
	I mean, it's a great tool, but it is definitely blurring that line between what is human and what is not human, what is technology.
	And that is going to become dramatically more the case going forward as these technologies become more and more indistinguishable from human beings.
	So you were definitely going to have issues with people, for example, developing relationships with machines rather than people.
	And those relationships certainly may not always be healthy.
	They may interfere with the ability of people to form relationships with other people and so forth, healthy relationships.
	So these are all issues that we're going to have to confront.
	And that's a big issue because the technology is, if it's programmed, right, it's always going to tell you what you want to hear.
	And as we both know, it's not always the case that being told what you want to hear is good for you.
	Exactly right.
	Human beings need challenges and we need to face reality, right?
	So one of the big problems I see is that these technologies, artificial intelligence, virtual reality and so forth are going to, in the future, create a very, very attractive fantasy world that many, many people are likely to be attracted to.
	And we could have an issue where people kind of check out from the real world and spend a lot of times in what literally is going to be something like the Matrix, right?
	Where they're living in this kind of alternate universe that is totally detached from the real world.
	And as a result of that, they're not going to be productive in the real world.
	And that's one of the problems that I think we're going to face for sure going forward.
	When you say because of it, they're not going to be productive in the real world.
	Is that because, like you said, they've checked out, they're existing in a fantasy world almost.
	I mean, we would have called it that 40 years ago before we knew about all this stuff.
	Is that what you're saying?
	Yeah, and these technologies will eventually be addictive like a drug.
	And I mean, to some extent, they already are.
	I mean, you've already, for example, there have been studies that have shown that young men are basically dropping out of the workforce.
	They're not looking for jobs or holding jobs, both in the US and in the UK.
	And there have been studies that have looked at that to try to determine how they're spending their time.
	What are they doing if they're not working or looking for a job?
	And it turns out that overwhelmingly what they're doing is they're playing video games.
	And we're already seeing this.
	And going forward, technologies like video games are going to become dramatically more attractive, more realistic, more addictive as they begin to incorporate both artificial intelligence and virtual reality.
	And so this will become a bigger and bigger problem for sure.
	Quoting from the book here, Martin, it's therefore, you say, an extraordinarily bold claim to argue that artificial intelligence will evolve into a general purpose technology of such scale and power that it can reasonably compared to electricity.
	Nonetheless, there are good reasons to believe that this is the path whereon AI, much like electricity, will eventually touch and transform virtually everything.
	It's a big statement.
	Yes, it is.
	It is a very big statement, especially when you really sit down and think about how important electricity is to us.
	I mean, just think of a single day in your life, how many times you interact with some technology that relies on electricity.
	I mean, how life would you, how different would your life be if the electric power really went out and it wasn't there?
	I mean, obviously, this is, you know, when this happens, it's considered to be an emergency.
	I mean, we worry about electric power in hospitals, for example.
	You know, lives are really, you know, at risk when the electricity goes out.
	And the point I'm making is that we are definitely entering a world where artificial intelligence is going to become that important.
	It's going to touch everything.
	Everything that we do is going to rely to some extent on AI.
	And if for some reason that artificial intelligence were to disappear, it would have a very dramatic and negative impact on us.
	So that's the world that we're entering.
	And we need to think seriously about that.
	And we need to embrace that, I think, because this is going to be an extraordinarily powerful, helpful technology in the same way that electricity is in order to help us solve a lot of the Problems that we face.
	You know, things like climate change, for example, we're going to need a lot of innovation.
	And I think artificial intelligence will be absolutely critical to that.
	But at the same time, artificial intelligence is obviously, even though I'm comparing it to electricity, it's also obviously in many ways very, very different from electricity.
	It's not simply a commodity power source.
	It's a very dynamic, disruptive technology that actually encapsulates intelligence.
	And much more pervasive for that reason.
	You know, the comparison with electricity, I think, is a great one.
	I've never thought of it until I read your book today.
	And now I think, yeah, this is like the early days of electricity.
	Now, in the early days of electricity, we tend to forget that a lot of people got electrocuted.
	You know, that devices were perhaps not as well made as they might be.
	People weren't told how to use them.
	And consequently, devices like early, you know, ions and that sort of thing, they had a tendency to electrocute people from time to time.
	They weren't as sophisticated.
	I think, although it's a big leap, and as you say, the technology that we're dealing with now is much more pervasive for a whole ton of reasons.
	You know, there are dangers that we need to be aware of, just as people needed back in the early days of electricity to be aware of those dangers.
	Exactly.
	And the truth is that with AI, the dangers, you know, there are more of them than with electricity, where you simply have to worry about being electrocuted.
	There are a whole range of concerns and risks that come with artificial intelligence.
	Everything from risks to our security, the potential for algorithms to be biased, the potential for this technology to be used directly as a weapon.
	These are all things that we need to worry about.
	So it's much more complicated than electricity because this is a much more dynamic, disruptive technology than electricity is.
	You spend a lot of time in the book, and I think everybody should read this part of the book for definite.
	They should read it all, but this part particularly, charting the developments that have led us to where we are.
	And I think those kind of start with the rise of deep neural networks, starting with the use of, I didn't know what these things were until I read the book, the use of GPUs or graphics processing units instead of the traditional CPUs.
	Now, you know, some of that is familiar to me because I do a certain amount of computing, but not all of it.
	But you say that that's one of the stepping stones.
	Yes, I mean, the basic thing that happened here is there has been for a long time going all the way back to the 1940s, this idea that you could build a system that was loosely based on the way the neurons in your brain are connected.
	And what computer scientists tried to do is to, at a kind of very rough, rudimentary level, they tried to replicate that in this system.
	So that's not a new idea.
	There have been people experimenting with that for many decades.
	But what happened just in the last decade, right around 2012, is that finally computers became powerful enough as the result of Moore's Law, right?
	This acceleration in computing power, they became powerful enough to really make this into a practical technology.
	And that's what happened.
	And then combined with that, we also have enormous amounts of data now that we didn't have before.
	And all of that data becomes essentially the training input for these algorithms.
	So you've got very fast computers running these powerful neural network algorithms that can hoover up enormous amounts of data and learn from it.
	And that is what caused this disruptive breakthrough in artificial intelligence.
	And it's this technology that is now powering all these things we see around us, like Amazon's Alexa and the self-driving cars and the translation between languages and so forth.
	It's all the result of this disruptive advance in artificial neural networks.
	Another mile marker you talk about is cloud computing.
	Most people, I mean, I tend not to use the cloud for storing my data.
	I tend to have hard physical drives backing stuff up here.
	But I'm seen as a dinosaur.
	Most of my friends put everything up in the cloud and they think anybody who doesn't is nuts.
	But you say that cloud computing is also a factor.
	Yeah, cloud computing is really becoming, you know, if you think of artificial intelligence as being like electricity, then cloud computing is like the electrical wires.
	It is what is delivering this technology.
	It's the conduit, the infrastructure that's making it available everywhere.
	And essentially, cloud computing is when computing becomes centralized into these massive computing facilities that are owned by companies like Amazon or Microsoft or Google.
	And these companies are competing with each other to offer more and more advanced capability.
	And they're offering compute power, they're offering artificial intelligence.
	And this has become now so widespread that virtually every business of any size across the developed world is now accessing these resources, utilizing cloud computing.
	And as a result of that, they have direct access to artificial intelligence.
	And that technology is getting better and better as these companies, Amazon and Microsoft and Google and so forth, compete to offer more and more powerful technology.
	So that's really becoming kind of the driving force behind this advance.
	You say, and I don't quite understand this.
	So if you could explain this for me, you say there is one thing that no company gives away for free.
	It's data.
	Well, we know that bit.
	You say that this means that the powerful synergy between AI technology and the vast quantities of data it consumes will inevitably be skewed in one direction.
	Nearly all the value generated will be captured by whoever owns the data.
	So are you saying that by centralizing data in this way, there are risks for ourselves that we're throwing up?
	Well, what I'm saying there is that in order to really leverage artificial intelligence, in order to really generate value from this technology, you need three things.
	You need very fast computers, strong computational resources.
	You need these strong, powerful artificial intelligence algorithms.
	And the third thing you need is lots and lots of data that's useful data.
	Now, the first two things, the compute power, the computational resources, and the algorithms are being widely distributed.
	They're available through cloud computing.
	In many cases, the software code for these algorithms is even given away for free.
	So the cost of accessing these first two components is very low.
	It's going to be very accessible to any business anywhere.
	So the thing that really is going to matter is going to be that third component, the data.
	So what you're going to see is that companies that own huge amounts of data are going to be able to create a lot of value from that.
	They're going to become potentially very wealthy based on utilizing that data.
	So it might be data owned by a health insurance company or by a financial company that owns data.
	But the point is that the value is really going to be in that data.
	That's interesting.
	But if I, you know, say I was creating great works of intelligence or I was creating products and I was coming up with fantastic ideas, firing them out 500 pages at a time, you know, every month, and I was putting it all up in the cloud.
	Are you saying that if I give over that data and put it in a, you know, secure cloud storage, that the benefits of that, whether I know it or not, will be accrued to the owner of the data, but the possessor of the data?
	Well, I wouldn't go that far.
	I think if you're creating, for example, a work of art, a book or something, and it's hosted in the cloud, you don't have to worry about that being stolen from you or anything like that.
	But I do think that companies that have access to our data are going to generate enormous value from that.
	But it's data that is collective, right?
	You need millions and millions of people participating in this and contributing their data in order for that data to be of value.
	So it's actually going to be large organizations and governments that control lots and lots of data from huge numbers of people or from other sources that are really going to see a lot of benefit from it.
	And the detail of the data is not, we're told what they're specifically interested in, but it's the trends and that sort of thing that the data throws up.
	It's almost what's contained in the metadata, yeah?
	Exactly.
	The value is not in any individual person's data.
	The value is in the collective data and the intelligence, the insights that you can gain from all of that aggregated data.
	Why is that a bad thing?
	Oh, it's not.
	I'm not saying that that's a bad thing.
	What I'm saying is that in terms of the way the value of artificial intelligence is distributed across our economy, it's going to be captured by companies that own lots of data.
	They're going to get a lot of value from this.
	So this is more of an insight for the business world, for people that want to know which companies are going to benefit.
	That sort of thing.
	This is an insight into who's going to be getting rich in the future.
	Exactly, right, right.
	You also say a very key part of all of this that is almost coming in under the radar because it's being introduced pretty much everywhere.
	And I'm certainly looking forward to being able to get these speeds.
	But 5G, you say 5G is a very important part of this.
	Exactly, because it's going to enable much greater connectivity.
	People talk about the Internet of Things, for example, which just means that every machine, device, appliance, industrial machine are all going to be connected to the Internet.
	And there are going to be algorithms that operate across these networks that can, for example, diagnose problems in devices or machines and sometimes fix them, that can control everything, make everything more efficient.
	And all of that is going to be powered by artificial intelligence, but you need very strong connectivity for that.
	And, you know, 5G will definitely make that possible.
	And that will be a great thing in terms of efficiency and scale, but it also will introduce concerns, especially around security.
	Well, it would be.
	Because as everything becomes more connected and more automated and powered by algorithms, there's, of course, a real danger there that someone could hack into that or attack that, you know, a cyber attack, for example.
	So these things.
	I've never seen in Canada, haven't they?
	Part of the Canadian health system.
	I think Labrador and Newfoundland, I think it was, were affected by a major hack in the last week or so.
	So all of this stuff is with us now.
	But as you say, if large amounts of data are going to be centralized in just a few places, comparatively, then we have to be at least concerned about the security.
	What can we do about that?
	Well, one of the things that we'll do is that we'll use artificial intelligence to defend these systems, to keep them safe from cyber attack.
	That's happening already.
	The companies, for example, that specialize in cybersecurity are definitely already leveraging artificial intelligence to do this.
	But at the same time, it's likely that the black hat people, the people that want to attack these systems, are also going to leverage AI.
	So there's going to be kind of an arms race utilizing this technology.
	So this is something that, you know, we definitely need to be investing resources.
	We need government oversight.
	We need regulation to ensure that there's an appropriate level of cybersecurity built into all of these systems that are going to become much more prevalent in the future.
	Sounds like a good idea.
	And we need to be on that now.
	Problem is, I think evidence shows that our lawmakers, that's a big sweeping generalization.
	I'm sure some of them are, but a lot of our lawmakers just don't understand this stuff.
	So a great deal of these things may be happening under the radar and the necessity to protect data, to protect systems, to understand that bad guys will be able to use artificial intelligence in the future to defeat our artificial intelligence potentially.
	A lot of that Is just not being recognized because the people who make the laws maybe don't understand this.
	Yeah, that's a real problem.
	I mean, if you happen to watch any of the events when Mark Zuckerberg from Facebook was testifying in front of the U.S. Congress, it becomes pretty obvious that they're simply not that technologically savvy.
	They can't realistically keep up with this technology, whether it's Congress or Parliament or really any legislative body.
	So what I think we're going to need in the future is specific regulatory agencies that do have the necessary expertise to form regulations, you know, something akin to the Federal Aviation Administration in the United States or the counterpart in the UK that governs, you know, aircraft, right, and makes rules for how that works.
	And it's in the same way you have agencies that govern securities trading on Wall Street and so forth.
	And you've got agencies that govern medicine and health and regulate drugs and so forth.
	They all have special expertise in those fields that allow them to do that job.
	We're going to eventually, I think, need an agency that can do a similar thing around the issues associated with artificial intelligence.
	It sounds like, Martin, that we needed this yesterday, not today.
	Exactly.
	Yeah.
	We're already, I think, behind the gun there.
	So we should be moving on that very rapidly.
	Martin Ford on the rise and rise and possible rule of technology.
	He's in California with us now on The Unexplained.
	And I suppose the bit where the rubber meets the road, as they say, is how this affects our day-to-day, worker day lives.
	Now, I've spoken to a lot of experts in this field.
	You know, some of them are people that you see on television that are well known for their scientific expertise.
	And a lot of them have told me that actually all of this is a great thing for us because it's not going to decimate jobs because the jobs that will be created in the future are the jobs servicing the robot and artificial intelligence systems.
	So that's going to be okay.
	We just have to educate people to take on those roles.
	And then all the horrible stuff we don't like doing, all the menial stuff, the manual stuff, and the getting, you know, getting your hands dirty and stuff like that will be undertaken by automated systems.
	And we'll be programming them.
	And there'll be plenty of jobs for everybody.
	I don't think you see it necessarily that way, do you?
	Not quite.
	I mean, I would agree that potentially artificial intelligence could be a great thing for us and it could bring many of those benefits for sure.
	I just don't think that we can sit back and just let this play out without worrying about any issues and expect that outcome.
	I think that if we do nothing, that artificial intelligence absolutely will destroy an enormous number of jobs.
	And it certainly will create new jobs too.
	But I think there are real questions as to whether there will be enough of those new jobs to really absorb all the potentially tens of millions of workers that could lose positions as this develops.
	And I think there's also some real questions around how successful many of those workers that have currently predictable, routine, to some extent, repetitive jobs, how many of those people are really going to be able to effectively transition into the new types of works that are going to be created?
	Because in many cases, the new jobs that are created are going to need very specific talents and capabilities.
	They may require, for example, that you be a very creative person.
	They may require that you have very good interactive skills with other people, that you're able to form sophisticated relationships with other people and so forth.
	And not everyone is going to have these types of talents.
	So I think that we're at a very real risk that a lot of people could potentially be left behind.
	And that's going to make our society much more unequal.
	Right.
	You analyze, I think it's Amazon's warehouse system, which is robotically controlled.
	I think it's them.
	Tell me if I'm wrong, but I'm going to quote you.
	In the book, you say, quotes, workers are gradually losing their agency and being transformed into what essentially amounts to plug-in biological neural networks that fill the gaps in a largely mechanized process by rendering the capabilities that are so far beyond the reach of machine intelligence.
	That sounds pretty bleak.
	Yeah, I mean, and that's what's happening.
	If you go inside one of these Amazon warehouses, you will find thousands of robots already and you have also found thousands of workers already.
	But what's happening is that the warehouses are really organized around the capabilities of the robots.
	And what the robots do is they basically move inventory shelves around and take them to workers who then stand there and either put items in the shelves or retrieve items from the shelves to fulfill orders for customers.
	And the reason that the workers are there doing this work is that the robots are not yet able to do what the workers do, which requires human level visual perception and dexterity and hand-eye coordination, things like this.
	The robots are not yet advanced enough to do this kind of work.
	You use the word yet.
	Yet, right, right.
	But right now, the people are doing this work.
	You know, they're literally kind of a cog in the machine doing this repetitive work again and again.
	This can be very dispiriting.
	Many people have complained that they feel like robots just doing this very repetitive task that so far the robots can't do.
	And of course, we don't have Amazon on the show tonight to speak for themselves, but I know that quite often I've seen people interviewed on television saying that they have very fruitful and rewarding working lives there.
	So there are two sides to the coin, I guess.
	Yes, exactly.
	But the point I'm making is that eventually, probably over the next 10 years, the robots are going to get a lot more proficient and they will then be able to do the tasks that right now require people.
	So as a result of that, I think over the next decade or so, you're going to see these Amazon warehouses become a lot less labor intensive.
	There are going to be fewer jobs there for people.
	Eventually, and it may take more than a decade, you know, these warehouses are likely going to begin to approach full automation, where there simply won't be many jobs there at all.
	There'll basically be a few people overseeing the machines, but it will be almost entirely fully automated.
	And that, of course, depends, and this is the holy grail, isn't it, for people developing these things.
	That depends on the robotic machinery being able to develop the sorts of perceptions, the kind of judgment that human beings are capable of.
	That's the cherry on the cake, but it's a big and valuable cherry, and it's a real hard one to attain.
	Yes, I mean, within an Amazon warehouse, you don't necessarily need all the judgment and capability of a human being, right?
	You don't need human-level intelligence, but you definitely need robots that can match or exceed human-level dexterity, you know, manipulating objects and so forth.
	And this is something that Jeff Bezos actually said he expected to achieve within the next decade.
	So this is not science fiction, the idea that you could have robots that can do the jobs that Amazon workers are now doing within 10 years or so.
	Or indeed, any other similar organizations.
	Exactly.
	Perhaps not quite as big as Amazon because what is, but yeah, I understand.
	You say in the book, overcoming the current limitations of deep learning systems will require innovations that bring machine intelligence inexorably closer to the capabilities of the human brain.
	And you quote, the leader in this area is DeepMind, a company founded by a man called Demis Hassabis.
	Unusually for an AI researcher, you say he received his graduate training not in technology, but in neuroscience.
	That's important.
	Yes.
	I mean, Demis Hassabis is, you know, definitely one of the absolute smartest people working in artificial intelligence.
	And he's, you know, in the UK.
	DeepMind is based in London.
	And, you know, his vision is to create what's called artificial general intelligence, which essentially means artificial intelligence that is at a human level.
	So going far beyond anything that we have today.
	What we have now are AI systems that are very narrow and specialized.
	They can do very specific things.
	What DeepMind ultimately would like to do is create a system that replicates general human intelligence, where you can think in a very holistic way.
	You can create new ideas.
	You can solve any problem that a human being could solve.
	And this would be, you know, if this occurs, it will be probably the most consequential invention that humankind has ever produced.
	I mean, it will be an enormously important breakthrough with incredible implications.
	Sorry to jump in.
	Will the machines ever be able to use the kind of gut instinct that we do?
	I mean, for example, I'll give you a good example here.
	You know, the famous landing of the plane on the Hudson River, you know, Sully, the pilot, who was a hero portrayed by Tom Hanks in the movie.
	Now, he was being told by systems and by people in his ears talking to him on radio that he should make for LaGuardia.
	But he knew that the plane wasn't going to make LaGuardia.
	And he had to do the only thing that was open to him, use his training and bravely, miraculously ditch that plane in the Hudson River.
	And everybody survived.
	An astonishing thing.
	But if an AI system was flying a plane, and they are talking about pilotless planes one of these days, I think there'll be a lot of public resistance to that.
	An AI system might have made the decision to go to LaGuardia, but that system would have been wrong because the plane wouldn't have been able, and it was proved in the subsequent hearing, the FAA hearing, that the plane wouldn't have been able to make it to LaGuardia or wherever else they were asking it to go.
	So these systems are a long way from being able, aren't they, to match that kind of instinct?
	Yes, absolutely.
	We are a long way from that.
	But still, that is the holy grail of artificial intelligence to build a machine that would be able to think in the way that a human being thinks and maybe would be able to solve a problem like that or would be able to solve many other problems in the realm of geopolitics,
	figuring out what strategy should your country have in terms of geopolitics, in terms of being creative and solving problems in fields of science and medicine and so forth.
	That's the holy grail of the field.
	And at a minimum, we're decades away from that, I would say.
	Maybe we're 50 years away, maybe even 100 years.
	But most people that work in artificial intelligence do believe that someday we will build a machine that is at that level, something that will match and ultimately exceed human intelligence.
	So someday we will have an artificial intelligence that could actually do a better job in a situation like happened with that airplane than a human being could do, because it will be smarter and able to think, obviously, much faster.
	So what happens to us?
	That's the vision.
	Well, that's the question.
	I mean, once technology gets to that point, then you can really ask the question, who is left that really has a job at all?
	Because you can imagine at that point that machines could literally do just about everything that any human being could do and probably do it better.
	And that would even apply to areas that you might think would be reserved for human beings.
	For example, art, for example, acting, you know, but there's already evidence that we're going to have synthetic actors in movies and things like that powered by artificial intelligence.
	So this is a real question for us is how are we going to live with the technology that in many ways might potentially exceed us?
	And again, I don't think that that's something we need to worry about in the immediate future.
	That's something that's still out there.
	But 30s, 40s, 50 years.
	It does raise real questions for their future.
	And, you know, this is something that we need to plan for.
	People, I think, were bemused or some of them shocked by we have this thing every Christmas Day, 3 p.m. on the TV here, the Queen's Christmas Message.
	A lot of people gather around ever since I was a kid, the TV set, to listen to Her Majesty the Queen give her view of the past year and look to the future.
	There was an alternative version of this that you might have heard about on Channel 4 television in the UK that was created by deep fake technology, controlled by humans, all of it.
	But it had the Queen dancing around Buckingham Palace, you know, sort of doing flips and all backflips, all kinds of stuff.
	But, you know, we know that with the best will in the world, our Queen is an amazing woman, but she can't do that at 95.
	You know, an astonishing thing.
	If we get to the stage where a computer programmer doesn't have to do that, but the actual characters that you've created, or maybe they've created themselves, think for themselves.
	I don't know what that means for any of us, Martin.
	Yes, I mean, this is certainly one of the biggest threats that we face.
	I mean, obviously, the example you're giving there is humorous and fun, and that's fine.
	There's nothing wrong with that.
	But these technologies could be used in nefarious ways, right?
	In the same way that a deep fake can make the queen dance and say funny things, a deep fake in the future might be able to make a politician say very destructive things or things that would be destructive to his career, perhaps just before an election or might be destructive to the country that he's running, he or she.
	This is a real danger.
	And the thing is that these, you know, the technology in particular of deepfakes is going to get better and better over time.
	And it's going to reach the point where it probably will be virtually impossible to tell what is real and what is the fabrication.
	And that will pertain to audio.
	You could make someone say something that they would never actually say.
	It could apply to video.
	You'll be able to create deep fake videos and so forth.
	And there are real dangers there.
	I mean, we've seen, for example, how disruptive a viral video can be with videos of police interactions in the United States, for example, have led to literal social unrest.
	Indeed, as we saw over the last year or two.
	But imagine if those things were artificially, if some technology was able to create those things artificially, then you may think that we've got problems now with people spreading fake news online by writing stuff.
	Just imagine if videos could be created or audio content that might generate itself appears out there.
	It doesn't seem to me that we're too far away from that.
	What on earth do we do about that?
	Somebody once talked to me about possibly using watermarking technology in videos and audio, deeply embedded, so that you're able to tell by going into the data stream whether something is genuine or not.
	I don't know.
	Is that an idea we need?
	That's one of the approaches that is being investigated.
	And there are some companies in specific applications that are already doing these kinds of things.
	For example, there's a company that is used to do that kind of thing with photographs that are used for insurance claims and so forth, that are used as evidence for insurance policies.
	But whether that can be done more generally is kind of an open question.
	I mean, the people that are really involved in these technologies have said that probably there's not going to necessarily be a technological solution to this.
	We may just have to live in a world where you cannot really believe what you say, what you see and what you hear.
	It could always potentially be a variocation.
	Of course, that's going to be enormously disruptive.
	Absolutely.
	That's where the wheels come off the bus.
	I think if we get to that position, I think we've got to do an awful lot of thinking.
	And our lawmakers, who've already got a lot to deal with, need to be giving some consideration to this before it just is too late.
	Martin Ford, thank you very much, Need.
	What's the title of the book?
	You tell me?
	Rule of the Robots, How Artificial Intelligence Will Transform Everything.
	And we can get that both sides of the Atlantic.
	Yes, absolutely.
	All right.
	Martin Ford, food for thought.
	Thank you very much for speaking with me.
	Thank you.
	It's been great to talk to you.
	Your thoughts.
	Welcome about Martin Ford, author of The Rule of the Robots.
	Some chilling information there, some things that we maybe all need to be aware of as we move forward in this world.
	And we've got multiple problems to deal with anyway, haven't we?
	Great opportunities offered by technology.
	But also, as we've seen in many occasions just reading the news, there are also downsides and we've got to be savvy about those.
	More great guests in the pipeline here at the Home of the Unexplained.
	So until next we meet, my name is Howard Hughes.
	This has been The Unexplained Online.
	And please, whatever you do, stay safe, stay calm, and above all, please stay in touch.
	Thank you very much.
	Take care.

▲