This is a download from the FM89.9, the Business Station. The FM89.9, the Business Station. Okay. The distinction between human and machine seems to be a clear one.
But what happens in a world where the machines are more human than the people and the people are more automated than the machines? That's right, it's time to Matt's spleen. Matt, I seem to remember the last time I studied with Jeff's undo. We were talking about killer robots.
Then I took a look at today's topic and it seems we're talking about killer robots. You want to explain that? Hey, Rich. Well, yeah, I'm a one-trick pony or maybe a one-trick mechanical bull given that it's machines we're talking about.
The last couple of weeks with Jeff, we've been trying to connect the dots from the Facebook data scandal to the data-driven world of tomorrow that will start to blur the lines between artificial intelligence and natural intelligence or human intelligence if you want. So this is the third week that I've been trying to tackle this subject. It's so big that we never seem to get more than halfway through an argument before we have to call time. All right, for those of you who are just joining in or maybe tuned in for this part, let's have a quick summary of where we've been and where we are now.
Okay, two weeks ago, we started by talking about Facebook, Cambridge Analytica, and the data that is spilling out of all kinds of online services and into the hands of parties that we know about and also parties that we don't know about. In the intervening time, Facebook has done some incredibly clever things like move about one and a half billion of its users' information around the world so that they wouldn't be subject to the strict European data and privacy control sort of coming up in May, which is always a great thing to do when the world's governments are looking at the way that you treat privacy and data control. Mark Zuckerberg's testimony before Congress was also criticised in some quarters for being less than transparent on certain issues. But what seemed most clear about those hearings is that when lawmakers and Facebook talk about privacy, they're not actually talking about the same thing.
You know, lawmakers are talking about this old-fashioned sense of privacy, the kind of thing you would find as a dictionary definition of the term. When Mark Zuckerberg and companies like Facebook talk about privacy, there's purely a technological and business framing of the issue. And that's kind of a new definition of the word that the dictionaries and of course lawmakers have still to catch up on. So last week, you were talking more about how this flow of big data will fuel the world tomorrow?
That's right. And I want to try and widen the scope a little bit more and confuse people a bit further. Also to demonstrate why we can't wait to sort out who is going to own our information tomorrow, because the big companies like Facebook, Amazon, Google, they're all already building the machines and the data gathering systems of tomorrow. You know, we can't just close our eyes and hope it's not going to happen.
So where that information ends up and who controls it could end up shaping the future we live in tomorrow, and it will decide whether we are passive data gathering engines or whether we're actually the driving force of a new kind of social and economic system. Nothing too ambitious, then. I know. I have to bring some sense of closure to the topics.
I'm going to finish it today whether we wrap it up or not. But this is certainly something I think we're going to be coming back to again and again. OK. I suppose we better start with robots.
Last time we spoke, I think I tried to make a case that robots aren't just the machines we find in factories or creatures from the science lab and science fiction that look like us or machines. So when we say a robot, we often mean a clever machine, a spade is a machine, a loom is a machine, a printing press is a machine, but we tend not to think of those machines as robots. They do one thing. They don't move around and they don't talk to you.
Yeah. The clever machine is not necessarily an accurate depiction, though, because we tend to imagine a machine that is actually like us in some ways. We imagine it to be flexible or multi-purpose. We imagine it may be to move in the same way we move, or maybe because it has, you know, some kind of intelligence and knowledge.
The world of tomorrow is going to be full of clever machines that meet some of these criteria, but without necessarily being a walking, talking robot. Cloud intelligence, is that what you mean? Yeah. I mean, the way that we're currently structuring machine, machine intelligence services, yes, I do mean the cloud.
Many of us already interact with cloud-based artificial intelligence, even if we don't realize it. The obvious ones are Siri and Alexa and DeepMind. They're also the less obvious ones, which I mean all the back-end services. So we're talking about things that run on Amazon or Google's cloud, or any number of other commercial platforms, and they use machine intelligence, which is actually built into those services and actually helped to run all the back-end operations.
So in effect, you could be interacting with AI multiple times a day, simply by browsing online and never actually be aware of it. Do you think that intelligence is always going to seem to come out? Well, that's actually a really cool question and concept to think about. And that's where we start to get into the more speculative areas.
Certainly, I think that it's the kind of system that a lot of commercial companies would favor, because it centralizes not just the intelligence, but also the information. So that model of the cloud suits a lot of companies like Facebook and Google and Amazon. But this is really a kind of short-term versus long-term argument. In the short-term, artificial intelligence requires a lot of resources.
So it makes sense to have those resources located centrally, and for users to tap into them as and when they need them. In a way, it's a bit like the 1950s, where it would have been impossible for somebody to have a computer in their home. It simply took up too much space. It was too expensive.
And it needed a lot of people and a lot of ingenuity to keep the machines running. You fast forward to now, and we've got computers that are far more powerful than those building-sized 1950s models. And they're just sitting in our pockets. At the moment, some parts of the technology are accessible and cheap, and that's the internet connectivity and the chips.
So when we talk about smart home devices, it doesn't make sense to shoehorn all the power of the cloud into an oven or a fridge freezer, especially when it's going to take up 15 people fields worth of storage space. Especially when you can just add a couple of chips that will connect it to that cloud and give it access to all the intelligence and functionality. Do you think that will change in the future? I don't see why it wouldn't.
That's one of the reasons that the conversation about the ownership of data is so important. If we can't assert ownership of our own data, the chances are we're always going to be locked into these commercial intelligence systems. And that would be a real pity, because when you look at the intersection of all the technology, both online and offline, you start to see a picture where power and governance could be far more decentralized. We start to build this picture where power grids or health provision, manufacturing and retail start to become much more localised again.
You also start to build a picture where these services are maybe a bit less commercial in nature, where energy is generated from renewable sources and it's distributed around a community, where automation and healthcare puts a specialist neurosurgeon at the end of every street, and we routinely go and see geneticists rather than doctors as we focus more on prevention than cure. In a Facebook and Amazon controlled world, you would probably be buying all of those services from subsidiaries of Facebook and Amazon, rather than accessing them as publicly owned resources, which would probably come up close to zero cost. Can governments really provide that kind of infrastructure? Yeah, of course.
I mean, the idea that they can't really is hogwash. You know, this idea that only market capitalism can provide essential services is really a new one. And I don't mean a couple of centuries, I mean, since the 1980s, when you look at most of the major advances in technology over the last 100 years, those advances have been sponsored or entirely developed by public bodies using public funds, whether in medicine or technology, all these different areas. Even today, a lot of the patents and research that feed into our technological field world start-off as research projects at public universities funded with public money, although potentially in collaboration with commercial companies.
So things like the internet, touch screens and GPS started off as projects within the public realm, and often they had a military or space research application as well. So absolutely, the future of artificial intelligence could be in public rather than private hands. And actually, a story I read this morning, the day after I wrote this, says that the EU is actually looking at spending close to €20 billion in investing in artificial intelligence as a public resource for Europe. So people are starting to look at that.
Because it's a case of us forgetting our own history and not just forgetting. There's also a concerted campaign to remind us in inverted commas that the market is the only efficient provider of innovation, which of course is more hogwash because the market innovates for profit, and most, if not all innovation, has no immediate commercial potential. But you're painting a utopian picture, right? A bit of a dream?
I know it sounds like it, but it doesn't actually have to be. That's why I keep stressing this issue of ownership of data. I know it doesn't seem like an obvious connection. But if our data is owned by us and shared publicly because we choose to share it, we stand a better chance of unlocking the potential of a lot of the technology that is literally just around the corner.
In the same way that some politicians are starting to view sites like Facebookers, potential public utilities rather than as a commercial service. The freedom of this information to move from place to place and not to be locked into the service of one private company or another. We exert a large influence on the way societies are going to look and are going to be shaped. And they will also help to determine for whose benefit those societies operate.
And one of the reasons that ownership of this information is so important is because the machines of the future are going to look a lot more like us just as we start to resemble them. And I'm well aware that that last sentence either won't mean much or will be very confusing to a lot of people. So I will try and unwrap it a little bit after the break. Brilliant.
We are listening to Matt Spleyne here on the FMP 2 9.9, of course, with Matt Armitage from Culture Part My Name is Rich Brandt. We'll be back just after these messages. Don't worry, this is the BFM 89.9. Breakfast for Masters, BFM 89.9.
BFM 89.9 at the Business Station. My name is Rich Brandt. I'm standing here for Jeff Sandhu in the studio with Culture Part My Name is the Matt Armitage we've been talking about or still continuing to talk about killer robots and the like. So before the break, you left us with a rather confusing thought that humans and machines will start to merge.
Can you just explain that to me and people at home a little more? With a great deal of pleasure. When I started this series to look at data and robotics, now one of the starting places was Sex Robots, because I'd been asked to talk about Sex Robots on a number of platforms. By most definitions of the term, these machines really aren't robots.
If they have moving parts, it's generally limited to the head. Whether it's automation and a connection to some kind of computing system that simulates personality, what no one could do is mistake one of those machines for an actual person, except for maybe in one respect, which is an unexpected one, the skin. The development of realistic skin is not simply a task for the adult industry. It's very important for prosthetics and also for the machines we may interact with in the future.
Because we recoil from things that feel strange or artificial. When you shake hands with someone and their hand is sweaty or moist, you have to try not to show that feeling on your face, you have to try and stop yourself wiping your hand afterwards. So when machines start to be amongst us, they will either have to look completely unlike us, or they will have to start to mimic us in ways that makes the human brain happy, and the human brain is very hard to trick. And of course, there may also be applications for synthetic skin, for skin grafts, and a lot of surgical procedures, as well as just for these kind of prosthetics limbs.
So is that how you see moving forward then through medical procedures? Well, where we've landed right now is very firmly in the world of transhumanism, which I know is going to confuse people even more, because in a sense we're talking about what comes after us, which is why the movement is also sometimes called post-humanism as well. This is a subject that we've talked about on the show a few times over the years, and sometimes it's kind of hard to imagine, especially when you look at news headlines that are full of impending nuclear war in the Middle East or North Asia, or even between North and South Carolina, don't know anymore. At a time when it seems we're potentially poised to destroy ourselves, it might seem to be a very weird time to declare that we're about to make the technology-fuel jump, or a technology-fueled evolutionary jump.
Right, so this is the convergence you've been talking about for the past few weeks. Yeah, I mean, if we go back to the medical aspects, we usually look at medical procedures as cures. You have something wrong with you and... Take medicines.
Yeah, you fix it. You take medicines. If you can't be fixed with the chemicals, then we try and fix you with technology. So if you lose a limb, you get an artificial one.
We can put you on dialysis and replace your kidneys. We can give you an artificial heart. This model of medicine is a bit like fixing your car after the engines fall in out. It makes a lot more sense to go for regular servicing and maintenance than to repair catastrophic damage.
And it's really weird, because what we consider to be common sense for a car, we consider to be an indulgence for our own bodies. And our bodies are kind of a bit more important to us. Right, so convergence then achieves this how exactly? Well, what we're going to see in the future is gene tools like CRISPR are going to allow us to design the next generation in ways that were previously unthinkable.
We can potentially select and promote specific genes, not just to choose the sex of the child. We can tweak the genes potentially to promote intelligence or height, muscle mass, beauty, hair colour, not to mention all the illnesses and diseases and conditions. We can either make irrelevant, or we can at least try and make people resistant to. We can just engineer them out.
It's kind of a bit like a prenatal pesticide. The end result is that those babies are going to have enormous advantages over children whose genes have not been tweaked or enhanced. Then how does this come back to the data? Because of the enhancement aspect, I know it's not the best example, but look at Oscar Pistorius.
His blade legs made him a faster runner than even the very fastest normally able sprinters. So imagine that you have implants or entire limb replacements. Chips could be embedded in muscle tissue. They could be used to switch off pain receptors at nerve endings, or they could be used to deliver chemicals or other enhancements.
You might favour some subdermal titanium implants, like Wolverine's style to make you a better cage fighter. When you talk about enhancing the human body with this kind of technology, because of all the machinery that's involved, you're still going to have to have some kind of control centre. That's why we're also likely to see chips being added to our brains to enhance both our memory capacity, but also to control all of these other enhancements. It does a little bit, but you have to also look at it from the point of view.
Our brains don't store memories in the way that computer stores them. Our memories are fragmented, and they've rebuilt every time we've recalled. They're actually stored in three different areas of the brain and pulled together. That's one reason why eyewitness testimony gets less reliable over time, because that memory is not a real thing.
It is actually a fiction of creation. Chips can help us to store memories and retrieve them as though they were on a video recorder. That's where we start to get data, an awful lot of data. Correct me if I'm wrong then.
All this is about ownership of our minds. Yeah, which is why I've taken three weeks to get here. Because if I said that three weeks ago, people would think I'm crazy, whereas now they're probably just too numb to care. I know it sounds very strange to think that the decisions we make now about data control in companies like Facebook could actually have an influence on who owns our memories in the future.
That's a very weird concept. If it's on a chip and that chip is in your head, then surely those memories are yours? Well, you would think so, but then have a look at your smartphone. You own the device, which is a bunch of chipsets and a screen, but without the operating system, that device is very expensive, make-up mirror, not even a particularly good one.
It's very dark. We don't own the operating systems. In fact, in most instances, we're not even allowed to play around with those operating systems. True.
And those operating systems gather data. They store photos, they monitor our keystrokes, they track our location, and not always with our express permission. And that's essentially what we're talking about putting in our heads. Those chips will be operating systems.
They will be monitoring and gathering data and potentially transmitting it to other places. And because our brain is an intelligent thing, we would also expect the operating system on a chip inside our head to be smart as well. So our minds would be connected to the cloud? The cloud or some other hub?
I mean, obviously those details are a long way off. But you can see the ownership of those operating systems and that information is extremely important. So the precedents we set now are going to play a large role in determining how those systems and that information are handled in the future. And how does that then bring us closer to these machines?
I know a few people's heads might be spinning already. So if they are, I suggest you hold on tight because it's actually the ground is going to shift further away. I mentioned last week that artificial intelligence is only going to be really useful to humanity once it's truly smart, once those machines are self-aware and have independent thought, they have a conscience and they have some kind of personality. They will in effect be living beings.
And these are going to be another core component of this post-human world. I'm not sure we have the mechanisms to cope with that kind of work. We don't. And that's one of the reasons we have to look at the issues now.
Some people may follow news stories about animal rights activists trying to extend human rights to say higher ape species or mammals like dolphins. What we're talking about goes way, way beyond that. We're talking about creating beings that are as intelligent as us or possibly smarter, whether they're stored in the cloud or in your toaster, those machines are going to have rights. Currently, the relationship we have with AI and robots is one of master and servant.
They do our bidding. So we have to start making decisions about how intelligent we're going to allow these machines to become. Because the sentient and self-aware machine that serves you will be a slave. And as we know from history, society's kind of pay a really high price for the legacy of slavery.
Are you saying that in the future, machines can own property instead of being property? Yeah, this is all part of the decisions we're going to have to make. We have to decide whether these machines will earn money and can buy and own things. Will they be sovereign citizens with voting rights, for example?
Especially as we're going to see these machines penetrate a lot of different aspects of our lives. With an aging population, machines like these connectors can panyons and carers. They're going to be friends. They may well be our bosses.
They may have bodies. They may not. But whatever the form they take, they will be conscious beings. So referring back to your previous topic of sex robots, potentially people can choose to form relationships with robots rather than people.
Well, not just the physical ones. I mean, this is where we start to see fewer differences between these two kind of post-human species. As we become increasingly enhanced by machines, the machines may also be able to make use of organic technologies to become more like us if they want to. So that's where you start to get this intersection between humans and machines.
What divides us becomes smaller than the things that unite us. I know that I'm looking at this purely from a technological aspect. There's a lot of work that's going to need to be put in by people with the grounding in, you know, all the moral and ethical issues. Hopefully before any of this comes to pass or it really is going to be a mess.
But I think there will be genuine relationships, you know, life partnerships if you like between humans and machines. But when we go back to those sex robots, if those robots are conscious and aware, then they will have to consent to any sexual activity. Otherwise, they are effectively sex slaves. So you see that the sheer volume of laws that would have to be changed to accommodate truly conscious machines, I mean, the scale of it is absolutely enormous.
If you switched one off against its wishes or destroyed it, would that count as murder? What would you do with a machine that knowingly committed a crime, especially if that machine was housed in the cloud? How do you punish that kind of machine? I mean, I'm not going to go through a huge list.
I just wanted to point out how big and complex this issue is going to be. Okay, will we be putting self-aware machines then in our own heads? Well, I know you said you found this freaky. So yeah, I've kind of saved the biggest and scariest one for last.
As I said, yeah, our brains are clever things. So if you want to put a chip in your brain, you probably want that chip to be clever as well. So the answer in short is yes. In the future, people will probably be putting other sentient beings inside the heads.
You know, you're really going to have to like and trust someone or something if you're going to give them access to the innermost secrets of your mind. There will also be the issue of whether it'll be a two-way process. Will you get access to the artificial intelligence's thoughts? Or is there going to have to be some kind of unbreachable firewall that exists between the two of you to preserve the privacy on both sides?
I know a lot of people are going to find this creepy because it's a bit like being possessed. And it's easy for me to sit here and tell you it's going to happen because I'm not the one who's going to have to sort out any of the complicated bits. But you can see clear reasons why something like that might help someone with dementia, for example. But it comes back to that thing that technology is never single use.
That's why I've been metaphorically stabbing people with my fork for the past few weeks. The technology that helps a sick person will enhance a healthy one. So no matter how weird or unpalatable we find it now, there will be people who are going to be willing to take that leap. Because it's hard to resist the idea of putting something with the power of Stephen Hawking's mind inside your own head.
And these mind-blowing possibilities are why we need to make the right decisions about big data now. Because those thoughts and ideas should be yours, even if you have to share them with an artificial intelligence. But it's better than some kind of co-ownership with Amazon where you have, you know, Jeff Bezos not getting on your... Sorry.
Better than some kind of co-ownership with Amazon and having Jeff Bezos knocking on your noggin and demanding a penny for your thoughts. Thank you very much Matt. Of course that was Matt from CulturePop here on Matt's Wing of Listen to BFM 89.9. The Business Station.
Don't go anywhere. I'm all stuff going right up with Geek Squad in just a few minutes. BFM 89.9.