00:00:09
[Music] Hello and welcome back to TechTV. Uh today we’re going to be talking about a special report that we’re doing on AI and anthroorphism. U this obviously is the uh tendency that people have to treat uh objects and technology as if they have a personality and as if they are are living beings. um very much in the way that Pete that sometimes we pretend that you’re alive and and living. Absolutely. I mean I I must be a um a victory of anthropomorphization or whatever you want to call it. I mean
00:00:53
it’s basically I mean I bet you didn’t know this. The the actual word comes from um making things look like God because God made us in his own image. So there you go. That’s what we’re doing. I I’m not sure how godlike we look, but I believe that this is a a growing problem. Um uh the way in which we uh depend on technology, whether we’re relying on it to help us with reports, whether we’re relying on it to help us map read or or a whole load of things that um uh we’re becoming increasingly
00:01:25
reliant on uh technology for. Um if we’re engaging in a way, what is the way in uh in which we’re engaging in it where we imbue it with some level of personality? Well, I’ve seen people I’ve seen company directors actually telling me that chat GPT thinks like them. I’ve seen there there was a famous case of the uh the guy working for Google who said that uh the Google AI that he was working with, he was an AI engineer, he said that it was alive. Um this is a big issue and it’s
00:02:02
it’s a huge issue, Bill, if you think about it. I mean, because this sort of thing has been going on forever. Yeah. Children used to have imaginary friends. The imaginary friend was a staple of fiction up to a point. And it it it goes back into history. We we’ve liked to create uh things that come alive, toys that come alive, and we made actual operas and pieces of music about toys that come alive. So there’s almost this this inbuilt desire from people to actually imbue nonhuman objects or non-animate objects with some
00:02:47
sort of life. If you think about it, you know, you you a former sailor, you know, you used to call your boat or still do call your boats she, don’t you? Well, yeah, very much so. But uh that’s largely because it it is a tradition and and being a nautical chap I’ I’d feel remiss if I wasn’t sticking to the necessary etiquette. Um but um I I don’t feel necessarily um an impulse to call to give my car a name or any other object. Um I would call them it and and and think of most inanimate objects or
00:03:25
or or technologies um in terms of an it or something there. I mean, I don’t think should we be referring them to as it or is it a something else? What is the the the correct way to to address these objects? I think you you you’ve hit the nail on the it head. Um, it is a problem. It is a problem because people are starting to call things it. Now, one of the things that the EU uh was actually insisting on in the early days of talking about AI is that there needed to be some clear differentiation
00:04:07
between a human being and a machine so that you actually knew there was that distinction and that discrimination. We are allowing that to be uh diluted or to to uh just decay a little because of this way that we have of you using technology. And it’s not right. There’s going to be two issues. One issue is there are mental health concerns because if you start to attribute attributes that do not exist to a piece of technology then you know you you you have mental health problems along the
00:04:47
way or down the line. But built there’s a second point and that second point is that it costs money. One of the things that Samman actually acknowledged is that people are saying please and thank you to IT systems. I’m not somebody who had an imaginary friend as a child and I I probably uh would be um guilty of of of not really really taking others seriously who did. Um uh my concern in this space is that at least when you had an imaginary friend that person um was entirely in your mind whereas if you’re talking to
00:05:30
um Siri or Alexa or any other um AI um and you’re having a conversation with it, it it is replying to you and and to a large extent it will be reinforcing um ideas you have and that might take you down a rabbit hole. Um uh and my concern here is that um uh uh we can become overly reliant on some of these technologies. Um I’m a as a former naval officer, I’m uh uh pride myself on my ability to read a chart or a map and to navigate. Um but I see so many uh of the younger generation these days who’ve
00:06:07
grown up with um satnav and over reliance on that and and can’t read a map. Um, and I and I I worry that there are so many other aspects of what we do in society where we’re becoming exceedingly reliant on the technology and where I I think we’re losing some of the skills that would allow us to uh function properly if the technology was interrupted or failed at any point. Yeah. you you’re picking up on some of the points that we picked up in the research that we did which is eventually
00:06:41
become this radio program. Yeah. Edward Moser, the Nobel Prize winning neurologist, he says that one of the things that we will lose if we don’t use maps is our inner mental map. That that’s an important cognitive thing to actually develop. One of the things ways that you learn your way around because Moses research was into how you navigate through spaces, you create signposts in your brain, you create landmarks, and that’s one of the things that we do when you’re map reading. So yeah, that though that isn’t
00:07:20
in the program, but Angela Sassi, Professor Angela Sassi makes the point that one of the reasons that we do this, one of the reasons why we anthropomorphize is we want to have a familiarity with something. It reassures us. And so we then become to rely on this thing. We rely on the technology and we think that we’re part of it. Whereas in fact, when it goes wrong, that’s the point that you’re talking about. That’s when you’re in trouble and you’re in trouble because you don’t
00:07:50
understand the underlying mechanism which is going back to what uh professor Edward Mos said but what about not just relying it to uh conduct a function for you like map reading what if you become emotionally reliant or emotionally entangled with a technology um do you see that being as a danger that is a big problem it’s a problem. The the again it’s complex. Um let’s take one point and let’s pick up on that complex word. Uh last week Apple said that AI failed when it had to deal with
00:08:29
complex problems. Now that could just be a a factor of the fact that it hasn’t actually developed enough yet. However, at the moment, you shouldn’t trust it. Yet a guy in New York discovered that his daughter had been talking to Siri for over 2 years and that it had been asking Siri to give her advice on how she should, you know, make life choices. One of the problems there is that these systems actually they they reflect back to you. They’re taking their cues from you. And that was one of the things that
00:09:07
psychologists found really early on with AI systems, the way that they develop them is so they get more information about you by reflecting back to you what you’re saying and you’re unaware of that. Now, that’s one problem. It’s a problem because some people are beginning to use AI systems to psychotherapy. Bringing reflected back to you what your problems are isn’t necessarily a good idea. I mean, you if you think about somebody who’s got suicidal tendencies and who wanted to talk about those
00:09:40
suicidal tendencies, you could actually get an AI system encouraging them to commit suicide. And there have been a few examples of this that have occurred on the web already. And we we talked to a psychotherapist about this who says that that’s a problem. You’ve also got a problem of emotional dependency. Now you could say hang on for some people this could be very very helpful. You could get very lonely people and apparently AI systems are actually better at helping um lonely people. they could be more
00:10:14
empathetic uh than a human being. Which does actually though, Bill, that that throws up a really interesting question, doesn’t it? Well, are you if you’re talking to a machine that I I’m I’m very familiar with um what I think are far too many people who use um screens and other forms of electronic entertainment as as artificial child care. Um, and and what I think you’re probably hinting at here in terms of company is that there is going to be a large number of older people who who desperately need company
00:10:47
and that AI may have a role to play here. But in terms of what you talked about psychology, I believe that you spoke to Luis Skinner, a a psychologist and and a specialist in this particular arena. Um, here’s a a little clip and maybe you can introduce her as as we start a little clip of exactly what she had to say. No, LOAS was taking therapy. Obviously, this is like a concern of that people would not uh get sort of professional help and instead would just feed their problems into an AI tool which obviously
00:11:21
has many issues around confidentiality and privacy and things like that. But also as you mentioned the one of the issues is that um these tools are generally trained to affirm your perspective. So what will happen is it will start to repeat your narrative back to you. And for people who are stuck in sort of patterns that they actually want to break, sometimes that really isn’t what you need from therapy. Or in fact, you don’t really need advice at all. You actually need someone to listen and then
00:11:43
to give you some more thoughts and then to listen again. So it can be Yeah, there can be areas in which this can be quite dangerous, I think. Yeah. I mean there’s there’s Eloise. She very kindly agreed to participate immediately after her yoga lesson, which is the reason why the the you can see her in that setting. uh yeah she’s making this point that you you can actually confide information the other thing is the the these AI systems they don’t have legal entity bill this is a big issue in Donald Trump
00:12:22
who is the US president and in his big beautiful bill he’s saying that AI will get a get out of jail free card for 10 is there won’t be any regulation that will apply to it for 10 years so that AI can develop. It was a position that JD Vance elaborated on in February at the Munich summit. So there’s a problem there. You can’t you can’t sue them. So you can’t sue the the manufacturer. They’re not liable. They could be giving you bad information and you have no comeback whatsoever. Surely
00:13:06
you still have a comeback um depending on which jurisdiction you’re in. I mean, I doubt the European Union is going to follow in in Donald Trump’s shoes in giving a get out of uh free uh uh card for um for for companies um uh for for 10 years for all aspects of what they do in every jurisdiction. I think we’re probably going to see some AI regulation in other key jurisdictions such as the UK or the EU. Um, obviously uh this the the the big beautiful bill has yet to pass in the US and it may yet be
00:13:36
amended. But I I would certainly be concerned if if this provision to to uh take all responsibility from AI and AI companies for for 10 years were enacted because I I think we we need a level of responsibility and we need a level of regulation. I don’t know what your thoughts are. Well, I absolutely agree with you. However, um you know, Kier Starmer and the Labor administration have signaled that they’re not too keen on regulating AI. There is currently a battle going on between quite a few
00:14:09
people in the House of Lords and the government about AI and copyright. And that’s even that bill is an issue for this because one of the reasons why the AI companies want to be able to cut into copyright, they want to actually take on the artists is because they want to humanize that AI interface so that you think it is human that it’s the artists and the creatives who are very very good at creating that discourse. That’s why it’s very valuable and I but this takes us really much into
00:14:47
the sort of realm of ethics and and actually what the motivation is for uh the the large AI companies. Uh and I I believe that it’s largely commercial and and you spoke I believe to uh Johnny Penn from from Cambridge University on some of this. Um uh uh what’s Johnny’s role for Cambridge and and what were his thoughts? He’s he’s the professor of AI ethics and society and he was basically making the point that you’re saying he was actually saying you have to Harry Evans so Harry Evans the very famous
00:15:22
investigative journalist said things are not what they appear on the surface you must always dig deeper dig deeper and that’s one of the things yeah why are these companies offering you these AI system users of chat GPT don’t realize is the system is studying them uh and changing itself to um sweeten the in the the experience of using the product. And the long-term kind of aim here is to be the first port of call between you and online commerce. And obviously the the the fact that um uh the online commerce is the ultimate
00:16:06
goal here is what’s driving all investment ultimately in in AI and and all technology. Um but surely regulation needs to come hand inhand with this. But of course one of the what okay so what does AI do? Uh Peter Kyle, the uh the the the technology minister, UK technology minister earlier this week in the Guardian newspaper, he was saying that Chat GPT knows him. Okay. Well, what’s his relationship with Chat GPT? The problem is that Chat GPT does know him. It knows what he wants. It knows
00:16:44
what he might be wanting to buy. It knows what he thinks that he needs and it sells that information. It sells that information because that’s in the terms and conditions that you’ve given permission for and that information is being passed on to other companies. It’s also letting you have this service and in return for that service you’re giving it information. You’re giving it more information about you and you’re giving it more information that can be capitalized on. Ah, but so what’s the
00:17:16
solution? The solution to this is you should own your own AI. You should own your own data. Data ownership is going to be one of the biggest issues of the next 10 years. You should actually decide how your your data gets used, where it gets used, and you should also get a significant return for that. It’s your data. But unfortunately the the the driving thing behind all of this is entirely commercial and what we’re seeing certainly in the US in in the the legislation there is the fact that the
00:17:54
the large tech firms have the upper hand in influencing what the destination is. Um there aren’t enough uh uh people in public and in society um calling out for uh ethics in in AI and and regulation of AI. Uh uh it’d be far better if there were. Um but to a large extent the vast majority of the population are stumbling blindly towards a a destination that I think we’ll all live to regret. Um uh uh do you think um that there is likely to be um major incidents, things going terribly wrong just as we’ve seen in
00:18:32
cyber security that might be a wakeup call? Yeah, I absolutely do. I think there’s going to be some terrible things that happen. So, just to go to one point that you raised, there are a lot of people who are actually raising concerns about this. We actually have found a lot of people that we’re going to be interviewing over the next few days who’ve got some very relevant points to make about this. The problem is that the government seems to be, and that’s governments across the world, many of
00:19:03
them seem to be wanting to close down this discussion. And I don’t think they’re going to be successful with that. Um to go to your point about the population. No, there there is a very famous quote from the uh literary critic Kenneth Tinan who said that a critic is someone who knows the way but can’t drive the car. The population neither knows the way nor can drive the car. If you take what you were talking saying about maps earlier and that’s got to be changed. Awareness of this has to
00:19:37
be raised among the public at large because they’re actually being duped and they’re being duped by these big tech companies. Well, this is a fascinating topic and and I’m sure that we’re going to return to it and we’ve got there’s a whole set of related issues around AI and its implementations around all of the ethics here and regulations. So, um no doubt it’ll be a topic we’ll be revisiting a number of times. Um I’m grateful for you to share some of your thoughts and and
00:20:06
those interviews that you had with Louise and and with Johnny. Um and I look forward to picking up on this in the near future. Oh yeah. And also though remember the radio program can also be listened to via the same website that you can actually pick this TV program up from. Well, there we are. We are truly multimedia, not only uh TV and radio, but you’ll also have see articles published on some of our websites. So um we welcome you to come back and join us and we’ll pick up this debate no doubt
00:20:39
in the very near future. Thank you. And also for the reasons for reasons of utter transparency. Just before we go, we will also put links up to the interviews that we did so that people can also make up their own minds. Well, if that is if they are making up their own minds and not being uh uh induced uh to do so by some sort of artificial intelligence. But who knows? Anyway, thank you all. See you again soon. [Music] [Music]
