AI: opportunity or danger? Expert clarifies (#10)

00:00:00: Swiss Cyber Security Days Talk, powered by Handelsseiton.

00:00:09: Welcome to today's episode, recorded live at the Swiss Cyber Security Days in Bern.

00:00:15: I'm Tobias Polzon and joining me is Karin Frik, former head principal researcher at

00:00:21: the Gottlieb-Dutweiler-Institut in Rüschlikon.

00:00:25: Karin is one of Switzerland's leading futurists specializing in trends in technology, society

00:00:31: and economy.

00:00:32: She has spent decades analyzing how innovations shape our world and has authored numerous

00:00:38: studies and topics such as AI, digital trust and the future of work.

00:00:43: Today we will explore the impact of artificial intelligence, what should we prepare for,

00:00:50: where should we set limits and how can we balance innovation with responsibility.

00:00:56: Karin, welcome to the show.

00:00:59: Thank you, welcome.

00:01:01: Hello.

00:01:02: When you look back on your childhood, were there any moments when you were interested

00:01:07: in the future, what fascinated you back then?

00:01:11: I mean, I was a child always interested in what will be when I'm an adult, when I grow

00:01:18: up, so what will be, what will I become, how will the world look like.

00:01:23: So, but of course I didn't know the term futurist.

00:01:30: Let's just dive in and talk about the big topic, artificial intelligence, AI.

00:01:35: Do you remember the moment when you first realized that AI would probably fundamentally

00:01:41: change our lives and what went through your mind at the time?

00:01:44: I mean, I really started to work on this topic more than 30 years ago when I started to work

00:01:52: and I started with AI, so just really in the end of the 80s.

00:01:59: It was, I had an interview with Hans Morowitz, who is a leading robotic researcher at the

00:02:07: Carnegie Mellon University and he had this book, "Mind Children" and it was really about

00:02:12: singularity, the idea and Hans was then about, he was in his 40s and he then told me that

00:02:20: when he will manage to live till he's 90 or so, he will live forever because then he

00:02:28: can upload his mind onto a computer and it was really the first topic when I started

00:02:34: my professional work and since then I'm always fascinated with future tech and the options

00:02:43: and yeah, you must, when you think back, I mean, in the end 80s, we did not have Google,

00:02:49: we did not have Wikipedia, we did not have iPhones, just I had a really simple Mac with

00:02:56: diskettes.

00:02:57: I mean, I remember those days in the 90s as well, that's when I grew up with technology

00:03:04: but maybe on the broadest spectrum, I've read that you've said that people normally are

00:03:09: not afraid of technology itself but rather its consequences.

00:03:13: Yeah, yeah, I think this is an important point.

00:03:17: We are not afraid, we are afraid who will use the technology maybe against us.

00:03:23: I mean, it's quite easy to see what's in favour for you, what the benefits are.

00:03:28: Actually everybody used iPhones, one thing like that, okay, it's easy, we use GPS navigation

00:03:34: because you understand it, when the tablets came as well, the older people started to

00:03:39: use it because you don't need to know how to programme, not press any keys but just take

00:03:45: your fingers and do it.

00:03:46: So actually, I think for most people it's very independent from their education, it's

00:03:51: very easy to understand the immediate benefits and so they are not afraid of technology but

00:03:59: you are afraid of it and you see all these health scenarios.

00:04:03: If technology is used as a weapon to manage you, to control you, to control your life,

00:04:10: to take all the benefits away and make you a slave.

00:04:14: I mean, it takes your job, maybe it controls you, it controls your mind, whatever you can

00:04:21: imagine and we are really afraid of the impact, of the negative impact of the technology,

00:04:29: not really of the technology itself.

00:04:32: That brings me right to the next question.

00:04:35: Do you think in general we should embrace AI, the possibilities or should we be scared

00:04:42: of it or at least aware of the risks?

00:04:46: I think, I mean, we should be aware of all the options we get and usually it is true

00:04:52: for all kind of technology.

00:04:54: You can use every kind of technology, you can use it as a weapon or you can use it as

00:04:59: a tool to somehow make things better, more efficient and so actually you can use a car

00:05:05: even to get faster but you can use it as a weapon, every kind of technology.

00:05:12: And I think as we get a very powerful technology such as AI, we should be aware what we do

00:05:19: with it and that's true for all kind of technology.

00:05:22: I mean, yeah, the higher the risk, if you use it, the more you have to think about what

00:05:28: you will do with it and that's true for some technology, we weapons, we need licenses,

00:05:34: we need license to drive a car.

00:05:36: For different kind of, if you work in a factory with big machines, robots, you of course need

00:05:42: a training, how to use it and all of that kind of things.

00:05:46: And so I think technology AI can be a powerful tool but we must be aware that we can use

00:05:57: for bad and make accidents maybe just by accident, not because we want to.

00:06:05: If you had to pinpoint on a timescale, when do you think AI will cause the most significant

00:06:12: societal shift?

00:06:14: I think it already did.

00:06:16: We are already here because yeah, this is the famous quote of, yeah, the future is already

00:06:23: here but unevenly distributed and yes, to some degree, actually if you look who designs

00:06:32: the direction where we go, what we do now, what we discuss, I mean, it's driven by this

00:06:38: future AI and so this is the first decision, I mean, who invests and somebody, I mean,

00:06:46: usually if you have enough people who see possible profits that they can make out of

00:06:52: it, so actually they invest and then you have everything and then you can make money.

00:06:57: I mean, we are a market society and there we need opportunities to make money for every

00:07:05: kind.

00:07:06: The people will follow and do their part, maybe not aware of all over and that's true

00:07:12: as we had before for the whole idea of sustainability.

00:07:16: I mean, you exploit the resources you get and you do some good, you profit because you can

00:07:23: feed more people and so and you see the negatives bill over maybe only years later and I think

00:07:31: this is true as well for AI and we are already there.

00:07:35: We will not go back.

00:07:36: I mean, you cannot really turn off the GPT and it's other agents that are here to come.

00:07:47: You just said agents 2025 will be the year presumably of AI agents.

00:07:55: As AI shifts from say task execution to more oversight, how do you think the role of us

00:08:03: humans might evolve?

00:08:06: I mean, in different direction.

00:08:09: I mean, I still think humans are as well able to learn, we learn, we learn not everybody

00:08:16: in the same speed, but we can learn, we can use it, we can use tools.

00:08:20: I mean, we could use all the other tools we invented before.

00:08:24: So we will be kind of a boss.

00:08:28: Because we all will get kind of slave or servants or assistants and then you have to think if

00:08:35: you don't do the job yourself, you have to think about how.

00:08:39: If you give somebody the job, you have to think more from above, more on the metal level,

00:08:47: because then you are in control of the system.

00:08:50: I suppose maybe not.

00:08:53: But maybe there are two ways.

00:08:55: You can say, "Okay, it's mega convenient.

00:08:57: Let the agent do.

00:08:59: I don't want to think.

00:09:00: I just want to relax and the agent is smart and it knows what I need.

00:09:06: It knows my budget and it will spend it maybe more rational than I do."

00:09:11: And actually, I outsource my life to the agent.

00:09:14: Would be one option.

00:09:16: The other option, as always, when some people are lazy, they just sit down and use all the

00:09:22: entertainment that the internet you can get and think they have a good time doing sweet

00:09:28: nothing, Deutsche Van Niente, kind of.

00:09:32: And the other way is, "Oh, I got such powerful tools.

00:09:36: Where can I get with that?

00:09:37: What can I make better?

00:09:39: What can I make different?

00:09:40: How could I get ahead?

00:09:42: What could I learn if I really have this power to go to the next kind of, if you take the

00:09:50: terms of the gaming.

00:09:52: Let me go to the next level."

00:09:53: And some people, and I think, "These are many.

00:09:57: They aspire to the next level."

00:09:59: I say, "Okay, I have this powerful tool.

00:10:01: I mean, we are here with these tools.

00:10:04: What can I do with it?"

00:10:07: Maybe on a more personal level, I see people rely more and more on generative AI tools in

00:10:14: my circle.

00:10:16: How do you foresee those tools affecting our critical thinking skills over time?

00:10:23: If you're not ...

00:10:24: Yes, there is one concept I think it's interesting or inspiring.

00:10:31: I mean, everybody knows the system one, two, thinking of Kahneman, which means system one

00:10:37: is when you are just emotional and intuitive.

00:10:43: System two is the one when you have to work.

00:10:46: It takes some power and effort to really for rational thinking.

00:10:51: And now there is a concept, maybe there will be as well, we'll get in the world with system

00:10:56: zero.

00:10:57: System zero means actually you outsource.

00:11:00: I don't want to bother.

00:11:02: I don't want trouble.

00:11:04: I don't have now the energy to really bring my mind to this problem, to solve this problem.

00:11:12: But I will outsource it and it's very convenient.

00:11:14: And convenience is always, as I once said, convenience always wins.

00:11:20: Of course, if you don't have to move yourself and you can use ... I mean, if you don't have

00:11:27: to make a fire and just have a switch, you will use the switch for heating or cooling

00:11:32: the room and so on.

00:11:34: And if you have a washing machine, of course, you don't wash by hand because it has not

00:11:41: really sex appeal washing by hand.

00:11:44: Maybe a candlelight dinner might be interesting from time to time, but otherwise not.

00:11:52: So actually maybe we will switch often to the system zero, which is we don't think at

00:11:58: all and we let the system think for us.

00:12:03: This could be a different society where we have ... maybe then we serve only on the

00:12:11: emotion, because it's okay.

00:12:14: We will become much more emotional because all the rational thinking has been outsourced.

00:12:19: And then we think, okay, maybe it's like a child because they hope maybe the mom or

00:12:27: dad, they call me, stop if it gets dangerous or so.

00:12:32: But this is just one scenario.

00:12:35: I still think ... On the other hand, I mean, today we ...

00:12:41: We don't need our body to work, most of us not.

00:12:45: Actually we just could sit here and do our work, our jobs.

00:12:50: And still now we go all to ... We do our sports, you need to go running, we go to the gym,

00:12:56: we do all these kind of sports, actually, but we do it active because we know if I don't

00:13:01: use my body, I will lose my muscles and I will get sick to some degree.

00:13:08: And maybe this same thing is going to happen that we get more and more aware or at least

00:13:14: some first people that you get aware of when you don't use your mind.

00:13:19: You will lose it as with muscles.

00:13:22: And then I think it will be even ... I mean, intelligence is cool and natural intelligence

00:13:29: is even cooler.

00:13:30: Actually, if you lose it, it's like a body, which is not fit.

00:13:36: Your research also shows that technological progress unfolds differently or more slowly

00:13:42: than initially expected.

00:13:45: What are some current AI hypes that might not materialize as expected?

00:13:51: I think it will take much, much longer because they are ... As soon as you get to more critical

00:13:59: problems, let's say really health decisions, the AI is not there yet.

00:14:07: You already see when you ask today, LGBT or perplexity or whatever, you can do it for

00:14:13: some research and you see the mistakes.

00:14:16: If you are a note of the discipline, you are asking, you realize all the mistakes, it's

00:14:23: making and saying, "Okay, if it would be a decision about my health, I really would

00:14:28: prefer somebody who makes a decision, who is more sure about what he or she says or decides."

00:14:38: And so, I mean, as I said, we talk these ideas and we really find back in this text with

00:14:46: Hans Morowitz for more than 30 years ago, all these ideas were there.

00:14:52: And even the technology was far away, but we are still not yet there.

00:14:57: And I mean, you can see, you can be amazed what's possible today, but it's still, I mean,

00:15:05: really difficult tasks.

00:15:09: It takes time, it takes energy, we don't know if we have enough data and so on.

00:15:15: I mean, we are amazed, but it will take longer.

00:15:21: And then you have all these problems.

00:15:23: The question is real ability, I mean, when it makes a mess, who is responsible?

00:15:32: Let's close the circle.

00:15:34: If you think about year 2035 and imagine how you will explain the world to a young person,

00:15:43: what will you tell them about the development of AI?

00:15:46: I mean, they know more than me.

00:15:49: Actually, today, my grandchild was just born.

00:15:54: I mean, she grows up in a world with just AI.

00:15:57: She does not know a world without, I mean, Google for her, that's really, like for me,

00:16:03: a kind of very old phone or whatever you can imagine.

00:16:07: I mean, she will not know a world without machines that answer her or maybe, yeah, mirror

00:16:16: us, know what she thinks, what she wants.

00:16:19: I mean, you don't have to explain them because that's kind of, they understand the world

00:16:26: they are in.

00:16:27: And I think the youngsters then, they are even more critical as we see it.

00:16:33: Yeah, it's kind of the next generation AI, maybe we need to call them.

00:16:40: Thank you, Karin Frik for your insights.

00:16:42: Thank you.

00:16:43: And thank you to your audience for tuning in to this episode from the Swiss Cyber Security

00:16:48: Days in Bern.

00:16:49: If you enjoyed the discussion, don't forget to subscribe.

00:16:52: And until next time.

00:16:53: (chimes)

00:16:55: [BLANK_AUDIO]