We caught up with Eberhard Schoeneburg, CEO of Cognitive Systems Lab - LIVE from Hong Kong FinTech Week.
Walter Jennings: Welcome to Waves in the Finoverse. I'm Walter Jennings, the host of a podcast brought to you by Finoverse. We're talking with the wave makers creating ripples, waves and tsunamis across finance, crypto, FinTech, Web3 and beyond. Listen weekly to hear the changemakers talk firsthand about their experiences in this dynamic industry. Thanks for tuning into Waves in the Finoverse. I'm your host, Walter Jennings. I'm joined in our podcast booth at Hong Kong FinTech week, by Eberhard Schoeneburg, Chairman and CEO of Cognitive Systems Lab. Welcome to our show, Eberhard Schoeneburg
Eberhard Schoeneburg: Great to be here.
Walter Jennings: Thank you. I understand that you created chatbots in the 1990s. But you said in the past, they never fulfilled their true potential. Could you tell us a little bit of the story of the invention and what missed potential they've had?
Eberhard Schoeneburg: Yeah, it is actually surprising for many people that the bot technology is actually quite old already. And I was one of the pioneers of the whole technology in the 90s; as you say it at the time, there was really many, many technical limitations, you know what you could do? And they still are there, we re-experienced this now with the metaverse what happens when you scale up and you have millions of users, you cannot do that. In my time, then, in the 90s, when we had like five concurrent users that was already huge. So the chatbots had lots of limitations on the technical side, on the banquet side, but mostly on the intelligence side, you know, because there was not this big progress that we see now in artificial intelligence.
You know, I'm one of the AI pioneers as well, and not just doing bots, bots was one of the key things I did. So we had to hand programme all these eSports, the avatars, you know what they say? And it was a very elaborate task. Now, you can automate a lot of this.
Walter Jennings:
So you're scripting versus today, now you can have predictive language, etc.
Eberhard Schoeneburg: So scripting the dialogues, which you still see today, you know, but it's just not a business model that works when you need 100 people to build up a bot or so that makes no sense. But nowadays, things have changed dramatically, especially since about 2017-2018.
Walter Jennings: Eberhard, your company is Cognitive Systems Lab and Cognitive technology, from what I understand is when a machine does what only a human could have done before whether that's see, smell, think, hear, walk, is that? Am I oversimplifying?
Eberhard Schoeneburg: No, no, no, that's the problem is the descriptions is easy. And the questions are easy, the answers are just horrible. So it's really what I'm trying to do. As I said, you know, I'm one of the surviving dinosaurs of AI and doing it since 40 years, and we have seen a lot of progress in certain areas in other areas, it's just horribly frustrating. Nothing's progressing. I remember when it all started, for example, with autonomous cars, vehicles. We had the first autonomous cars already in the 1970s or so you know. And then there was a big hype, especially in the early 2000s. A huge hype the last 10 years or so.
But you see it stalling now, because, you know, it's not so difficult to go to 60-70-80% of, you know, autonomy. But the last 20% especially the last 2-3%, you know, that make these things absolutely secure. It's nearly impossible to achieve. And that's why it's stalling out. Nobody predicted now with Elon Musk. Elon Musk has predicted it, I don't know 10 times, you know that we would have self driving autonomous cars now. It's not happening right now. Right?
So there's all these ups and downs and backs and forwards where you really don't know, you know. And that's the problem with intelligence. Right? It's intelligence is a very, very tricky topic.
Walter Jennings: Well, Eberhard, just with the self-driving cars, I think we've all seen massive improvements in driver-assisted technology. So while we may not have reached the ultimate goal of driverless, we're at the penultimate goal of driver assistance. So perhaps a major motor companies are recognising we might be at that assistant level for a few more years.
Eberhard Schoeneburg: But we have been at the assistant level for 30 years already, a lot of especially the German, right, a lot of the German cars have driver assistance since literally 30 years. The problem is when you over promised and then under deliver that as a whole is a big problem, right? I would rather do it the other way around. I don't promise too much and then try to overachieve somehow but..
Walter Jennings: Now, we're here at Hong Kong FinTech week, what are some of the more promising artificial intelligence applications in FinTech?
Eberhard Schoeneburg: Yeah, I mean, that's another area I've been working on a lot. It's obviously, everything related to banking. You know, the whole Metaverse issue that we're discussing, and I will talk about later today. The issue is that you can relate AI to all these new FinTech technologies, where if you walk around here, you will see that most of these boots and startup boots there, especially everybody says they do AI, I doubt it, to be honest with you, I really doubt it.
And, but it is obvious that if you had AI and the right level of AI, you could definitely make a lot of things much easier, better, faster, more efficient. And that's where you know, the trend goes. So for example, I've also installed the first robo-advisors, for example, for financial services, like for banking, retirement planning and stuff online. And a lot of companies go there now and then all the blockchain issues, you know, that make things more secure. And so, so there is a big overlap of FinTech, financial applications and AI, and the whole Metaverse thing on top of it. So there's a lot of synergy coming up right now that's the most promising thing.
Walter Jennings: And within Cognitive Systems Lab, is AI, the primary focus or are there a wide range of other applications you're looking at.
Eberhard Schoeneburg: We are going far beyond AI. So classical AI is always focused on computer science, right? So you think about building intelligence in a computer system. But I mean, if you look around in our world, a computer system is a dead system, right? It's not alive. But most things that are intelligent are alive, I would say every everything that is alive is intelligent in a certain way. So we are trying to go more towards biology oriented and based models that are not relying on computer architectures and server farms and stuff like that, right.
So I'm working with some of the top people in the world to build, actually new types of brains literally, you know, not simulate brains building new type of brains, so called organoids. These are brains mini brains that are derived from stem cells using body cells and union, you reprogram them, and then you're differentiate them so that the body cells can turn into brain cells. So you can grow brain tissue, and then you grow mini brains into the size of a pea a little bit bigger. And then you can connect these mini brains and make computing that way. So it's a completely new revolutionary thing.
Walter Jennings: That sounds then like attaching biological material and to..
Eberhard Schoeneburg: My bio computing, literally bio computing without computers.
Walter Jennings: Fantastic. Well, I'm wondering how you keep the processor fresh and
Eberhard Schoeneburg: Oh, it is actually a big problem, right? So you have to use the brains and the problems why? Why you can grow brains, like the size of a cat brain or mouse brain or human brain is the vascular system. So you can grow the tissue and you can go 3d. So it's literally like a brain. But the problem right now is they haven't figured out yet how you can go rhodium so that they also maintain their own blood flow. So that is the only limiting factor right now we have, but, you know, over time that's going to happen.
Walter Jennings: Now, I've read about a concept called alternative AI, can you explain that?
Eberhard Schoeneburg: This goes in the same direction, that's a term that I've raised, and I invented that whole area also. So my first real AI company was called Artificial Life, but artificial intelligence and an artificial life, what you try to do is you try to build alternative models of life. So,with, you know, people trying to go to Mars and so on. And the whole topic of what is life? Where can life be detected? How can we detect it and so on its own what are the key features of life, for example, is intelligence, an unavoidable consequence of life? These kinds of questions are extremely interesting. And we are trying to build emulators, simulators and so on for artificial life environments.
And the metaverse is one very interesting concept as a basis because you can think of a Metaverse as kind of a simulation environment of artificial life. So it's a confinement in which life can develop in completely different and new ways.
Walter Jennings: You've written extensively about whether machines can develop a consciousness, what are some of your views on the subject?
Eberhard Schoeneburg: Yeah, that's a very tricky topic, as you know, the question is, what is consciousness? Right? Can you even define it, and there's people out there that I know are thinking that language models are conscience, which is in my eyes, completely nonsense. But the problem is, you know, it's hard to define what consciousness is, but it is relatively easy to emulate conscious behaviour.
And I think that's what you will see a lot of systems like for example, avatars or bots, that behave as if they had a conscious now towards you, if you communicate with them, or even for entertainment purposes, right, they will have a character, they will have a feel to it. And you will see a lot in the future a lot of applications that will make use of this that you have kind of a personalization and humanization of technology in the sense
Walter Jennings: Yeah, big fan of contemporary literature. And one of the Booker Prize finalists last year was Clara and the Sun by the Japanese author Kazuo Ishiguro and there they discussed the main character is a humanoid who comes to life in the sun and then follows her journey. And whether or not that's consciousness or not, we're not quite sure. What do you say to those who are worried AI technology could get out of hand? Why is it? Is it important to build a failsafe?
Eberhard Schoeneburg: Yeah, absolutely, I think. But in this respect, AI is no different than any other high tech, you can turn any high tech device machine into a weapon if you want to, I mean, you can use your pen and kills.
Walter Jennings: I'm not James Bond. So I think you're safe for the moment. Now you've said intelligence is like music. What do you mean by that?
Eberhard Schoeneburg: I wouldn't put it that way. But intelligence and music have a lot of things in common. And it's, you can use musical concepts to describe intelligence. For example, if you have to play an instrument?
Walter Jennings: Just the voice trained.
Eberhard Schoeneburg: It's good. But the point is that, you know, music has a logic. You know, I don't know, if you feel that when you listen to a piece of music, especially classical music, listen to Bach, you can predict the tongues that are coming up and down in perfect harmony, right? And there's a logic you can just put node x, after node y, it has to match somehow. Right? And how does it work? And what causes that feeling for us that this is the right tone, right? This is the right harmony? And another one is not as disharmonic what why do we begin? Where does it come from? What causes these effects and then interdependencies? And that is very similar to thoughts, you know, how thoughts emerge in your brain? Why is one thought following in other thoughts, you know, it's...
Walter Jennings: Eberhard, what's your greatest wish now for your work?
Eberhard Schoeneburg: Oh, that I don't know, I'm so long in this period. Hopefully, that some of the things that we're working on is not a waste of time and then it doesn't turn into any bad things. You know, at the end of the day...
Walter Jennings: Any technology you're looking out for that you think will be game changing coming in the near future?
Eberhard Schoeneburg: I think brain to brain, direct brain to brain communication is the coolest thing you can imagine. So, you know, we have nowadays brain to computer interfaces, where you can read brain activities, EEG signals, and so on and convert them into computers. But then if you can do that direction, you can also do the reverse direction, like you can have the computer directly talk to your brain.
And if you combine it, you can have one brain talk to another pathway without talking literally, you can transfer thoughts, you know, in a certain way, there is no already solutions where you can see what someone sees. So I can see without being next to you, I could be somewhere, I could see what you see in your eyes with your eyes. And that is just fantastic, man. If you think about that, I could feel what you feel, for example, things like that. That is the real future of AI.
Walter Jennings:
I'm looking forward to that. But the before that though, I'm looking forward to LinkedIn updating me about all the people I see on the floor here and reminding me of their names and companies. So no, it's a it's quite an exciting world ahead. Well, thank you Eberhard Schoeneburg for joining us today on Waves in the Finoverse. Chairman and CEO of Cognitive Systems Lab. Thank you so much. Appreciate it. Thank you.
Eberhard Schoeneburg: Thank you.
