Subscribe
Share
Share
Embed
Born & Kepler is named after the mathematician and scientists Max Born and Johannes Kepler. This bilingual podcast, offered in both German and English, dives into the expansive world of Artificial Intelligence (AI), exploring its foundations, evolving technology trends, academic search, and its impact on businesses and society.
Born & Kepler will feature a diverse lineup of experts from academia, venture capital, private equity, journalism, entrepreneurship, CTOs, and policymakers. Each guest offers unique insights into how AI is reshaping their sectors and what we might expect in the future.
Our goal is to provide a deep understanding of the core principles and breakthroughs in AI, enabling you to stay updated with the latest advancements in AI technologies and how they are transforming industries. During our episodes, we will explore how AI is influencing business strategies, optimizing operations, and driving innovation. We will also explore the ethical, social, and regulatory aspects of AI in everyday life.
Wolf Ketter (01:14.573)
Thank you for the invitation Andreas.
Andreas Deptolla (01:16.718)
Thank you. Maybe for start, us an overview of your academic career and especially the connection to topic of AI.
Wolf Ketter (01:30.035)
Yes, that's it. So I never planned to become economics computer professor That's how it so often in life. a portion of coincidence, also inspiration. It was more or less a zigzag course into the business school. Yes, started studying electrical engineering in news technology.
Andreas Deptolla (01:33.55)
you
Wolf Ketter (01:56.915)
And when you work it was like that. I it at the end of 80s, beginning of the 90s. And then was like that, I to hardware. But I did But over time, more software was into electrical engineering, programming. I also fun doing that. But I really to touch something, build something. But after all these years, when the file...
so that you so-called programmable logic gates. And then I realized that it would actually nice if I understood about how to build large systems in software. And so I for a scholarship in the USA at the University of St. Thomas. about large-scale software engineering. That was then one of largest software engineering programs in world.
And then I started software engineering at the master and then had a job at Lawson Software and PeopleSoft. And the software was originally written in English because they from the USA and they wanted to internationalize they to, if you this software then everything is in German or whatever.
And we two other so-called software architectures. And about 150 other people programmed And then it clear to that nothing is constantly changing. And some point, the software systems have to be renewed. And then we thought, okay, that would really cool if there an AI that that for us. And then I just got more and
Andreas Deptolla (03:39.682)
Yeah
Wolf Ketter (03:53.235)
in the subject. And then one of my female professors at the University of St. Thomas, Bonnie Bennett, studied at the University of Minnesota and did her PhD at Maria Ginni. And that was really pure luck, because Maria Ginni, even now, she's turning 80 soon, she's still active and she's one of the co-op-fans of the AI worldwide. And was really for me...
is a distance less than Cologne and Bonn. And then I just moved to Minneapolis and studied Art Intelligence there, the focus on learning. And there I focused mainly on electronic markets. And we had a startup that, it also about procurement, supply chain management. And then I of the ideas of our startup.
I But then I realized, similar to the programming, in electrical engineering, we also VWL BWL in the secondary subjects, but of course not too many lectures. So I and BWL and VWL So I a double promotion. And that was then one thing another.
Dr. Alok Gupta, one of the leading business informatics in the world, will a fellow Seattle Gerhard where I will fly to. He said, want to professor in business school. On a Sunday morning Las Vegas, I a lecture I nobody I in this room and then I a coffee and I came back there 150 people. I was wow, what's going on
Andreas Deptolla (05:37.678)
Thank
Wolf Ketter (05:46.905)
After giving the lecture, I 10 job offers. Six were from University and four the industry. One of was from the Rotterdam School of Management in the Netherlands. Then I thought, after eight nine years in the USA, would a good time to return to Europe.
Andreas Deptolla (05:54.606)
You
Andreas Deptolla (06:02.921)
Mm. Mm-hmm.
Wolf Ketter (06:14.695)
That's we did. We then returned and then back to the Netherlands in 2007 I set a laboratory, Learning Agencies Research Group, a machine learning group, research group in special area of artificial intelligence. And all applications focused on electronic markets, especially in field of supply chain management.
Andreas Deptolla (06:41.582)
If you the swing to today, you are in Cologne, what the current highlights, what is your research, what is the teaching?
Wolf Ketter (06:54.611)
Yes, so I am actually in Cologne, if you want to or in one or the other then will probably to a lecture or seminar with me, around. So, my teaching chair is called Information Systems for Sustainable Society. teaching chair is a computer for a sustainable society.
develop, design new algorithms, also marketplaces, e.g. marketplaces, and integration of BWL and VWL and mainly this, 90 % or more, to sustainable energy and sustainable traffic. These are really 90 % of our applications in the area of energy transition and traffic transition.
And about 10 % are generally in the field of human-machine interaction. how can the human and AI work together? And I sometimes an article that we had discussed which also inspired by Bill Joe, the co-founder of Sun Microsystems. He an article in the Wired magazine, I think in April 1, 2000.
published a very famous article, a seminal article called Why the Future Doesn't Need Us. And Bill argues there that we over the last few and that was 2000, that some point the machines, the the essential characteristics of people take namely thinking. And that then, way, it only the hill
Andreas Deptolla (08:30.946)
Mm-hmm.
Andreas Deptolla (08:48.014)
Mm-hmm.
Wolf Ketter (08:51.763)
And I don't agree with I just think that AI has incredibly great possibilities. Of course, you to be You have to that a certain extent. We now also have the EU AI Act, which has course had by the various parliaments and member states. But I think that's a good instrument.
AI has an incredible number of possibilities. But it's like with any other technology or biochemistry, that you just have responsible. As they say, responsible usage. That's important. But it's also really, as they say, even a teaspoon can be used So you have responsible with the technology.
We already in engineering studies I started. beginning of the 90s, the professor always told no what technology you as engineers are responsible for how you use it and use it. And if you look at AI, this week, the Nobel Prize in, so all Nobel prizes are awarded and the economic prize, I think, Monday. And yesterday was the chemistry Nobel prize, the yesterday was the physics Nobel prize.
Andreas Deptolla (10:06.263)
Mm-hmm.
Wolf Ketter (10:17.139)
And I never thought that as an AI researcher I would see that really computer scientists, so colleagues of mine, there are actually three colleagues of mine, or if I remember correctly, he is not really computer but he got it for neural network, will four Nobel prizes. These areas of physics and chemistry. That's absolutely fantastic. That's unique. And then you can also see which social advantages are That's why when I work
in the field of chemistry, the protein folding of Dennis Hassebin, then it's unbelievable. You would never have that this protein folding has taken years to develop. And here with the algorithms, can this in shortest time, 200 million different protein species.
Andreas Deptolla (11:15.168)
Awesome.
Wolf Ketter (11:16.107)
You can all proteins and that is only possible with the help of so-called machine learning. That's just fantastic. And only with can develop new drugs, for have better active ingredients. So that you for example, protein and also specifically somehow to the organs, to doctor, so that they just work So I think that's unique or also right now with
For example, ChatchatGDP. There is a lot of hype about ChatchatGDP. You should also see that the AI is a giant... You can imagine as a giant ball. In there is a part quantity, that's the machine learning. there are the so-called deep neural networks. And there is ChatchatGDP. ChatchatGDP is a very small part of all these big beings that I...
artificial intelligence. But nonetheless, is important part, this generative AI, generative AI. And of course, is also an instrument. You have to be careful with the results, of but suddenly people who may or may democracies, may access to knowledge and also documents translated.
Andreas Deptolla (12:41.058)
Mm-hmm.
Wolf Ketter (12:44.147)
where they weren't trained enough before. That's not mean, because they can't anything because they access to the training system. But with these technologies it's like a door opener, window to world. That's you shouldn't all these technologies. You should see how every new technology makes them useful and in good use. That's important.
Andreas Deptolla (13:14.382)
You just said that this core thesis, that AI is smarter than us as humans and that it some dangers that you might 100 % with. What are the see with AI? What can AI not achieve?
Wolf Ketter (13:39.251)
Well, that's hard to say. There were colleagues of mine, I was also asked to sign, but didn't do But very famous colleagues of mine, a few hundred in the USA, AI researchers, voted for a half year to
Andreas Deptolla (13:58.478)
Mm-hmm.
Wolf Ketter (13:59.987)
I honestly don't so. It's a bit like the James Bond movie, where the evil just on developing. As soon as the Pandora opened, it's difficult. Then you have to keep going and it really well. In my opinion, And just to more in the areas. And all, have to lot in Europe.
Andreas Deptolla (14:07.028)
Yeah.
Wolf Ketter (14:30.003)
I have a member since January 1st year in the expert commission of the Oslo Fund Align, the EU commission, mainly in the areas of energy, mobility and climate. But there I am also responsible for the core areas, machine learning, AI. And then we also wrote on how the expenditure for research in next years, until 2040, 2050, should be distributed. And of course, very large
eye on the area AI. That's I often with the investments we in Europe, but especially in Germany, not all them. They not sufficient. I am not a politician and I also politically neutral. I to any But I have say that this black zero is not particularly good Because really to large-scale, very large investments.
especially on the field of infrastructure. This goes into the hundreds of billions, especially on the field. We talked about AI, then we need a very strong broadband, but mobile internet expansion. We need other infrastructures like road construction and above all we need rail construction if we the traffic walls. Only through these three
infrastructure measures that have strong influence and that also cost lot of money We don't to prepare anything. But every year when we don't manage to you can't necessarily count but somewhere, I a time last it costs 9 to 10 billion. And that only up.
Andreas Deptolla (16:13.358)
You just your career. If you reflect were there any projects or things you focused on that maybe failed? Maybe there were some learnings or breakthroughs on? Are there any topics you think about?
Wolf Ketter (16:42.355)
Yes, of It was briefly, but I think if you say that some projects, not all projects It's like in life. You start some things, set big hopes and some become very, very good and some not. And vice versa, it's also where you think, I'll now and then develops into a mega project. It's not always all that predictable.
But a project that didn't work out but I a lot from is that I the of 2000s, speak. It was an electronic procurement platform.
And there we up electronic market and made a matchmaking between sellers and companies could offer their most powerful capacities example services or products. And other people say, okay, I need so so many bottles or heaters. And that was a very good idea. Even to this I see this
Andreas Deptolla (17:58.158)
Mm-hmm.
Wolf Ketter (17:58.387)
Especially in B2B, not quite But we had a fantastic start-up board. We the N Windblatt from Hammond Windblatt. That's really one of the most successful venture capitalists in the USA. We the former governor of Minnesota, Jesse Ventura. We had a top lawyer, Ryback from Minnesota. It was just a great time. We worked
But you learned and even if the project was not successful, it was a good foundation, including for my doctoral studies and for being my professor. if I example, among other things, Executive Education, so when I middle to top management, it's worth if you... I founded two companies over time, I'm not a serial founder now, but I founded and I have...
Andreas Deptolla (18:44.162)
Mm-hmm.
Wolf Ketter (18:58.279)
Many people as advisors or researchers in industry, as engineers. So from that perspective, it helps.
Andreas Deptolla (19:08.11)
Often it's the with these marketplaces that the startups either don't the supply or demand side pulled That's a difficult thing. Was it because or was there some other problem?
Wolf Ketter (19:22.899)
Yes, think it was simply because that. Yes, think it was simply because that. Yes, of course, that was maybe one reason to it up, but the other was simply that we were really unhappy with the timing. That was end of the dot com bubble and then some point so many startups just really from the table. That was just an unhappy timing.
Andreas Deptolla (19:39.159)
Mmm.
Wolf Ketter (19:50.587)
I think if we that a few years later, it would have through the roof. But I didn't regret as I said, we learned We had a lot of stress, but then you learned to deal We met a lot of interesting people, across the United States, pitched as you imagine it, elevator pitches and things that.
Andreas Deptolla (19:51.832)
Sure.
Wolf Ketter (20:15.919)
So it was an exciting time. Ten years later we did another startup That was much better. But I myself out of Because the startup didn't anymore. I pulled myself out of because I on my academic career. Because I to my so-called tenure track. So you first assistant professor, then associate professor, full and share professor. And then I to concentrate it.
Andreas Deptolla (20:37.774)
Mm.
Wolf Ketter (20:44.423)
But I always in touch and have worked with the industry. That's I'm today.
Andreas Deptolla (20:51.054)
I find very exciting, this connection between research and industry. think we have Germany. Thank you again for your experience exchange here.
Wolf Ketter (20:59.73)
No.
Wolf Ketter (21:04.851)
So, I can maybe say something about that, because what I really in the USA, but also actually in the Netherlands, where I lived and worked is that I was very, we just about failures, as they say in America, and this culture of is in the USA, but also the Netherlands.
Andreas Deptolla (21:26.818)
Mm.
Wolf Ketter (21:36.071)
much more forgiving.
Andreas Deptolla (21:38.562)
You know.
Wolf Ketter (22:04.659)
And in the and in the Netherlands it's opposite. They cool, he tried He really tried it. He did And then you learn outside. I actually wrote a LinkedIn post by a lady. I forgot. The co-founder of Zweihorn Energy told something or less similar. They up after a year, but she said, okay, I've so much now.
Andreas Deptolla (22:12.332)
He did something!
Andreas Deptolla (22:26.531)
Mm-hmm.
Wolf Ketter (22:34.899)
I learned and that just me for new things. I mean, you yourself have started to say, Andreas, you know that even better than me. I think that's especially when you're younger. I've always said we only this one life, right? And I'm more a risk taker, in positive sense, not totally crazy risks, just trying out things where I think, that's really a super interesting thing. That could impact.
and then I it, even if I aware that it can wrong. For example, when I from the USA and did a in Business School. Whenever you a top business school, I had in February 2007 with the research dean and the feature research dean in the Netherlands at Rotterdam School of Management. And then I told what I wanted to
also from the AI side and from the sustainability side, energy mobility. And then he said to Wolf, you do great work, but abundant is research. Abundant is, this is way too risky. So I said, thank you much, I'll do anyway. then years later I him, well, I him all the time, but once in a And then he said, about the research, it's really pretty good. Maybe we something together now.
Andreas Deptolla (23:45.176)
You
Wolf Ketter (24:02.675)
For me, when I believe I'm about You can a lot of but if always say, okay, that doesn't then it's useless.
Wolf Ketter (24:25.885)
So, I think we're back.
Andreas Deptolla (24:28.494)
So, now it's little bit rumbled, we can out I think that's also an interesting topic, you're talking about risks in life. Maybe you can also talk what you with your students, with top students here Germany. We a few weeks ago. What's that, I'll in the AI sector, we're actually very good
research but let's say, PS, when it to startups, which don't really on the road and which ultimately lot of economic growth in Silicon Valley. Or always, yes, over last...
Wolf Ketter (25:10.802)
Yeah.
Andreas Deptolla (25:11.31)
20 years and we Germany, except SAP, didn't much new in which then was scaled.
Wolf Ketter (25:25.145)
I think there are probably more things, but two things that come mind are actually the risk readiness of the students. I often have top students, as I do They are really very, very intelligent. But often this risk readiness, which is either to become professor at a top business school
Andreas Deptolla (25:34.584)
Mm-hmm.
Wolf Ketter (25:54.291)
or anywhere else, universities, engineering sciences or to start-up customers. That is still not so widespread Not that you say, that's coming more. Definitely, I that too. We also the so-called innovation dome We rebuilt it two or three years ago. It's here now.
Cologne, was also a ranking, I think, a few days ago, is now also the third best start-up university in Germany. So that's good to Another thing that I really notice, where we strongly behind in Germany in general, that is true for my university, but in general, that is really the promotion of companies, so the industry. We are one of the few teachers who really
Andreas Deptolla (26:44.974)
Mm-hmm.
Wolf Ketter (26:51.347)
actively and really deeply working with the industry. On the one hand, we basic research, which we relatively high or at that's always our goal. And then, where many researchers and professors stop, but we, especially me, that's also my mission, want to feed
in the industry, but also for society. Because that's concern, that we contribute to a sustainable society. And so I've quite well with companies to work and also there, for example, to the so-called field experiments, which are really for the absolute top journals like Management Science or Nature. Yes, you just for
And, but many companies here Germany don't have this culture that you universities or that you base So that's coming now. This culture is deeply anchored in the USA. For example, the one university where I was, University St. Thomas, was one founder of Best Buy, Bill Schulz. I now
Pimal Down, 52 million US dollars and the other one donated half a billion. And there are several. And then they just a new campus or new research center. So without public... And they can really afford The only thing that takes is... Then it's called... I don't know, Smith Hall, somehow.
Andreas Deptolla (28:29.858)
Hmm.
Andreas Deptolla (28:44.131)
Hmm.
Wolf Ketter (28:44.148)
So if I to building B or Smiths or something, I know what Hall is called, Collins Hall, then I don't that's, assuming that they all integral people and companies, which they are all given. So that's just one thing that's in Germany. And even now big companies, I also had, even in the
Andreas Deptolla (28:53.016)
Mm-hmm.
Andreas Deptolla (29:05.567)
That's missing.
Wolf Ketter (29:11.379)
I had a lot contact with Bosch, example, a global company that we here Germany. Bosch actually invests, has its own research centers, two or three top universities in the US, including Carnegie Mellon or the University of California Berkeley, where I was a professor I can understand why, but on other hand, we also top talent in Germany.
Andreas Deptolla (29:14.689)
Mm-hmm.
Andreas Deptolla (29:28.502)
Mm-hmm.
Wolf Ketter (29:40.499)
say, only in the USA do people top performance, then it's bad. Even our own German companies there in the USA, the research centers, for example, have these Bosch Labs or other, I don't know, Siemens or other companies. Then it difficult, So that's what I as a
Andreas Deptolla (30:04.302)
You have the view of companies, universities, what they can What between the is that students take more risks, things that are not a classic way. Which other advice would you to...
and ambitious students in Germany who are now in age of AI. And maybe a second question, what common advice would you give that you would simply ignore?
Wolf Ketter (30:47.323)
That's always nice with wheel hoses. mean, there's a very famous quote from Oscar Wilde. The only thing you can do is advise to give it further. That you just keep which doesn't follow. But well, one thing that at least has, I can people what worked That's not the case, because I think that, well, everyone has to individually. But I would actually recommend if it works one to two semesters, best two.
Andreas Deptolla (31:04.194)
Mm-hmm.
Wolf Ketter (31:16.179)
But least once to study Then it doesn't where abroad, but that you just get this experience. It's not just studying. It's also just a different culture, it together. I think if there were really a more people, not from Germany, so that's just a worldwide phenomenon, then we would have...
Andreas Deptolla (31:18.89)
Mm-hmm. Mm-hmm.
Wolf Ketter (31:41.683)
much more understanding and less, well now I may bit naive, but less these conflicts worldwide. So if people have much more exchange, then they for example, in the USA years, 10 years the a year in England, I actually lived in Russia then it's just such a thing, if you know that they actually not bad people. that is of course the political system that sometimes many.
Many people, especially educated people, want their countries, or are already leaving. And that's a problem It's difficult. think this really opening up, going for a semester or a good university, also taking with friendships, that's an experience that you will
Andreas Deptolla (32:29.9)
Mm-hmm.
Andreas Deptolla (32:36.428)
Network connection,
Wolf Ketter (32:38.835)
Yes, and one thing that I very important, especially in the top magazines, in the top journals worldwide, is always said, you should interdisciplinary. But in reality, many these top magazines are still very monodisciplinar. But nonetheless, entire research, since the beginning, is highly interdisciplinary. So especially now the
If you at the energy and traffic syndicates, they are not only electric engineers, are only energy economists, are also psychologists, they are lot of informatics, they mathematicians, they physicists, are businessmen, they are people's businessmen, new markets. That is also the exciting thing. We are now in a time of great transformation of many systems.
the transformation of the health system, transformation of the energy system, the mobility, the cities, the cities of future will completely different. It's a difficult time we're living in, it's not easy, not like all the conflicts, also climate change. That's really the goal of our group, that we contribute to this so-called wicked problem.
to some solutions. We can't the whole problem, but can least try parts of it. And they all interdisciplinary. And I'm just to every day because you so many interesting people who in many areas than one person. Or at least they more in some areas. And I know things they don't know. And that's just the case with these
Wicked problems are climate change, health systems, systems, of These big social, also economic, technical phenomena. can't easily. In business school, often the two-by-two matrix.
Wolf Ketter (35:03.603)
So, that's a problem now and there's a solution to That's not the case. For example, I had an interaction with Ms. Merkel, former chancellor in 2008. She said, they always these serious games, they could make for the energy transition. And that was one the...
Andreas Deptolla (35:04.073)
Mm-hmm
Andreas Deptolla (35:18.274)
Mm-hmm. Mm-hmm.
Wolf Ketter (35:30.803)
So an initial ignition for our platform Powertech. So if you're interested in a serious game for the energy transition, you can take at it. Powertech, so power, so P-O-W-E-R, then tech, T-A-C, dot org, powertech dot org. That's an open source platform. You can the software and can basically simulate the future in the field of energy mobility. Interesting thing, we have also been
world championships, dozens of universities and research institutions and companies that in the whole world. And now we in the next phase, we are doing specific research projects. That is really very exciting. I think 600, 700 people it, and master's A few dozen doctoral graduates have been written about it, several startups have been founded. So that's where I'm very happy.
Andreas Deptolla (36:25.464)
Mm-hmm.
Wolf Ketter (36:28.307)
These are projects that far beyond research and magazines, i.e. research publications. They really have a social and economic benefit. And I'm very happy about that and thankful that this has
Andreas Deptolla (36:45.166)
You do a lot in the field of energy. Maybe you can explain why the AI is so energy-efficient. There are articles that say that data centers in the need as much power as a small town. What is the background?
Wolf Ketter (37:07.48)
We should distinguish, Andreas, which kind of AI.
You shoot with guns at spats, If I like this week, the Nobel Prize for various AI researchers, they have course very complex algorithms and turn them into very complex problems. But we can often use very simple AI technologies, for example, very simply, which is not even AI, but often a regression is often done in many
Andreas Deptolla (37:42.913)
Mm-hmm.
Wolf Ketter (37:43.125)
We don't need deep reinforcement learning or other large-scale ensemble methods. I had the lecture for our Master's Learning at the University Cologne I said that we to a real sense of and knowledge of where to use them.
Especially this chat GDP, the generative AI, which is very power-eating, or this blockchain mining, is very power-eating, so that's not all. And you just it sensibly, but for the energy transition you have to make that you can often use so-called light-weight algorithms. you don't need a huge data center for On the contrary,
Often, things locally relatively easily, because you have to able to react What has happened over last few we going? We from a demand-driven energy landscape, society, into a supply-driven society. Originally, we had a vertical integrated energy supply, where a company
electricity, example, produced, then distributed, then to customers, made and over the years this network has been and now, and that was all fossil fuels moving, coal was burned or nuclear power, we don't anymore. And what we have do now is course lot of wind, solar, water power.
Andreas Deptolla (39:26.796)
Mm-hmm.
Wolf Ketter (39:36.787)
But then we no control over when electricity is because they very dependent And if that is case, then have to set when the energy is available. And if there so much available, can it little cheaper. Because energy must always be a very tight balance.
Andreas Deptolla (39:41.26)
Mm-hmm.
Wolf Ketter (40:06.259)
of demand and supply. If that's not the case, we'll get so-called Brown or Blackouts, and we don't of course. And could hours, we don't Maybe we can deeper into But we need AI that then acts in because of dynamic prices. So we really have working on so-called...
high-intelligence software agents. These artificial intelligence, a software that our preferences, our preferences and goals over time and then acts our name on a planning horizon or in real time, buying electricity or by
Our cars charge, electric cars charge or the washing machine runs, so the washing machine. And of course we have the so-called smart gateway, smart meters. But we in Germany, it's a very sad game really. We are still in the one-stop percentage range in the smart meter rollout, because many people, first of it is legal, I think first of all 300,000 kilowatt hours, which I don't is reasonable. And secondly, many people are
because they say, the data protection, the data protection. We're talking about an energy electricity profile. I don't makes sense, because if we to the energy shift, we to with dynamic prices. And honestly, five years we this house, and it's an energy efficiency house. And then we so-called many smart devices.
And they have for a long time, even if we not at home. So these fears that people have. someone can a break-in now. I'm home now and then I'm not home. us, there is always something going because then the agent decides, okay, I'll the washing machine. I'll the electric car now. The vacuum cleaner robot drives. It all autonomously. Nobody has at On contrary, when we at home, is at least very the devices.
Wolf Ketter (42:24.371)
We have small children, that's something else. These often the fears. There is indeed much more work to I hope I can contribute example, it's April 1st next year, I will also give one of three four keynotes on topic of artificial intelligence, chances and risks. And I see, of course,
Andreas Deptolla (42:27.502)
You
Wolf Ketter (42:54.035)
I have been working with technology for 89 years, that's a long time. And over time I a lot of that, but not that AI really dominates But will probably be that over time. Our research has that over the last seven years, especially in the field of machine interaction, that there are some tasks that AI can us. I would say that my everyday life,
Andreas Deptolla (43:21.912)
Mm-hmm.
Wolf Ketter (43:24.083)
Two to three hours a day, can probably to the AI. First, the emails through, some even answer pre-formulate, make Why should I deal And we in our research that for some tasks it in the future that, for example, the AI
Andreas Deptolla (43:34.168)
Mm-hmm.
Wolf Ketter (43:52.947)
the for some tasks. And then, when there is the people ask. But also vice versa. And I don't that as something I see that as something good. Today we have, even if you at simple examples, this right-writings culture or these text programs where you say, okay, maybe I write or like So that's nothing negative. You just to technology and then apply
Andreas Deptolla (44:10.722)
Mm-hmm.
Wolf Ketter (44:21.171)
And of course, have to, depending on the copyright, that's a completely different topic. And of course, a big issue with the KI is also the ethics. How to use In the USA there were many prominent cases. For example, Amazon was by several women ago because they said, okay, the settings, they were not considered. That was actually the case.
Andreas Deptolla (44:47.715)
Mm-hmm.
Wolf Ketter (44:50.579)
But the AI couldn't anything about because the AI is based on the so-called training data. And it was just that in the past, lot more men were white men of a certain age. And when the AI sees this data, it learns that too. And then, for example, recruiter suggests that. And then it's just like that.
Andreas Deptolla (44:58.52)
So, mm-hmm.
Andreas Deptolla (45:07.342)
Mm-hmm.
Wolf Ketter (45:19.699)
If you it's fulfilling profanity. That's the way What did they do? They women to court. Now they go there and positive way they edit the training data set. You of You have to that responsibly. Then women are also suggested
or in American courts, African Americans were condemned. They said they have dark skin color, they must be murderers or the criminals. Often it was white Americans who the crime. And that also cleaned There are also companies that have credit cards. As a person, credit card applied to the companies.
Andreas Deptolla (46:03.15)
Hmm.
Wolf Ketter (46:17.683)
have the neighborhood around asking how credible they are. But they the privacy of these neighbors without them wanting to and knowing And these things that you have to pay These ethical principles in AI are important. These are technologies that you have learn over time.
Andreas Deptolla (46:44.27)
I see that we could do 3 or 4 more podcasts. are so many topics. Maybe, to conclude, you work a with politics, Ursula von der Leiden, Olaf Scholz etc. In the area AI, you speak very plastic, these are...
three advices you would to the German or European politics. What would those be? Where okay, here we a lever, here we have to tackle
Wolf Ketter (47:19.539)
There are many things, but probably the massive expansion of the internet structure, broadband and mobile internet, definitely. Then the funding from universities, so much more money into education. Not only the government, but also the companies have to much more money into further education of their employees.
Andreas Deptolla (47:27.118)
Hmm.
Wolf Ketter (47:48.339)
Yes, just set so that companies do other things, so that yes, there is a lot that you just get more competitive If you see that the top 20 companies in world, 10 of the USA, 10 of China, then you almost in Europe. And that can't be our goal. That can't our claim.
We have get out of we have to do You can that the last few years many start-ups have as universities in Cologne. I think this is a step in the right direction, it long-term. Many companies have invest more in our countries, in the EU.
Andreas Deptolla (48:34.606)
Thank for your work. We just to talk together. I'm it worked out
Wolf Ketter (48:42.803)
Yes, thank much Andreas for the invitation. It fun and we can talk again in the future if you like. Yes. All the best. Bye.
Andreas Deptolla (48:52.632)
So let's it. Thanks, ciao!