Can AI 'feel' emotions? Video interview with Prof. Dr Thorsten Zander and Robin Jost (with English subtitles)

26.11.2024

• Facebook
• Twitter
• Xing
• LinkedIn

Video interview Prof. Dr Thorsten Zander and Robin Jost on artificial intelligence and emotions

In November 2024, Prof. Dr Thorsten Zander and Robin Jost met for a video interview in the film studio at the Lusatian University. The Lüdenscheid-born professor and internationally recognised pioneer in the field of brain-computer interfaces shed light on the opportunities and challenges associated with the emotional dimension of AI. Together, they discussed the developments in artificial intelligence with regard to observing and analysing human emotions. Robin Jost, science communicator at BTU Cottbus-Senftenberg, moderated the interview, which can be viewed on the official YouTube channel of BTU Cottbus-Senftenberg. You can find the complete video interview with English subtitles here.

Robin Jost:
Dear viewers, dear Professor Zander, I am very pleased to welcome you here today at the campus of the Brandenburg University of Technology Cottbus-Senftenberg. You have been the Volkswagen Foundation's Lichtenberg Professor for Neuroadaptive Human-Technology Interaction at BTU since 2020. And it is safe to say that you are a pioneer in the field of passive brain-computer interfaces. You are also co-founder and director of the Society for Neuroadaptive Technology, on whose behalf you organise conferences and other events to promote research and outreach activities on all aspects of neuroadaptive technology.

Your key research areas are the implicit interaction with technology, cognitive probing for automated guidance of AI, but also the ethics of neuroadaptive technologies. Also: Your start-up Zander Labs was awarded over 30 million euros in funding from the cyber agency last year. Just to provide some context: This is the highest single grant that has been awarded in the EU. Our viewers can see that: Groundbreaking research is being conducted here at BTU Cottbus-Senftenberg - and Professor Zander is at the forefront of this.

Therefore: Welcome to you once again! Let's get right into it. What are neuroadaptive technologies?

Prof. Zander:
Thank you very much for the invitation, Mr Jost. I am very pleased to be here today. Regarding your question: neuroadaptive technologies are technologies that adapt to the neuronal state of humans, i.e. to our brain activity. To put it abstractly, in a certain way they also adapt to what we are thinking. We cannot read thoughts directly, that is very important. But we can hear the "operating noise" of the brain, so to speak. The machines can recognise whether we are surprised, stressed or happy and can incorporate this information into the way they work and adapt to us. So the machine could understand that if I have just read something and interpret it as incorrect, it could then give me corrections or provide further information.

Robin Jost:
And your research also deals a lot with the observation and evaluation of human emotions and you had already mentioned thoughts. What does emotion mean in relation to artificial intelligence? Can machines really feel emotion or is that an illusion, so to speak?

Prof. Zander:
Whether machines can really feel anything is a question that we as humans cannot answer. It's also a question about being human. Is there something in us that is not purely physical? Or are we based on physical processes and our emotions too? If that is the case, then I would say that a machine can also learn to feel emotions. So I don't see any reason why this can't work. Of course, it could take a very long time before it really works in such a way that a machine feels emotions in the same way as we humans do. In philosophy, this is in the area of qualia, where you ask: "Are my own sensations really exactly the same as your sensations? Can sensations really be transferred?"

And I can't answer these questions, and I don't think any human being on this planet can answer them at the moment. However, I think that a machine (if we're talking about artificial intelligence) can understand what we mean by emotions and can at least simulate them. If you have a robot that can be shown certain facial expressions, for example, and it understands that we are friendly or happy, then it could also smile. And whether that is the same as feeling it - I wouldn't want to explain that. But I think that we can enrich machines so that they can understand and respond to emotions.

Robin Jost:
At a certain point, that's a philosophical question, right?

Prof. Zander:
Absolutely. It's purely philosophical.

Robin Jost:
That's actually a nice segue into the current state of research, when it comes to equipping artificial intelligence or AI systems with emotional intelligence. Can you give us an overview of the current state of research?

Prof. Zander:
Yes, I think that's rather difficult at the moment. There are certain approaches in which robots or machines react to certain impulses, but there is no real understanding. It's more of a direct reaction that is already pre-programmed, so to speak. Behind this is what is known as "alignment" or "human compatibility". This is the English term and covers the extent to which a machine can adapt to us, to our values and understand what is important to us. And of course this also includes our emotions. So what are we happy about? What are we sad about?

And solving this is a huge topic in AI research, because it is of course also assumed that artificial intelligence will become more powerful, i.e. have a greater influence on the world. And if they don't understand us, there's a big risk that this could be a problem.

Perhaps for some background: companies like OpenAI, which produced ChatGPT, had a so-called "Super Alignment Team", in other words a team that really put a lot of effort into bringing emotions and an understanding of our emotions and values into the machine. This was then disbanded after four years because they simply hadn't made any progress.So to summarise, it's a hugely important issue that we should all be concerned with. I think we should all be looking at AI more anyway, because it will become very important for us in the near future. But there are no clear concepts being implemented at the moment that can solve the problem of the machine being able to understand, comprehend or perhaps even emit emotions. I have a concept, but it hasn't been realised yet. We are currently in the process of realising it.

Robin Jost:
And what role does science have here at the Brandenburg University of Technology Cottbus-Senftenberg and also the region of Lusatia? What role does this centre of innovation play in this topic, so to speak?

Prof. Zander:
It's a two-pronged approach. On the one hand, it is, in my opinion, the only place in the world where the interaction with AIs is really focussed on neuroadaptive systems, i.e. where there really is a "track", i.e. a series of lectures that really deal with the topic. We have an "Artificial Intelligence" study programme, which is in English.

Robin Jost:
And this study programme is also unique, or at least there are very few, right?

Prof. Zander:
Exactly. There are many programmes that call themselves AI, but this focus on neuroadaptive Systems is unique. As far as I know, there is no other like it in the world and of course you could say, "Well, then it's not an important topic if it only exists once" - but that's the wrong way to look at it. It is something that is completely new and where we really are at the forefront of this development.

On the other hand, it's not being recognised as much. So we see - and this is also a very interesting phenomenon - in every country except Germany, this research and my research is clearly being recognised and is also being strongly discussed and there is a great deal of interest. And here in Germany there is an enormous lack of interest. I thought about it for a long time because I personally find it surprising, but interest is just beginning to emerge here. The 30 million euros in funding from the cyber agency has naturally generated a lot of interest. But as far as the students are concerned, there are around 300 students in total in my lectures each semester. Two or three of them are from Germany.

Robin Jost:
Only two or three?

Prof. Zander:
Yes, the rest are from other countries. They are all coming to Cottbus. That's quite interesting. But it clearly indicates that the perception in other countries is much stronger than in Germany.

Robin Jost:
Madness. But I think that the funding will also draw attention to the region and to research here and will certainly also help to ensure that research here in Germany continues to be recognised and respected.

Prof. Zander:
Absolutely, I think so too and the hope is really there. But of course it will take a while. But when it comes to the moment, it's more the foreign countries that are interested in what we're doing.

Robin Jost:
That's exciting. And if you can perhaps give us an outlook for the future: How much will artificial intelligence influence our lives in the next 5, 10 or perhaps 20 years? To what extent will it shape us? How far will emotional artificial intelligence develop - do you perhaps have a feeling as to how things might progress?

Prof. Zander:
Yes, I have a clear feeling. Talking about times is always difficult. In my career as a scientist, I've seen how people have made incredible mistakes. It's just so complex, you can't predict it. But I think in our lifetime we will see massive changes that will take place in almost all areas of our lives. So the question is: what will change and why will it change? And these AIs will bring a new perspective on the world and they will come into our lives like we might imagine aliens coming to this planet. They will be able to act and understand the world, but they will probably see the world differently than we do. And they will also be able to act on a similar level to us.

In some things they will be better, in some things they will be worse. So you have to imagine it: All of a sudden, a new species arrives on this planet. And it has a right to be here because it is intelligent and perceives things, and it will simply be here and interact with us. There is a positive perspective that we can see where we as humans can contribute to the well-being of the planet and where the machines can also contribute to the well-being of the planet. If we can manage this communication and develop an understanding for each other, we can work together positively on what makes our lives beautiful.

But of course that's too positive. It will of course be abused - it is already being abused. The political situation in Germany is so tense because there is so much fake news out there. And of course this is already being dealt with via AI, but this is just the first, shallow beginning. We have seen from Cambridge Analytics how elections can be influenced without any major AI influence. That means we also have to be careful. We will see in all areas - both negative and positive - that AIs will have consequences. I think that's what we need to understand. Our actions have consequences, at the moment the actions of machines, which are autonomous, don't have big consequences. So it's happening and it's coming more and more, but there's still a human behind it. We always say: "Humans have to decide" - that will go away completely.

We will have machines that make completely autonomous decisions. And then they will also have to take responsibility for the consequences completely autonomously. And that is of course extremely difficult when you imagine a society that suddenly has a new entity that can also do things, and we have to learn that. We have to deal with the fact that we are facing these challenges and instead of always slapping each other in the face, we have to work together to make it work. And if we can do that, then we will certainly have a share in this positive situation, that we can solve problems together with machines that we couldn't solve before, so that all of our lives will be better.

So I think, in the long term, that's exactly what it will come down to. There will be a few back and forths where, as with all new technologies, we will have problems and then get annoyed about them. But ultimately, in the long term, we will be able to solve problems that we can't solve at the moment. I believe that together with AIs and human intelligence, we can make a huge contribution to solving problems such as climate change or world hunger or equal distribution in society or other challenges in a way that we can't even imagine yet. I don't believe that AIs will be able to do this on their own. I think they simply lack the understanding to do so. But if we learn to work together with the AIs and bring our two forms of intelligence together, I have a lot of confidence. And that's also a key motivator for my own research.

Robin Jost:
Yes, that's nice that there's also a certain hope involved. One thing is for sure, artificial intelligence will have a huge impact on our lives in the coming years and decades. And I think we could talk about this topic for a very long time, because it's really exciting.

Prof. Zander:
I'd love to.

Robin Jost:
But perhaps that's a good final word, so I'd like to thank you once again for your time and for the totally exciting conversation, and also to our viewers, and I wish you all the best with your research. See you next time!

Prof. Zander:
With pleasure. Thank you very much for the invitation and all the best to you too. .

World Compassion Day

World Compassion Day is a global day of action to fund compassion and empathy. It is celebrated annually on 28 November and serves to inspire people to create a more compassionate world - be it through animal welfare, sustainable action or social justice. The day invites people to not only think about compassion, but to actively live it, be it through small gestures, charity work or more conscious decisions in everyday life. Events, lectures and campaigns around the world aim to raise awareness of the importance of compassion in modern society. World Compassion Day underlines the relevance of Prof. Zander's AI research, as his approaches show how compassion and ethical behaviour can be supported and funded by AI. In areas such as healthcare, education and social services, it is becoming increasingly clear how important it is to develop technologies that are not only functional, but also empathetic to people's needs and feelings. The day of action illustrates why such perspectives in AI development are central to a more humane future.

Contact us

Press contact

Further information