Neuroscience, DeepSeek & the Future of Europe

On 21 March, neuroscientist and AI visionary Benjamin Bargetzi will be speaking atthe Unternehmertag about Artificial Intelligence. Here he shares his vision forthe future of AI for society, the economy and for each and every one of us –and what Europe and companies need to do to keep up with AI.‍

‍

U-Tag:What do you expect from the German and European AI start-up scene, and whatfuture does AI have in Europe, despite all the cutting-edge research?

Benjamin Bargetzi: Europe lacks its own Silicon Valley. While there are innovation hotspots in Bangalorein India, Shenzhen in China and, of course, in California, there is nothing comparable in Europe. Berlin and London are making a start, but they don't offer the same kind of vibrant environment where creative people looking for this kind of thing can go.
Risk aversion in Europe is a major problem. While investment in the US, China or even India flows more easily with government support, founders here have to fight tooth and nail to get investment and support. Europe undoubtedly has excellent research, but it lacks a culture that really promotes entrepreneurship and innovation.
If Europe would finally agree to take innovation seriously, I see a bright future for us, but it would just have to happen slowly. And the chances are good, as shown by the Chinese Deep Seek-R1 model, which works just as well as ChatGPT but cost only six million dollars to develop – and doesn't need fancy Nvidia chips either. This means that Silicon Valley's lead is large, but not insurmountable, and that should be a motivation for Europe.

‍

U-Tag: Can AI develop visions for business strategies?

BB: Absolutely. If I am starting a new business, I can use AI as a brainstorming partner to identify potential weaknesses or generate new ideas. In this way, AI can act as a kind of mentor, available 24/7.
However, it is important to remember that once uploaded, all sensitive information is stored on Open AI and Microsoft servers in the United States. You should therefore consider using a personalised custom GPT where the data is stored locally and securely.

‍

U-Tag: Will AI disrupt the management level?Which managers and directors will stay, which ones will have to go?

BB: The real upheaval in management didn't start with AI, but already in 2020: digitalisation due to the coronavirus, global conflicts, supply chain problemsand energy crises have forced companies to radically rethink their strategies.AI will only accelerate this change. But I don't think it will be AI that replaces management, but rather the zeitgeist that has prevailed since 2020, which, for example, calls for new working models such as remote work and is intended to make companies more efficient and leaner. Tech companies such as Google or Meta are already increasingly laying off middle management as AI takes over many administrative processes.
Future leaders will need a spirit of innovation, trust in the future and the ability to embrace change. Inspiring confidence in the future will be incredibly important for leaders. Those who just manage will be replaced.

‍

U-Tag:Is AI also conceivable as a valid investment manager? Will AI be more successful than human investors in the financial markets?

BB: I see no reason why machines shouldn't make better investment decisions than humans. AI can spot patterns and analyse markets faster and more accurately than any human.
However, acceptance will take time. People prefer to trust other people, especially when it comes to money or vital decisions. But as soon as AI is proven to deliver better returns, it will take off.
The long-term perspective is exciting: initially, AI investors will have an advantage. There will be a transition period in which AI systems perform much better than humans, but when everyone is using AI, it will stabilise, the human factor will come back into focus, and we will return to a new equilibrium.

‍

U-Tag: Many people fear an omnipotent, generative intelligence. Is this concern justified?

BB: There are two extreme positions: on the one hand, AI is seen as a harmless tool that merely increases productivity. On the other, there are fears of a Terminator-style threat that will destroy jobs and destabilise society. As is often the case, the truth lies somewhere in the middle.
I'm less interested in consumer applications like ChatGPT or Midjourney than in bringing machine learning algorithms into research. There are incredible opportunities here, for example in cancer research, where we still don't fully understand what's going on. An AI that can analyse huge amounts of data could uncover new connections and enable breakthroughs. One example is Google DeepMind's AlphaFold 2, which can predict the folding of proteins with a high degree of accuracy – a milestone for medicine. The big opportunity for AI is to develop a machine that can one day produce drugs for Alzheimer's or cancer.
Of course, AI also poses dangers. For example, I am strictly against the use of AI in autonomous weapons systems – I see no added value here, only risks. But basically, AI is like money or weapons – whether it is used for good or evil depends on the people who use it.

‍‍

U-Tag: That leads to the question: Can AI solve the fundamental mystery of free will?

BB: I worked for a long time in neuroscience and psychology before turning to the world of machines and robots. So, I am very ambivalent about this question. We are now trying to find answers to the big philosophical questions through all the AI topics – but the problem is that neuroscience has not yet been able to give a clear answer to whether humans have free will or how exactly consciousness arises. There are many theories, but our understanding of what the brain actually does and why it can think consciously remains incomplete.
What we do know from neuroscience is that much of our thinking is automated. Our brains work with patterns, creating models of the world and adapting them to new insights. Whether this can be understood as free will, or whether we just want to believe that we have it, is another question.
One exciting theory from neuroscience sees consciousness as the result of extremeinformation compression. The brain works mechanically, almost like a robot, butat a certain density of information, something like consciousness could emerge.If you apply this theory to AI, a machine that processes a similar amount ofinformation in a limited space could also develop consciousness.
At the same time, this raises the question: if the brain works purely mechanistically and you don't believe in a soul or something supernatural, does that mean that free will in absolute form does not exist? Maybe there is a limited form – the ability to stop and decide whether you want to do something or not – but on the whole you are remote-controlled and act according to predetermined patterns.
That's why I don't think AI will give us the final answer to the question of free will. But what it will do is take the debate about consciousness to a new level. If, for example, robots with consciousness work for us in the future, ethical questions will suddenly arise that are reminiscent of the problem of slavery.

‍‍

U-Tag: What role will AI assistants play in 2025?

BB: I predict that the most commonly used buzzword in 2025 will be 'AI agents'. You will have just one AI system that you can operate with one element, such asChatGPT, and it will connect multiple systems and perform actions and complete tasks autonomously. For example, I could say to my mobile phone, 'I need to get a haircut' and the system would take care of all the organisation. Or, 'I want to travel around Vietnam for three weeks' and the flights, hotels and itinerary are already taken care of. Manual booking will hardly be necessary.
These developments will drastically reduce bureaucracy and administration and make companies more productive. Those who close their minds to this development will not remain competitive in the long run.Thesedevelopments will drastically reduce bureaucracy and administration and makecompanies more productive. Those who close their minds to this development willnot remain competitive in the long run.  

‍

U-Tag: Could you please develop a real-life perspective for us of what the Unternehmertag will look like in five years?

BB: From a technological point of view, it would be possible to hold the Unternehmertag entirely in the metaverse. Participants could use VR glasses to attend from home, move interactively through the setting, follow the panels and not have to travel.
But I don't think that's going to happen. I think we'll still be meeting in person in five years' time because people like people. People are social creatures, and nothing can replace face-to-face contact. This is where my background as a neuroscientist meets my current career in AI: things are becoming possible with technology, but that doesn't mean that humans will do all of them. Perhaps there will be hybrid solutions – with a real conference and a parallel virtual version for those who cannot attend in person.
AI could simplify the booking process and organise the whole stay on its own. You could have AI-driven robots performing as a show act or serving drinks. But at the end of the day, the personal, direct interaction between people remains at the heart of such events – regardless of the technological possibilities.

‍