Ideas: Mr. Westerheide, you are founders and CEO of AI for Humans GmbH. Can you briefly introduce yourself and your company?
Fabian Westerheide: I have been active in German and European digital ecosystem for over 15 years – as a founder, investor, networker, author and entrepreneur. I have set up start-ups, initiated venture capital funds and has been focusing on the topic of artificial intelligence (AI) since 2013. Since then I have accompanied the development of this technology as one of the central drivers in German -speaking countries – with the aim of strengthening Germany, but also conveying their social importance.
Together with my wife Veronika Westerheide, I founded AI for Humans GmbH a few years ago. It is the legal expression of a common mission: to help Europe to preserve digital sovereignty in the age of AI. The Rise of AI Conference has emerged from a small meeting-up in 2014, in which we dealt with the question of whether the technological singularity isorders, has emerged in the past ten years-today the leading conference for the European AI ecosystem. On May 14, 2025, the next conference will take place in Berlin – with over 350 thought leaders, entrepreneurs and decision -makers on site. At the same time, we offer a free livestream. If you want to understand how Europe can survive in a world driven by AI, you are welcome – personally or digital.
AI for Humans GmbH organizes this conference and strategically accompanies selected companies, institutions and managers on their way through the AI transformation. It is not about short -term tools, but about long -term orientation, orientation and effect.
Artificial intelligence is the most powerful tool of this century – in a world that changes faster and more radically. It is up to us how we deal with it. This requires technological understanding, but also the will to deal. Because AI is never neutral – it reflects the culture, the values and morality of its origin. That is why it is crucial for our future that Europe develops its own AI systems: so that we understand our education, our administration, our healthcare system, our way of life-and ensure our freedom. At the same time, together with partners with AI.Fund, I initiated a venture capital fund that supports European AI companies in the growth phase. Because the more powerful the technology becomes, the more important the people they shape – and the capital that enables them.
So you have long dealt with artificial intelligence before chatt. Were you surprised by the rapid spread of Chatgpt?
I have been dealing with AI for over a decade. That was long before the term “chatt” existed at all. The years 2014 to 2017 were crucial: At that time we experienced the basics on which today’s Large Language Models (LLMS) build up. So the trend was recognizable early – and who was attentive at the time knew: the (r) evolution was coming.
The quality of the jump from GPT-3 to 3.5 (at the end of 2022) actually surprised us. Suddenly there was a model that was not only theoretically good, but practically usable – at a high level. The result: 100 million users in two months. A global mindshift. AI had arrived in the mainstream overnight. Media, companies, politicians – all spoke about what we had been preparing for years.
So the real surprise was not that it happened, but when and how quickly it was. At the same time, this moment triggered massive investments again and again catapulted AI to the strategic agenda of states, governments and companies. That was already the case in 2016 to 2019 – but then Covid came. Now the momentum is back, stronger than ever.
At the end of last year, Deepseek, the Chinese counterpart came to Chatgpt. It should be much more efficient in terms of computing power. Is that really the case from your perspective?
Deepseek – or at the moment also manus AI – are a strong sign that China not only catches up, but is also increasingly able to adapt, optimize and bring them back into the market more cheaply. What is particularly striking: Deepseek needs significantly less computing power, some with cost differences in a factor of 30x to 500x – an enormous efficiency gain that is particularly important in economic application.
This shows: China has built up a powerful AI ecosystem. Investments in research, education, capital markets and industry over a decade now bear fruit. At the same time, you have to see: Deepseek is based in many ways on American preparatory work- many data and “mother models” come directly or indirectly from the USA. But China has learned to make a functioning product out of it – for less money.
Technically speaking, this is a positive development because it revives the competition. For us in Europe, this means that we are not allowed to make any illusions. The innovation speed increases – and it is no longer enough to just watch or regulate. We have to become active ourselves.
Do you see it critically from a user perspective that Deepseek is a Chinese company?
Absolutely. The view of functionality is no longer sufficient. The question is: what and who is behind it? China uses Ki very specifically to control and monitor the population – algorithms as instruments for social control. We see a system there that does not primarily use technologies to increase the efficiency, but for political power. And these technologies are already exported – in over 60 countries, especially to Africa and Southeast Asia.
This is a geopolitical reality check: While we are still discussing ethics and governance in Europe, China creates facts – with favorable hardware, efficient software and authoritarian efficiency. Precisely because American products are often more expensive and regulatory, many states rely on Chinese AI infrastructure.
Therefore: We Europeans have to build our own systems – now. We have to secure sovereignty, also technologically. And we have to export ourselves – into the world, not just within our borders. Otherwise we can find ourselves in a digital world order that no longer corresponds to our values.
Do you think that in the long run another country can make the USA first place on AI?
In the short term: no. Medium to long term: only China. The United States is currently unobtrusive-not only because of its tech companies, but because its entire system is designed for innovation, speed and scaling. There, capital meets risk and quick implementation.
China catches up, however – with a plan, with power and with a focus. But Europe? Here we still discuss whether we need our own AI at all. This would be exactly the decisive factor: a European answer that balances innovation and regulation, strengthens national champions, secures access to computing power and consistently thinks of education.
If we can’t do this, we continue to lose technological influence – and thus in political ability to act. We need a European AI strategy that is quick, brave and coordinated. Otherwise others determine our digital future.
How do you rate Germany’s position?
Germany has enormous potential – and uses it far too little. We are too slow, too bureaucratic, too fragmented. It is clear: Without a digital leadership role in Germany, Europe becomes technologically irrelevant. That is exactly why we have to act now.
I hope that the upcoming federal government will finally create a Digital Ministry – with real implementation competence. Germany needs a state reform that digitized the administration, simplifies processes and finally creates space for innovation. This includes targeted investments in AI, semiconductor, data centers – and the establishment of European clusters.
The SPD must learn not to see AI as a threat, but as an opportunity for more quality of life, more efficient work, for just participation. The CDU has to take state conversion seriously – with digital ministries, digital schools and digital industrial policy. And everyone has to ensure that the wealth gain by AI not only ends up with a few shareholders, but also arrives in the width of society.
One often hears that users do not use the complete potential of AI tools, since the quality of the prompts, i.e. commands, is insufficient. Do you have a tip for our readers how you can train here?
“Prompting” is the new digital key competence – comparable to the moment when people have learned to use Google efficiently. It is about giving AI clearly, structured and goal -oriented instructions – whoever can do this works more productively, more precisely and creative.
My tip: Take a look at the “Ainauten” newsletter https://www.ainauten.com/. There are practical tips, tools and prompt, explained there every day. Anyone who reads this regularly stays up to date, learns intuitively and saves a lot of time when familiarizing with new AI tools.
And another tip: come to the Rise of Ai conference on May 14th in Berlin-live or online. There are over 350 leading heads gathered that show how AI specifically works in practice-from medium-sized companies to research, from start-up to politics. If you cannot come personally, you can register for the free livestream and later view all content on Demand. You can get more information at https://riseof.ai/.
Thank you for the conversation.
Anja Schneider led the interview.