Invited Talk by David Bessis

“If AI does the Science, what is left for us?“

“AI cracks mathematical conjecture“ / “LLMs produce PhD-level research“ / “The top physicists have already surrendered to AI superiority.“

Headlines like these are currently shaping the public debate. The capabilities of AI systems are advancing rapidly – and there is currently no indication that this development will come to an end. Does this mark the end of human science? What role does a PhD still play if research could become largely automated in the future? And why should anyone still make the effort to learn, or even write a PhD thesis?

These questions were at the center of the guest lecture “If AI does the Science, what is left for us?”, which was organized by the Center for Quantum Technologies (QuTe) in early July 2026. The guest speaker was mathematician, tech entrepreneur, and author David Bessis, who published the book Mathematica: A Secret World of Intuition and Curiosity in 2024. In his work, he explores the role of intuition, creativity, and human knowledge creation in mathematics – and thus addresses precisely the questions that are being raised anew by the rapid progress of artificial intelligence (AI).

More than 160 participants gathered in the lecture hall of the Max Planck Institute for Software Systems (MPI-SWS) to seek answers to these questions and discuss the role of AI in science and research. Following a welcome address and a brief introduction to the Center for Quantum Technologies (QuTe) by Managing Director Christian Hochbruck, the first session of the evening began: a moderated discussion between David Bessis and QuTe’s Scientific Director, Prof. Dr. Moritz Weber. The discussion focused on the impact of modern AI systems on scientific work. Topics included which achievements of current AI models are truly impressive, where their limitations lie, and why scientific understanding remains indispensable even as machines become increasingly capable of solving problems or suggesting new insights. The discussion also addressed which skills researchers will need in the future, how the role of doctoral training and scientific education may change, and why human creativity, intuition, and judgment will remain essential even in an era of advanced AI systems. Rather than portraying a scenario in which AI replaces scientists, the discussion instead highlighted a future research landscape in which AI opens up new possibilities and helps researchers address complex questions more quickly and comprehensively.

The discussion between Bessis and Weber was followed by a panel discussion entitled “How does AI change Research Practice?”, featuring researchers with different academic backgrounds and scientific perspectives. The aim was to move beyond general headlines and hype-driven debates about AI and to discuss concrete examples of how AI is already changing everyday scientific work – from structuring initial ideas and conducting literature reviews to planning experiments, analyzing data, and writing scientific articles and reviews. The discussion made clear that the panelists primarily view AI as a supportive tool that opens up new opportunities and can facilitate scientific processes in many areas. At the same time, the participants emphasized the need for transparent guidelines for the responsible use of AI, particularly regarding traceability, scientific integrity, and the disclosure of AI-assisted content.

The central message of the evening was clear: Artificial intelligence will fundamentally transform research in the coming years and simplify numerous workflows. However, scientific progress will continue to rely on human creativity, critical thinking, intuition, and the ability to ask new questions and critically evaluate insights.

The full recording of David Bessis’ talk is available here.