Artificial intelligence is increasingly becoming part of the scientific process itself. Beyond analyzing data or predicting molecular structures, emerging AI co-scientists can generate hypotheses, critique ideas, search the literature, propose experiments, and interpret results. In lab-in-the-loop approaches, these computational systems are connected with wet-lab experimentation and automation, creating iterative cycles in which models learn from experimental data and suggest what to test next. Recent systems from Google DeepMind, FutureHouse/Edison Scientific, and Stanford s illustrate how quickly this field is developing. Yet questions remain around scientific novelty, data quality, hallucinations, bias, and whether faster science necessarily means better science. With Dr. Thibault Geoui, life-science and AI expert at Zühlke and host of the Tech and Drugs podcast, we discuss the rise of AI co-scientists and self-driving labs, where the real bottlenecks in AI-driven drug discovery lie, and how the role of human scientists may change as AI moves from software tool to scientific collaborator. Find Thibault here: https://www.linkedin.com/in/thibaultgeoui/ Listen to the Tech and Drugs podcast here: https://youtube.com/playlist?list=OLALYPui2n_zY5D4atXf3KbIcPUSQcWN7-A&si=V5-C6KIIOMpGonI7 Read Thibault’s Newsletter here: https://www.linkedin.com/newsletters/7036361021543866369/?displayConfirmation=true

Disclaimer: Louise von Stechow & Andreas Horchler and their guests express their personal opinions, which are founded on research on the respective topics, but do not claim to give medical, investment or even life advice in the podcast.

Learn more about the future of biotech in our podcasts and keynotes. Contact us here: scientific communication: https://science-tales.com/ Podcasts: https://www.podcon.de/ Keynotes: https://www.zukunftsinstitut.de/louise-von-stechow

1. https://www.nature.com/articles/s41586-026-10644-y
2. https://www.nature.com/articles/s41586-026-10658-6
3. https://www.nature.com/articles/s41586-026-10652-y
4. https://www.fiercebiotech.com/biotech/openai-launches-biotech-specific-ai-model-gpt-rosalind
5. https://www.medrxiv.org/content/10.1101/2025.07.07.25331008v
6. https://www.nature.com/articles/d41586-025-02028-5
7. https://arxiv.org/abs/2410.09403
8. https://www.nature.com/articles/d41586-023-02218-z
9. https://valutric.com/googles-ai-co-scientist-vs-openais-deep-research-vs-perplexitys-deep-research/
10. https://www.nature.com/articles/d41586-025-00343-5
11. https://www.nature.com/articles/d41586-024-02842-3
12. https://sakana.ai/ai-scientist-first-publication/
13. https://www.wired.com/story/ai-scientist-ubc-lab/
14. https://research.google/blog/accelerating-scientific-breakthroughs-with-an-ai-co-scientist/
15. https://arxiv.org/abs/2502.18864
16. https://www.nature.com/articles/d41586-025-01485-2
17. https://www.science.org/content/blog-post/evaluation-deep-research-performance
18. https://modernsciences.org/openai-deep-research-agent-ai-tool-limitations-february-2025/
19. https://www.nature.com/articles/d41586-025-00377-9
20. https://arxiv.org/abs/2409.13740
21. https://www.nature.com/articles/d41586-024-01684-3
22. https://www.nature.com/articles/s44286-023-00002-4
23. https://pmc.ncbi.nlm.nih.gov/articles/PMC10700133/
24. https://www.nature.com/articles/s41586-025-09922-y
25. https://www.nature.com/articles/d41586-026-01551-3