
The Interview Mustafa Suleyman, Artificial Intelligence pioneer: People should be healthily afraid of AI
19 snips
Jan 9, 2026 Mustafa Suleyman, a British AI entrepreneur and co-founder of DeepMind, shares his insights on the dual nature of artificial intelligence. He emphasizes the need for a healthy skepticism towards AI's power, cautioning that it could overwhelm humanity without ethical safeguards. Suleyman discusses the rapid advancements in AI, the potential for creative breakthroughs, and the impact on jobs. He advocates for a humanist approach to AI and highlights the importance of preparing younger generations for a technology-driven future.
AI Snips
Chapters
Transcript
Episode notes
Healthy Fear Is A Useful Signal
- Mustafa Suleyman urges listeners to be "healthily afraid" of AI despite being techno-optimistic.
- He says that healthy fear should drive collective action and scepticism about AI's risks.
Exponentials Caused The Recent Leap
- Compute used to train top AI models has scaled exponentially, producing step changes in capability.
- By 2021–22 the same general methods began to master language, causing a rapid capability leap.
AI's Interpolation Feels Like Creativity
- AI now generates genuinely new interpolations across modalities, not just reproductions of past data.
- This generality enables creativity in images, text, and potentially molecules or scientific hypotheses.

