The Diary Of A CEO CEO Of Microsoft AI AI Is Becoming More Dangerous And Threatening! - Mustafa Suleyman
Jan 7, 2026
This discussion features Mustafa Suleyman, co-founder of DeepMind and advocate for AI safety. He shares his early fears and empowerment in guiding AI responsibly. Mustafa explores containment challenges and the risks of AI-enabled bioweapons. He envisions a future with humanoid robots and emphasizes the need for global regulation. The conversation dives into designing cooperative AIs and balancing short-term incentives with long-term safety. Mustafa urges individuals to engage in AI regulation and highlights the potential for both radical abundance and catastrophic failure.
AI Snips
Chapters
Books
Transcript
Episode notes
AI's Future Is A Human Choice
- AI's form and governance are not inevitable and will be decided by collective human choices.
- Mustafa Suleyman argues we must actively shape who owns, trains, and governs AI over the next decade.
The Digit That Opened A Decade
- Mustafa recalls watching a model generate a handwritten digit seven and feeling stunned by its emergent capability.
- He traces that simple moment to today's photorealistic image and video generation progress.
Scale Fueled Language Surprises
- Scaling compute and data produced unexpectedly strong language models that feel human-like.
- Suleyman highlights yearly 10x scaling as the driver behind today's conversational AIs like ChatGPT and Pi.
