tech
The Man Who Thought He Could Keep AI Safe
Demis Hassabis has devoted his life to advancing a technology he thinks could destroy the world.
TL;DR
- Demis Hassabis, co-founder of DeepMind, is developing superintelligence while fearing its destructive potential.
- His initial vision for a controlled, safe AI development via a 'singleton' scenario failed.
- Key events like Elon Musk founding OpenAI and Google's commercial interests complicated his safety agenda.
- Hassabis's attempts at establishing AI governance safeguards with Google and for public benefit (NHS) ultimately faltered.
- Despite setbacks, Hassabis remains driven by a competitive spirit and now focuses on gaining personal influence to ensure AI safety.
- He has shifted from an idealist to a realist, believing adaptability and a "seat at the table" are crucial for navigating AI's unknown consequences.