The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch 20VC: DeepMind's Demis Hassabis on Why AGI is Bigger than the Industrial Revolution | Why LLMs Will Not Commoditise & We Have Not Hit Scaling Laws | Bottlenecks in AI & The Energy Crisis Caused By AI | Whether AI Will Do More to Harm or Help Inequality
817 snips
Apr 7, 2026 
Demis Hassabis, co-founder and CEO of Google DeepMind and the mind behind AlphaGo and AlphaFold, explores how close AGI really is. He gets into scaling limits, why frontier labs may keep pulling ahead, and the missing pieces beyond today’s models. Also on the table: AI safety rules, drug discovery, energy strain, jobs, and inequality.
AI Snips
Chapters
Books
Transcript
Episode notes
Why Compute Still Gates AI Progress
- Compute remains the main AI bottleneck not just for training larger models, but for testing new algorithmic ideas at realistic scale.
- Demis Hassabis argues scaling gains have slowed from early dramatic jumps, yet returns are still substantial for frontier labs expanding compute.
Why Today's AI Still Feels Jagged
- AI is ahead in areas like video and interactive world models, but still lacks continual learning, robust memory, long-horizon planning, and consistency.
- Demis Hassabis calls current systems jagged intelligences because tiny prompt or environment changes can make them fail at elementary tasks.
Why Frontier Labs May Pull Further Ahead
- Demis Hassabis does not expect frontier models to commoditize evenly because labs that invent new algorithms will widen the gap as old ideas saturate.
- He expects open models to trail the frontier by about six months while staying valuable for startups, academics, and edge use through smaller systems like Gemma.
