Towards a Better Environment for Frontier AI in Europe
summary
Europe currently holds 4.8% global GPU share and planned gigafactories of ~100,000 H100 equivalents rank around 30–35, far smaller than projects hundreds of times larger; LLM capability for software tasks doubles roughly every seven months and GPT‑5 sits above that trend. Urgent priorities are rapid national leadership on compute, staffing technical policymakers (the UK hires ~70 technical AI researchers), a government technical foresight unit, and enforced frontier rules — the EU’s code of practice signed by Microsoft OpenAI Mistral XAI models workable regulation.
AI-written summary · about
chapters (8)
- 0:00Europe's AI frontier challenge
- 1:26Evidence against AI slowdown
- 6:13Rapid AI advancement consequences
- 8:22Europe's shrinking AI compute share
- 13:22Need for more AI infrastructure
- 15:52Government technical expertise
- 16:36EU frontier AI regulation success
- 19:51AI regulation requirements
related talks
27mMaking Europe Exponential
51mUnlocking Europe's AI Potential
17mState of AI 2025 – Opening Keynote
46mTowards Trustworthy AI: The European Approach
46mThe tech giants in AI - who is going to win the battle?
35mSovereign AI in Action: Advancing India’s Digital Future With Ope...
24mThe EU US Tech Gap
38mAI’s Evolving Landscape: Opportunities Strategy Economics and Sustainable Advantage