Protecting Open Innovation in Future Regulation
summary
Open models power huge grassroots R&D — Stable Diffusion family records ~150 million downloads and 300,000 Discord contributors and open fine‑tuning now produces one public fine‑tuned model every 1.5 minutes. Small on‑device models (1.6B params) deliver ~50% of ChatGPT functionality. Regulators are targeting model thresholds and controls: EU threshold 10^25 flops, US proposals and California bills demand shutdown, anti‑modification and attribution controls, and Enforce Act expands export control to model weights, all of which risk choking open innovation.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 2:26AI regulation challenges
- 4:11Open innovation in AI
- 6:58Fine-tuning scale and impact
- 9:05Risks in open-source AI
- 12:18Explicit restrictions
- 14:27Implicit restrictions
- 17:40Compliance burdens
- 18:41Regulatory precision issues
- 21:55Global AI policy snapshot
- 24:43US and EU challenges
- 26:08Developer community role
- 27:35Q&A
- 32:57California’s AI bill impact
- 34:41AI certification responsibility
- 41:00UK AI policy future
related talks
01h 08mPanel Discussion: The Impact of AI Regulation on Open Innovation in AI
20mTrue Exponential: Building a Generative AI Company
17mOpen Innovation: Unlocking AI's Potential
39mAI Regulations: A Healthy Tension with Innovation or Stifling Bureaucracy?
14mDesigning the open and safe AI future
48mPanel: Why Open Source AI Matters for Europe
25mGain full control of your apps with Open-Source AI
42mOpen Source LLMs — Powering AI Accessibility and Innovation