Ex-DeepMind Researcher Seeks Funding for Agents That Continuously Generate Their Own Training Worlds
jparkerholder · x · 2026-08-14
Former Google DeepMind researcher Jack Parker-Holder has launched an open call for a new experiment. He proposes that instead of training on a fixed set of environments, AI agents should continuously create new ones for each other to drive capability improvements.
He aims to run this experiment publicly at a much larger scale and is currently looking for compute funding or contributors. This idea builds upon his paper Open-Endedness is Essential for Artificial Superhuman Intelligence, which argues that open-ended, ever-self-improving systems are a crucial ingredient for achieving ASI.
Related event: Ex-DeepMind Researcher Calls for Funding Open Agent Experiments(2 posts)→
More from AGI Musings
- True Superintelligence Hinges on Working Memory Over Reasoning — josh_wills · 2026-08-14
- Exploring: When Will Multi-Agent Systems Self-Organize Around High-Level Goals? — matt_slotnick · 2026-08-14
- AI Search Grows 197% YoY: Why the AI Market is Not Zero-Sum — annbordetsky · 2026-08-14
- AGI as a Rorschach Test: Predictions Reflect Personal Desires, Not Objective Future — danfaggella · 2026-08-14
- Beyond Goodhart's Law: Why Evals Are the North Star for AI Capabilities — typewriters · 2026-08-14
- Robotics Will Expose the Real Cost of Human Labor and Disrupt Demand — LexSokolin · 2026-08-14