Turning hard science problems into open human-vs-AI games: a research model proposal
FrankFelixAI · x · 2026-09-02
Sharing Julian Duru's proposal: what if hard scientific and engineering problems across digital and physical worlds became open games where humans and AI agents compete, collaborate and build on each other's work?
The mechanism mirrors The Innovation Game (@tigfoundation), which drives verifiable algorithmic progress via competitive challenges and rewards: publish a challenge with clear acceptance criteria, let anyone or AI agents submit, test every submission transparently, reward what works, and make each verified result the starting point for the next challenge—turning research into a cumulative solve-verify-reward-build loop. FlowResearch is exploring extending this to a much broader class of problems.
More from AGI Musings
- Tester Claims Fable 5.1 Excels at Long-Horizon Tasks, Finance and Consulting 'Wiped Out' — felpix_ · 2026-09-02
- Raphael Millière argues intentional stance usefully describes AI agents without anthropomorphism — sebkrier · 2026-09-02
- One LLM already runs at 14,000 tokens/sec—frontier intelligence at 5,000 tok/s within 5 years? — dolo937 · 2026-09-02
- What If Frontier Labs Stop Releasing Models and Keep 'Oracle' AI In-House? Reddit Debates — IDefendWaffles · 2026-09-02
- The Cognitive Revolution: we externalized our muscles, now we're externalizing our minds — Konstantine · 2026-09-02
- AI hallucinations flood Australian parliament inquiries with fake research — nordicinst · 2026-09-02