Science Utopia: LLM agents simulate an entire academic research ecosystem
jindong_wang92 · x · 2026-10-20
Jindong Wang (William & Mary), James Evans (UChicago) and collaborators present Science Utopia (SUTO), a closed-loop LLM-agent framework simulating an academic ecosystem where AI massively enables paper writing, peer review and grant writing.
Highlights:
- Persistent researcher state across years: expertise, memory, reputation, budget, collaborations and COI.
- Full loop: direction/team formation → submission → double-blind review → decisions → funding allocation and attrition → revision/resubmission, iterated yearly.
- Controlled counterfactuals on the same initial world to study output, diversity, resource inequality and career sustainability.
The motivating question: when everyone can access top models for research, does a Matthew effect emerge where those with more API credits produce more? Paper, website and code are public.
More from AGI Musings
- ChatGPT flubs a simple crossword, shaping outsiders' skepticism about AI — iruletheworldmo · 2026-10-20
- The AI safety inversion: why alignment efforts keep increasing risk — mayfer · 2026-10-20
- AI data brokering splits: pre-training data commoditizes, RL post-training data is the real business — agihouse_org · 2026-10-20
- Google researcher: 'tradition' is the social mechanism absorbing displaced work — moultano · 2026-10-20
- Frontier labs don't want your chats — they want the world around you when you typed the prompt — marktenenholtz · 2026-10-20
- Ethan Mollick: multiplayer AI must move past group chat — its promise is working across teams — emollick · 2026-10-20