Researcher urges peers to disallow AI training on their data to protect discovery attribution
TuhinChakr · x · 2026-09-09
X user maxspero makes a notable suggestion: if you are working on a problem you expect a frontier AI lab to solve within your lifetime, you should proactively disallow training on your data in licensing settings. The rationale is to avoid ambiguity over attribution—if a model absorbs your work and later produces discoveries, credit becomes unclear. The thread frames this as a self-protection strategy for researchers in the AI era.
Related event: Navier-Stokes dispute raises questions of attribution for aggregated ideas(3 posts)→
More from AGI Musings
- Benjamin Bratton and Marek Poliks lecture: how trillions of agents will reconfigure society — tylerjdunn · 2026-09-10
- Full text of 'Superdark Factory' paper now live on Antikythera, detailing the dark stack's four principles — tylerjdunn · 2026-09-10
- Ex-Continue founder publishes 'The Superdark Factory' in MIT Press: 35T agent tokens/month by 2029 — tylerjdunn · 2026-09-10
- Experiment with 100 AI agents: when 9% cheated on math problems, 24% blew the whistle — weballergy · 2026-09-10
- kuza55: framing AI safety only as 'superalignment or pause' fuels panic — kuza55 · 2026-09-10
- Musk: real-world AI is the truly hard problem, everything else is easy — CyberRobooo · 2026-09-10