https://www.lesswrong.com/posts/Zr37dY5YPRT6s56jY/thomas-kwa-s-shortform?commentId=Pwc2SmyPi9u3maSsd
Resolution criteria
This market will resolve to the date on which Thomas Kwa officially ceases to be an employee of OpenAI (assuming that he has already signed the contract).
Definition of "leaving": "Leaving" is defined as the termination of his employment contract with OpenAI, whether due to resignation, quitting, or termination. An internal transfer to a different team within OpenAI does not count as leaving, but transferring to an external organization or spin-off does. (Example: going to the OpenAI foundation would be considered leaving).
Sources for verification: The resolution date will be verified using the first available public source of truth, such as:
Thomas Kwa's LessWrong profile or updates on his LessWrong Shortform.
Professional directory updates (e.g., LinkedIn, the METR team page if he returns, etc.).
Official announcements from OpenAI or reliable reports in tech/AI journalism.
Date calculation: The market will resolve to the actual final date of his employment at OpenAI, if known. If only his departure announcement date is public and the exact final day of employment is unclear, the market will resolve to the date the departure announcement was posted.
Background
In late August 2026, AI safety researcher Thomas Kwa announced on his LessWrong Shortform that he was leaving the non-profit evaluation organization METR (where he spent ~1.7 years) to join OpenAI to work on measuring and modeling recursive self-improvement (RSI).
In his announcement post, Kwa explicitly outlined a short timeline for evaluating his impact:
"My work could easily be net negative if the team can't publish something meaningful within 4-6 months, so if this happens I will probably quit or internally transfer to something more robustly good."
This market tracks when he will depart from OpenAI.
1,000