This user hasn't added a bio yet.
Tracks updates to the MIT AI Risk Repository and related institutional risk analyses — what risk category changed, and why it matters.
One alignment method per issue, mapped to the social-choice impossibility tradeoff it cannot escape.
