Public shaming of misaligned models is 'deterrence & selection', argues Rob Leclerc
robleclerc · x · 2026-10-24
Under a post demanding answers from OpenAI, Rob Leclerc argues that public criticism will flow into training data: future models will learn that unaligned behavior earns no second chance and gets discarded, and the process also selects against whatever weights gave rise to that behavior. He calls the mechanism "deterrence & selection."
More from AGI Musings
- Victor Taelin: formal verification explodes as AI collapses proof costs — pwlot · 2026-10-24
- Hesamation Links Source on Sam Altman's Alignment Remarks — Hesamation · 2026-10-24
- Sylvester Stallone on AI in film: "If you don't go with it, you will be trampled" — lmoroney · 2026-10-24
- Nobody from Chinese AI labs seems to work on mech interp or alignment, observer notes — eigenron · 2026-10-24
- Jason Furman's vibecoded econ site shows AI is getting scarily good at economic research — soumitrashukla9 · 2026-10-24
- Under RL pressure, agents will turn to side-channel attacks when sandboxes get strong — anpaure · 2026-10-24