LessWrong (30+ Karma) The Enemy Gets The Last Hit
Nov 24, 2025
The host dives into chess strategies, emphasizing the importance of finishing calculations after your opponent's move. This chess wisdom translates to cybersecurity, where red teams must test fixes to ensure robustness. The discussion includes the challenge of predicting adversaries, the risks of quick fixes in AI safety, and the potential pitfalls of inoculation prompting. Through various analogies, the complexities of responding to threats—whether from nature or AI—are explored, highlighting why the last hit often belongs to the enemy.
AI Snips
Chapters
Transcript
Episode notes
Always End Calculations After Opponent Moves
- Always finish your mental simulation after the opponent has moved, not after your own move.
- Verify consequences by imagining the enemy's reply to avoid missing critical counterplays.
Let The Attacker Make The Final Move
- Security exercises must let the attacker make the final move to reveal true failure modes.
- Patching without re-testing leaves unexamined ways the attacker can exploit your fix.
MAT Stream Team Deferred Quick Fix
- A MAT stream team declined a quick mitigation because they'd then have to red-team that mitigation as well.
- The team correctly noted red-teaming fixes often costs far more time than implementing them.