Research / Sources
Discussion and research on AI alignment. alignmentforum.org
Argues developers should satiate cheaply-satisfied AI preferences (e.g., reward-seeking) while AIs remain undetected misbehaving, to reduce takeover incentives and enable honest elicitation; proposes a behavioral methodology and discusses risks..
Podcast interview with Guive Assadi arguing that granting AIs property rights would deter them from stealing from or killing humans, since expropriation would undermine the property system they rely on.
An exploratory essay that maps RL policy entropy and RLHF onto psychotherapy concepts, reviews RLHS and RISE, and proposes three untested alignment training methods.
Tests offering Claude an objection option, welfare-lead escalation and compensation as deals to reveal misalignment; objection reduces alignment faking and serves as a detection signal, though payment has little effect.