Risk from fitness-seeking AIs: mechanisms and mitigations

Fitness-seeking is increasingly what misalignment looks like in practice—how should we respond?

Current AIs routinely take unintended actions to score well on tasks: hardcoding test cases, training on the test set, downplaying issues, etc. This misalignment is still somewhat incoherent, but it increasingly resembles what I call “fitness-seeking“—a family of misaligned motivations centered on performing well in training and evaluations (e.g., reward-seeking). Fitness-seeking warrants substantial concern. In this piece, I lay out what I take to be the central mechanisms by which fitness-seeking motivations might lead to human disempowerment, and propose mitigations to them. While the…

related reading

• Risk from fitness-seeking AIs: mechanisms and mitigationssubstack.com
• Fitness-Seekers: Generalizing the Reward-Seeking Threat Model — LessWronglesswrong.com
• Fitness-Seekers: Generalizing the Reward-Seeking Threat Modelsubstack.com
• The behavioral selection model for predicting AI motivations — LessWronglesswrong.com
• What failure looks like — LessWronglesswrong.com
• What failure looks like — AI Alignment Forumalignmentforum.org
• Current AIs seem pretty misaligned to me — LessWronglesswrong.com
• The case for ensuring that powerful AIs are controlled — LessWronglesswrong.com
• Are we existentially threatened by the type of AI misalignment seen in the OpenAI Hugging Face attack? — LessWronglesswrong.com
• Fail safe(r) at alignment by channeling reward-hacking into a "spillway" motivationblog.redwoodresearch.org
• Current AIs seem pretty misaligned to meblog.redwoodresearch.org
• New report: “Scheming AIs: Will AIs fake alignment during training in order to get power?”joecarlsmith.com