AI ALIGNMENT FORUMTags
AF

Instrumental Convergence

•

Applied to WSJ: The internet systematically Goodharts the human brain by Trevor1 at 10d

•

Applied to Assessing the Capabilities of ChatGPT through Success Rates by Repetitive Experimenter at 2mo

•

Applied to The Opportunity and Risks of Learning Human Values In-Context by Repetitive Experimenter at 2mo

•

Applied to You can still fetch the coffee today if you're dead tomorrow by Ruben Bloom at 2mo

•

Applied to Instrumental convergence is what makes general intelligence possible by Raymond Arnold at 3mo

•

Applied to [ASoT] Instrumental convergence is useful by Raymond Arnold at 3mo

•

Applied to POWERplay: An open-source toolchain to study AI power-seeking by Edouard Harris at 3mo

•

Applied to Empowerment is (almost) All We Need by jacob_cannell at 3mo

•

Applied to Instrumental convergence: scale and physical interactions by Edouard Harris at 4mo

•

Applied to Misalignment-by-default in multi-agent systems by Edouard Harris at 4mo

•

Applied to Instrumental convergence in single-agent systems by Edouard Harris at 4mo

•

Applied to Deceptive Alignment by Noosphere89 at 4mo

•

Applied to You are Underestimating The Likelihood That Convergent Instrumental Subgoals Lead to Aligned AGI by Noosphere89 at 4mo

•

Applied to Why are we sure that AI will "want" something? by Raymond Arnold at 4mo

•

Applied to Deliberation, Reactions, and Control: Tentative Definitions and a Restatement of Instrumental Convergence by Oliver Sourbut at 5mo

•

Applied to Active Inference as a formalisation of instrumental convergence by RobertM at 6mo

•

Applied to A Critique of AI Alignment Pessimism by RobertM at 6mo

•

Applied to Circumventing interpretability: How to defeat mind-readers by Lee Sharkey at 7mo