AI ALIGNMENT FORUMTags
AF

AI Boxing (Containment)

•

Applied to Protecting against sudden capability jumps during training by nikola 2d ago

•

Applied to Information-Theoretic Boxing of Superintelligences by JustinShovelain 3d ago

•

Applied to Self-shutdown AI by Jan Betley 3mo ago

•

Applied to Boxing by Raymond Arnold 4mo ago

•

Applied to Thoughts on “Process-Based Supervision” by Steve Byrnes 5mo ago

•

Applied to A way to make solving alignment 10.000 times easier. The shorter case for a massive open source simbox project. by AlexFromSafeTransition 5mo ago

•

Applied to [FICTION] Unboxing Elysium: An AI'S Escape by Super AGI 6mo ago

•

Applied to Ideas for studies on AGI risk by dr_s 7mo ago

•

Applied to ChatGPT getting out of the box by qbolec 9mo ago

•

Applied to ARC tests to see if GPT-4 can escape human control; GPT-4 failed to do so by Christopher King 9mo ago

•

Applied to Bing finding ways to bypass Microsoft's filters without being asked. Is it reproducible? by Christopher King 9mo ago

•

Applied to I Am Scared of Posting Negative Takes About Bing's AI by Yitzi Litt 10mo ago

•

Applied to How it feels to have your mind hacked by an AI by blaked 1y ago

•

Applied to Oracle AGI - How can it escape, other than security issues? (Steganography?) by RationalSieve 1y ago

•

Applied to I've updated towards AI boxing being surprisingly easy by Noosphere89 1y ago

•

Applied to Side-channels: input versus output by davidad (David A. Dalrymple) 1y ago

•

Applied to Prosaic misalignment from the Solomonoff Predictor by Cleo Nardo 1y ago

•

Applied to My take on Jacob Cannell’s take on AGI safety by Steve Byrnes 1y ago