Seven Strategies for Tackling the Hard Part of the Alignment Problem — AI Alignment Forum
Thanks to Michael Ripa for feedback. • TL;DR * There are two types of problems that AI systems can have: problems that we can observe during development and problems that we can’t. The hardest par…
x Eight Strategies for Tackling the Hard Part of the Alignment Problem — AI Alignment Forum AI Frontpage 18 Eight Strategies for Tackling the Hard Part of the Alignment Problem by scasper 8th Jul 2023 8 min read 11 18 Thanks to Michael Ripa for feedback. TL;DR There are two types of problems that AI systems can have: problems that we can observe during development and problems that we can’t. The hardest part of the alignment problem involves problems we can’t observe. I outline seven types of solutions for unobservable problems and taxonomize them based on whether they require that we know wha
related reading
- Alignment remains a hard, unsolved problem — LessWronglesswrong.com
- Alignment is not solved but it increasingly looks solvablealigned.substack.com
- What failure looks like — AI Alignment Forumalignmentforum.org
- Recommendations for Technical AI Safety Research Directionsalignment.anthropic.com
- Model Organisms of Misalignment: The Case for a New Pillar of Alignment Research — AI Alignment Forumalignmentforum.org
- Current AIs seem pretty misaligned to me — LessWronglesswrong.com
- Another (outer) alignment failure story — AI Alignment Forumalignmentforum.org
- Deep Deceptiveness — LessWronglesswrong.com
- Automated Alignment is Harder Than You Think — LessWronglesswrong.com
- Self-Fulfilling Misalignment Data Might Be Poisoning Our AI Modelsturntrout.com
- Can we safely automate alignment research? - Joe Carlsmithjoecarlsmith.com
- Alignment Faking Mitigationsalignment.anthropic.com