AI Alignment Is Turning from Alchemy Into Chemistry - Alexey Guzey
Introduction For years, I would encounter a paper about alignment — the field where people are working on making AI not take over humanity and/or kill us all — and my first reaction would be “oh my god why would you do this”. The entire field felt like bullshit. I felt like people had been working on this problem for ages: Yudkowsky, all of LessWrong, the effective altruists. The whole alignment discourse had been around for so long, yet there was basically no real progress; nothing …
AI Alignment Is Turning from Alchemy Into Chemistry created: 2023-04-25 ; modified: 2023-04-25 Table of Contents 5 random top posts > Why I believe in AGI (again) > Most Important Slate Star Codex Posts > Matthew Walker's "Why We Sleep" Is Riddled with Scientific and Factual Errors > You are a morale-driven machine > Theses on Sleep Introduction For years, I would encounter a paper about alignment — the field where people are working on making AI not take over humanity and/or kill us all — and my first reaction would be "oh my god why would you do this”. The entire field felt like bull
Explore this link on the map →related reading
- Alignment remains a hard, unsolved problem — LessWronglesswrong.com
- AGI Ruin: A List of Lethalities — LessWronglesswrong.com
- LessWronglesswrong.com
- The Field of AI Alignment: A Postmortem, and What To Do About It — LessWronglesswrong.com
- ALIGNMENT - by vincent huang - a slice of my mindmindslice.substack.com
- PSA: Almost nobody is directly working on superintelligent alignment — LessWronglesswrong.com
- Can we safely automate alignment research? - Joe Carlsmithjoecarlsmith.com
- AI as a science, and three obstacles to alignment strategies — LessWronglesswrong.com
- Shah and Yudkowsky on alignment failures — LessWronglesswrong.com
- Productive Mistakes, Not Perfect Answers — LessWronglesswrong.com
- I Would Have Solved Alignment, But I Was Worried That Would Advance Timelines — LessWronglesswrong.com
- Ngo and Yudkowsky on alignment difficulty — LessWronglesswrong.com