Proxy misspecification and the capabilities vs. value learning race

📖 Description

G Gordon Worley III [recently complained](https://www.alignmentforum.org/posts/sSpu2EABtTTDmBZ6T/g-gordon-worley-iii-s-shortform?commentId=tbj8R6K2q9rfhjtS7) about a lack of precision in discussions about whether Goodhart's Law will present a fatal problem for alignment in practice. After attending a talk in which Dylan Hadfield-Menell[[1]](#fntkb71ztt8d) presented the "Goodhart's Law will be a big deal" perspective, I came away with a relatively concrete formulation of where I disagree. In this post I'll try to explain my model for this, expanding on my short comment [here](https://www.lesswrong.com/posts/sSpu2EABtTTDmBZ6T/g-gordon-worley-iii-s-shortform?commentId=s4qu4hNykRmv8hALo).

📊 Game Impacts

💭 Reactions

🔗 Sources

• [1] https://www.alignmentforum.org/posts/tWpgtjRm9qwzxAZEi/proxy-misspecification-and-the-capabilities-vs-value