The 'Strong' Feature Hypothesis could be wrong.

Posted on August 2, 2024

You can read this essay on the Alignment Forum here, or on LessWrong here. LessWrong might be a bit better because there are some interesting comments. I posted this essay on LW instead of my own personal site because I thought this would give it a better chance of reaching it’s intended audience, but I’m finally getting around to linking it from my own personal site.