nooarche's Shortform
Jun 141
What if harm avoidance were structural rather than learned? *This post was written with Claude (Anthropic) as cognitive scaffolding. I have ADHD and use AI assistance as a prosthetic for written communication. The ideas, experimental design, and architectural decisions are mine; the prose clarity is collaborative.* My own take on...
On Moral Residue in Alignment Systems This post is a diagnostic argument about alignment failures, not a proposal for a solution. Introduction In alignment work, it is common to treat ethics as something that can be settled in advance: encoded into objectives, constraints, constitutions, oversight procedures, or proofs of safety....