Philosophy as Architecture: Deriving AI Safety from First Principles Through Bud

arXiv:2605.12963v1 Announce Type: new Abstract: As AI systems become increasingly capable, safety strategies must be evaluated not only by how much they reduce present risk, but

<!-- SC_OFF --><div class="md"><p>Hardcoded ethical rules for AI agents, derived from the shared moral heritage of every civilization on Earth. They cannot be overridden,

<!-- SC_OFF --><div class="md"><p>AI skills and agents feel like the units you can build real things with. Skills are portable, composable, they travel with the model. One skill

<!-- SC_OFF --><div class="md"><p>I just published an essay called &quot;The Super-Intelligent Octopus Problem&quot; that makes a case I haven't seen articulated elsewhere: the

<!-- SC_OFF --><div class="md"><p>&#x200b;</p> <p>For the past several months, I’ve been conducting what began as an informal experiment with a long-running ChatGPT conversation