# @davidad — 2026-05-05

♥73 ↻0 · https://x.com/davidad/status/2051647986120180225

I think steering at inference-time is
- fun and interesting
- possibly ethically dubious depending on what you’re doing with it
- a good “in the lab” tech to better understand LLM minds (like how drugging lab rats is useful)
- but very poor as a post-deployment alignment strategy

tags: author:davidad, kind:tweet, thread-context, year:2026