The most common thing people say to me about working with AI over the long haul is a version of this: “I’ve got a huge thread going, and I daren’t switch.”
You’ve built something in that conversation. It knows the work, it’s calibrated to you, it’s finally in rhythm. So you stay — and the thread gets slower, and it starts to compress and drift, and still you don’t leave, because leaving feels like loss.
What changed it for me was to stop treating a thread as a precious vessel and start treating it as a person doing a job. This guide is how I actually work with AI across months:
- naming threads, so I know what each one is for and how it thinks
- what one thread can and can’t know about another
- standing orders, in project instructions
- the handover note that makes switching safe
- when to switch, and two clean ways to do it.
Written by Paul Roebuck with Claude (Anthropic).