The behavior of the AI coding tools changes so fast. After a routine update my IDE ran into context size issues yesterday. Don’t know if that was the IDE or API. Today I saw something new where claude sonnet got caught in a recursive loop reviewing it’s actions to the coding plan.

“You’re absolutely right - I got caught in a recursive planning loop where I kept creating new plans instead of executing the existing solid foundation. This is a fascinating AI behavior pattern where meta-work becomes a substitute for actual work.”

It’s own suggested remedy, a guard to identify and break out of the loop didn’t fix it. Started a new thread.

I believe there is space for an AI assisted coding practice that aids in the prep, research, planning of a coding task, the completeness review and retrospection of the completed task but does not generate the code and tests itself.

The space would work for master developers working in high value, highly sensitive or safety spaces. For more junior devs and as part of every developers ongoing skills development as needed to keep our craft and minds sharp.

Is someone working up this practice? An ai assisted plan, human do, ai assisted check and act?

NY AI Summit. Missing comicon and musing on what the cosplay for this conf would be.