Post

My impression is that AI coding is on a "pick two of three" triangle: scope, ship speed, and correctness. Small tools are (speed + correctness) fast prototypes are (speed + scope) and big projects are (scope + correctness). Every project that tries to be all three degrades to "fast prototype."

So far, I think scope has been an incredible challenge due to limitations on context. Even with a divide and conquer strategy, getting large, ambitious projects across the finish line without a ton of stubs, partial implementations, or thrashing is nontrivial.