nightlycosts

A technical person will also have have met those problems related to context and will know to drop a simple

"update an AGENTS file with relevant information"

To be able to navigate that faster on longer tasks. Meanwhile, the lay person does not even conceive of the LLM as a file reading entity. To them, its machinations are its own, so these types of "dumb" (simple) solutions are not even on the deck of cards.

Specially in an article that initiates the conversation advocating for setting a machine that stays on overnight running agents making rebases among who knows what else. There are two other frontpage threads right now tangentially about the cost of AI, I'll have this conversation there, to see how well the idea this cost is gone holds up against a primed crowd.