I generated a script today to diff 2 CSVs into a Venn diagram, ran it twice, then deleted the code.

The LLM itself could do have donne it, maybe you didn't need the code at all

It's a language model, not a compiler. Which is what people get wrong.

Ask one to count the 'r's in "strawberry" and it may or may not get it right.

Ask it to create a program to do it, it'll get it right instantly and it'll work.

When we get to a point where "AI" can write a program like that in the background, run it and use its result as a tool, we'll get the next big leap in efficiency.

I think the future of computing is ephemeral code like this, created rapidly on demand, then done when the immediate task is done.