[discardable_dan]

I've had it write Scheme with little issue -- it even completely the latter half of a small toy compiler. I think the REPL is the issue, not the coding; forcing it to treat the REPL like another conversation participant is likely the only way for that to work, and this article does not handle it that way. Instead, hand it a compiler and let it use the workflow it is optimized for.

[antonvs]

Agreed. The article bemoans the fact that AIs don’t need to work in the inefficient way that most humans prefer, getting micro-level feedback from IDEs and REPLs to reduce our mistake count as we go.

If you take a hard look at that workflow, it implies a high degree of incompetence on the part of humans: the reason we generally don’t write thousands of lines without any automated feedback is because our mistake rate is too high.