On: Why LLMs can't really build software

[ambicapter]

> I had a whole laundry list of things I wanted to change about my codebase

I always have a whole bunch of things I want to change in the codebase I'm working on, and the bottleneck is review, not me changing that code.

[lordnacho → ambicapter]

Those are the same thing though? You change the code, but can't just edit it without testing it.

LLM also helps you test.

[marcosdumay → lordnacho]

Review is not test. Testing does almost not help making your program correct, and does not help at all making your code "good".

Almost every quality software has is designed in from a higher abstraction level. Almost nothing is put there by fixing error after error.