Your honor! I didn't steal anything!

I just mis-estimated where my gun was and where his wallet should go! I thought it belonged in my pocket.

I've never seen any court rule that compilation is a copy protection.

Sometimes these alerts don't work. A 9.9 earthquake and there would be no buildings left. It was a false alarm, but all the trains and such had to stop.

If you're going to make the reviewer responsible for quality, then you need to give them time to do a proper review. And a full review takes about as long as writing the code in the first place (you have to test every line, look for edge cases, negative tests, etc), so you need to schedule time in your budget for developers to do that.

And no one schedules enough time for a full review.

Competent programmers can work without really competent code review. They prefer to have code review, but if it's not present, their output is still consistently high quality. It's a multi-step process to ensure quality.

I found this other article that talks in more detail about what he's been doing with it.

You are right. Maybe Linux Foundation members have been pressuring him to use it for something in the kernel. They pressured him successfully in the past.

No one else wants them.

This article does a better job showing what Linus has been working on with LLMs.

I guess the key is to use AI to do something you're bad at. Apparently you are good at making lovely documents, so that wasn't effective.

People who talk about "junior devs" don't know much about the software industry.

Those are functional tests, not unit tests (unless your specification was so detailed it was written in UML).

That is begging the question. What I mean is, a lot of people want more complexity, because they want things to take longer, because they make more money. Middle managers want programmers to be busy but go slowly so they can justify hiring more people. There was an interview with Marc Andreesen about that, and Bjarne Stroustrup earlier.

Bottom line: faster isn't better.

Generally if you document your interfaces, that's enough. The comments end up being useful because then I can just read the comment to figure out what a function does (and what values are valid, how it behaves on error) instead of spending time looking through the code to figure it out.

Get the interfaces pretty. The internal code can be fixed later if needed.

If you want to do a good job with code review, it takes about as much time as writing the code. You need to test every line, etc.

Most of the time we trust the developers so as reviewers we just skim through the process, but that's harder to do with AI.

Probably not.