Working on a fantasy-console-thing and using AI to write a load of unit tests for me.

In my defence, I've told it not to tell me how to fix things if anything fails. I also don't want to know its opinion on what I've been doing. It creates the tests and I just run and fix.

647 tests created for my compiler and assembler. 318 tests are failing. I'm now enjoying fixing the things properly.

Instead of a project destined for the bin, I'm making progress.

Finally found a good use for AI!