I may be old school, but I still find it crazy that people just allow AI to write code. I use AI a lot when I'm coding, but I always ask, review the ouput then copy and paste what I actually need.
I constantly find it outputting things that are wrong, not handling errors properly, or not worrying about security.
yeah totally. an example from the article is that when when you're reviewing 31 lines of business logic instead of 150 lines of boilerplate it's a lot easier to catch bad error handling or security issues - which kinda goes hand in hand with what you're saying.
A bit of prompting to ask it to not throw away errors and write TDD-style helps imo.
yeah we're seeing the same thing from the infrastructure side. small opinionated api surface means less code and you can actually read what the agent wrote.