> this is where I think AI for coding is now. It gets things wrong enough that I have to manually check everything This might be dependant on the programming l

> this is where I think AI for coding is now. It gets things wrong enough that I have to manually check everything This might be dependant on the programming language, some languages are way more popular and have way more questions on StackOverflow and Reddit and repos on github, so the answers will be better. When I use copilot for JS it's right 90% of the time. And where it's 'wrong' it's usually just stuff it skipped over because it didnt have proper context.