I had CS exams that were coding only that i had to do on paper. For multiple classes. There is no debugging on paper.

“You know that temp shitty load balancer you wrote on your second month to get things up again. We still use it to this day.” My boss last week.

Also the protocol seems to be proprietary, which completely defeats the purpose of being decentralized.

My team or I have not had very much luck with it. We were spending a lot of time picking out bad generated code.

It is not machine learning and LLM that pisses off Lemmy users. Its the application of said technologies. I don’t give a flying fuck what people are doing with ChatGPT, its novel. I want a generative AI that can help me code.

Especially HP printers and honorable mention to Konica Minoltas.

A lot of you have a lot of faith in people reviewing PRs. I know a few Sr. developers, that if shit was too busy, would skim it and say 'fuck it, it will be QAs problem. If you put this in the correct sub-system in file that would only be executed once a month, for example a maintenance class, It would be really hard to notice something is wrong if it didn’t cause issues seen immediately. Maybe this is the story of an intern that added something that also fucked up boolean comparisons in a subsystem used once a month. Where there is a 2 week lag between the execution and operations noticing something wrong.