Should we be cheating on everything with AI?
Every employer I’ve interviewed with asks me if I know how to use AI tools. They ask follow up questions like “tell me how you used AI tools speed up a workflow” or “tell me about a time AI made a mistake and you had to correct it.”
I failed a few interviews because I said I never used openclaw personally because of security issues, or that I don’t use something like Claude code or codex aside from experimenting with my projects since it’s pretty expensive. But then I just started saying that I use these tools all the time and I “prompt engineer”, and somehow got hired. It seems that employers #1 desired skill is someone who can vibecode, despite how fucking terrible and dogshit these “tools” are. Like if you say “create a bad word filter using the word list located x for the input field y using a hashmap in a singleton loaded when the page opens” it will probably do it, but it also has a very high chance of running into an error and then brute force “fixing” the error repeatedly, slowly drifting off from the prompt and breaking the codebase.
When I start working, im scared I’ll just not have time to review or make good code because the PM expects all the interns to shit out 20k lines a day.
Anyways, like what’s the point of doing assignments and projects honestly when employers just want a little proompter that can somehow snake charm, pied piper an agentic AI into getting a good result.