Comment by darenr
1 hour ago
The question is really why are you asking your coworkers questions that can be answered by AI. Likely you're putting the burden on them because you're tired of talking to AI. This is learned helplessness.
If instead you ask questions that are targeted, like "did you look at the PR I pushed yesterday, do you think this is something we can get into production soon?" then you'll get a human answering. I would posit that anything than can be answered by an LLM adequately, should be, then when you ask a coworker a question be direct in what you want. Ask for their input. for example, instead of "what alternatives do we have to <tech of the day>?" instead ask "I've researched a few alternatives (to <tech>) but I'd like to see what your experience might tell us about each of these that seem promising..." It all comes down to what and how you're asking.
You'll burn through your goodwill with coworkers if you keep asking questions than can be adequately answered by an LLM.
> "did you look at the PR I pushed yesterday, do you think this is something we can get into production soon?"
ChatGPT, look at the status of the PR and compare it to the flow chart of how we merge into production. Check against the milestones for this quarter, too. Respond in a short but cheerful manner.
If you read their other article on AI, you see the general problem.
https://orchidfiles.com/ai-will-build-your-roadmap-in-ten-se...
> Build a 7-year growth roadmap for my project
Given a prompt like that, you are going to get equally trashy responses. You feed it slop, it's going to give back slop.
They aren't putting any effort into quality questions or requests. Maybe these aren't accurate to what they actually wrote or the actual response, but then that just means they are lying and can't be trusted.