To be fair, if he said what he wrote here in private, it’d still be extremely bad leadership.

Obviously he should correct them and point out why, but maybe not trash on them entierly.

Oh, so you want to make a website where you post personal pictures of youself and post updates of your everyday life, while connecting to your friends? It’ll be really big you say? There’s nothing else on the market lile this? Let’s go!!!

I agree! It has been a great help in those cases.

I just don’t believe that it can fullfill the actual need for sites like StackOverflow. It probably never will be able to either, unless we manage to make it learn new stuff without reliable sources like SO, while also allowing it to snap up these obscure answers to problems without burying it in tons of broken solutions.

I honestly believe people are way overvaluing the responses ChatGPT gives.

For a lot of boilerplating scenarios or trying to resolve some pretty standard stuff, it’s good.

I had an issue a while back with QueryDSL running towards an MSSQL instance, which I tried resolving by asking ChatGPT some pretty straightforward questions regarding the tool. Without going too much into detail, I basically got stuck in a loop where ChatGPT kept suggesting solutions that were not viable at all in QueryDSL. I pointed it out, trying to point out why what it did was wrong and it tried correcting itself suggesting the same broken solutions.

The AI is great until whatever it has been taught previously doesn’t cover your situation. My solution was a bit of digging in google away, which helpfully made me resolve the issue. But had I been stuck with only ChatGPT I’d still be going around in loops.