I got an email ban.
1609 hours logged 431 solved threads
If this is true, then we should prepare to be shout at by chatgpt why we didnt knew already that simple error.
ChatGPT now just says “read the docs!” To every question
Hey ChatGPT, how can I …
“Locking as this is a duplicate of [unrelated question]”
Chatgpt is going to get trained on thinking those two questions are duplicates and end up giving bullshit outdated answers to every question.
Always love those answers, well if you read the 700 page white paper on this one command set in one module then you would understand… do you think I have the time to read 37000 pages of bland ass documentation yearly on top of doing my actual job? Come the fuck on.
I guess some of these guys have so many heads on their crews that they don’t have much work to do anymore but that’s not the case for most
And then links to a similar sounding but ultimately totally unrelated site.
Stack overflow was the pioneer of hallucinations.
Already had that happen with perplexity, like, no mate, I’m asking you.
This message brought to you by chatgpt bot.
You joke.
This would have been probably early last year? Had to look up how to do something in fortran (because fortran) and the answer was very much in the voice of that one dude on the Intel forums who has been answering every single question for decades(?) at this point. Which means it also refused to do anything with features newer than 1992 and was worthless.
Tried again while chatting with an old work buddy a few months back and it looks like they updated to acknowledging f99 and f03 exist. So assume that was all stack overflow.
Take all you want, it will only take a few hallucinations before no one trusts LLMs to write code or give advice
[…]will only take a few hallucinations before no one trusts LLMs to write code or give advice
Because none of us have ever blindly pasted some code we got off google and crossed our fingers ;-)
It’s way easier to figure that out than check ChatGPT hallucinations. There’s usually someone saying why a response in SO is wrong, either in another response or a comment. You can filter most of the garbage right at that point, without having to put it in your codebase and discover that the hard way. You get none of that information with ChatGPT. The data spat out is not equivalent.
See, this is why we can’t have nice things. Money fucks it up, every time. Fuck money, it’s a shitty backwards idea. We can do better than this.
You can be killed with steel, which has a lot of other implications on what you do in order to avoid getting killed with steel.
Does steel fuck it all up?
Centralization is a shitty backwards idea. But you have to be very conscious of yourself and your instincts to neuter the part that tells you that it’s not to understand it.
Distributivism minus Catholicism is just so good. I always return to it when I give up on trying to find future in some other political ideology.
