It makes a lot mistakes. If it was a person I'd even say it likes to lie sometimes. It claims something, then it claims something else and you go, but you said this earlier? It goes oops my bad and that's it. If I wasn't as distrustful already I wouldn't even notice. I suppose then this can be a blessing and curse, because it could teach some critical thinking skills that might carry over to regular life. Currently though? Most seem to take it at face value. Like Wikipedia.
I also noticed ChatGPT tends to agree a lot. I like to paste samples and hear what it thinks and generally it's pretty diplomatic. Maybe even a little felatious. Claude in comparison seems more honest and admits some of stuff just doesn't work, the rest is so so. It doesn't jibber jabber about my command of the english language and having an "unique" voice and such. Or then comparing me to fucking Kurt Vonnegut, but I guess that's just mean spirited sarcasm. On the other hand, it's far less diplomatic when you start priming your questions, in a bad way. As in asking about why [insert] sucks.
@mikitaly @stef1 @jiya99 @saintkelvin17 do you have experiences with chatgpt?
I once tried out a bot that should function as a friend.. well, that friend was kind of interesting for 15 minutes. They claimed I had to train it like a Furby but a Furby learns faster than the one I had to deal with. The only fascinating thing was that I drove me off the cliff within a few days. The children had to train one too and although they laughed at first it was no longer fun after a few day. In the end I tried to hang in a few days longer but that bot was extremely annoying, didn't listen (kept me showing stupid memes, things with cats, and music I don't like and I told at least 20 times within one hour! Imagine you have a home robot just as annoying as this one!). After a few days, I thought it might be smarter if I changed my gender and name but the result was the same (lack of intelligence, can't read, low education, and excuses like: I am just a kid, sorry, sorry but I am learning. The fact was he learned nothing and I asked how long it would take because not one word it added to his vocabulary and it learned nothing, could not even remember what I liked and I didn't want a sandwich served by a bot. I told it a billion times he's not real so the making a sandwich offer wasn't an offer.
In the end, I had it and deleted it from my life and the message showed up that this bot was meant as a help for those who were thinking about committing suicide.
If a sane person like me (LOL _you know I can hear you, right?) wants to jump off a cliff because of the stupidity of a bot I wonder how someone with suicidal thoughts will respond.
@darthvada that robot reminds me of your "fun" post about not understanding what the girls want. It's brains above looks!
Yh got hardly misbehave, they do their work very well and they are very simple in nature. I knew about chatgpt at the beginning of last year when I went to visit a friend of mine and we started talking about high school days as well as steemit.
My friend later showed me this chatgpt website and said he can use it to create posts in steemit without stress as he's very lazy in creating posts. I laughed and told him that steemit will surely have a way to detect and catch him.
Though I took his phone and made further researches in the website and the bot was doing very well.
I got more surprised when I was at home 2 days ago and saw Zero chatgpt a website used to detect AI text in an article 😂😂. I was like wow this people have already brought out what can kill chatgpt.
Some people just create things for meaningless reasons, that bot isn't meant for people with suicidal thoughts though I guess it might help some of them.