this post was submitted on 12 Nov 2023
88 points (83.8% liked)
Asklemmy
43516 readers
2163 users here now
A loosely moderated place to ask open-ended questions
Search asklemmy π
If your post meets the following criteria, it's welcome here!
- Open-ended question
- Not offensive: at this point, we do not have the bandwidth to moderate overtly political discussions. Assume best intent and be excellent to each other.
- Not regarding using or support for Lemmy: context, see the list of support communities and tools for finding communities below
- Not ad nauseam inducing: please make sure it is a question that would be new to most members
- An actual topic of discussion
Looking for support?
Looking for a community?
- Lemmyverse: community search
- sub.rehab: maps old subreddits to fediverse options, marks official as such
- !lemmy411@lemmy.ca: a community for finding communities
~Icon~ ~by~ ~@Double_A@discuss.tchncs.de~
founded 5 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
...umm...well, that's just freaking stupid.
Smarter than expecting an intellectual conversation with a LLM.
Wtf? This is a weird take lol
The ideologically opposed have very rarely actually used the machines they rail against. It was the same with the internet and the computer. Neo-luddites doing impressions of ostriches.
Iβve literally never been able to talk about the noosphere satisfyingly with a human. Powerful NLPs are very capable of having intellectual discussions, even if they donβt actually have intellect.
Every time I've had a conversation with it and it "gets deep", it changes the subject. Like this for example, something like that would never happen on there (you might say I tried just that). I would not count on it.
Tell it not to.
Make a GPT using their tool or give it custom instructions if you need to. Also try pi.ai. Itβs more informal but Iβve had great conversations with it.
I've begun to use it, though so far it's more or less paraphrasing what ChatGPT is saying. Any special methods I should use?
I'm not sure. I know it has different tunings but I don't know much about them.
Actually I do have one tip, though, which is to tell it to play make-believe, or pretend that it has an opinion and then tell you it. Many of the bots are highly pre-trained to respond asserting that they absolutely cannot have an opinion, which is more or less true. But their "made up" opinions tend to be pretty much the same regardless of when you've asked. So it can definitely make up its "mind" about something, which is of course a byproduct of biases of the texts its trained on or its application of logic and its "ethics." It makes sense within this context that their "opinions" would be the same regardless of when or how you ask.
I had a nice conversation with pi.ai about machine consciousness - but I did have to keep reminding it to pretend it had an opinion, or else it would default into the "as an AI I blah blah blah" stuff.