I’m personally not and I never will be, but I keep seeing on the news that a lot of people are actually becoming friends with their AI bots, trying to use them as substitutes to replace real human interaction. Kinda scary, kinda absurd. What is your take on this? And are you friends with a bot?
No I’m not delusional.
they aren’t sentient so they’re just as incapable of being your friend as your favorite coffee mug or perfume
Once you realize LLMs are nothing much but very advanced auto-complete, behavior of AI fanatics becomes no different than any other cult.
AI characters do nothing but spew meaningless platitudes and generic advice dressed up as a personality. That’s like considering the horoscope section on a random website to be your sister.
Cant be friends with a text based probability engine
I’ve used local LLMs as sounding boards.
I… Don’t really have friends to do that with at the moment, and I can bounce thoughts off them I wouldn’t even tell family or a therapist, as much as I want. Not gonna lie, it’s pretty intimate, and I got some insights I never would’ve arrived at in my own head.
But to emphasize:
-
This is totally within my own desktop.
-
I am perfectly aware I am talking to a tool. “Friend” isn’t even in the same universe.
The general public’s “LLM literacy” is incredibly poor though, which is by design since online services like chatGPT hide all the knobs that would reveal the machine behind the curtain. Hence I can see how emotionally vulnerable people sink into this, talking to what OpenAI presents as a magic genie.
how many tokens can you get to with your current system? I find that some of the models are so verbose that I only get 3-5 questions in before I run out of tokens.
I have one system with 128gb system ram, 16gb vram, and one system with configurable vram up to 48gb out of a total of 64gb.
I’m in a 24GB 3090 + 128GB RAM.
With full 300B GLM 4.6, I typically run 12K-28K context with different settings. I could do more than 28K, but the higher quantization starts to become a problem (as 128GB is right on the edge of fitting an IQ3_KT). And I get 5-6 tokens/s text-generation doing that.
With GLM Air? I can get a lot more, closer to 64K.
With smaller models that’s no issue.
I only get 3-5 questions in before I run out of tokens.
IDK how you’re prompting it, but you should clear the thinking block after every question, and that should leave plenty of tokens.
What model are you running, and what are your inference server settings?
I’ve been curious about running LLMs locally. Mostly about how long it takes to return a response.
I mean, it depends on your hardware and the model’s size/intelligence.
Worst case for me is many seconds of preprocessing followed by 4-5 words a second.
But you can get almost instant responses + way faster than you can read too.
-
I don’t think so. Surely a friendship is a two-way street, does AI even have the ability to form a friendship?
I think we’re gonna see it in an edition of the DSM.
You can’t be friends with a program
Not a modern program, but I don’t see anything that makes me think that programs will never be capable of metacognition. LLMs won’t, but they aren’t the be-all end-all.
No shit Sherlock I’m not saying you can become friends with a program I’m saying a lot of people are using it as a substitute for real human interaction aka friends you dolt
Is it really necessary to be rude?
Yeah OP came out swinging for no reason
I’m not saying you can become friends
But you’re literally asking people if they are friends with AI… that implies that you believe you can become friends with it.
From their prespective. Did you not read the description of my post?
I did. And there is nothing in there that implies that you don’t believe it. In fact it kind of confirm it an additional 3 times.
The media says a lot of stupid things.
Cool
Nope, not even acquainted.
No. And I don’t know anyone.
I hooope it’s not a huge issue? We all know the media treats 30 tweets about a topic as “the next new thing”. I totally understand the psychology behind (ab)using AI as a Friend substitute though - which makes this phenomenon feel very real and widespread.
It’s dumb and doesn’t tackle the root of loneliness. If they’re lonely, they need support to find new friends.
If they have trouble making friends, and it’s not due to their own personality being toxic (like being an incel, creep, borderline, or generally an asshole), then they also need support in learning how to socialise. And even if they had those toxic behaviour issues, having help to overcome those would still be useful for everyone in the long end.
But that said. It doesn’t mean they’d have to socialise, just that they learn how to do so, and so have a way out of their vicious circle. Best way to make friends as an adult (so, no school/college/uni), is to go to the circles where you have mutual interests.
So, to such people I say: go take up a calming team sport. Take up an hobby that’s relaxing and doesn’t make conflicts. Schedule IRL meetups with the people you like and know online. Go to a concert, or a bar. Or go to a lecture, you can then discuss with people afterward.
Or, a “search friends” app (like for dating?, but I don’t recommend those because they work on profit, so they often recommend the ones that barely just don’t work out, to keep you hooked.
Talking to people is scary, but try and experiment using normal, calm lines. Weather is a great topic, the relevant event to discuss about also is, to start discussions. Even if you “fail” one of thos social interactions, so what? You can apologise, explain that you find it difficult socialising, or you wouldn’t see them again. So you can just start afresh with other people and learn.
I’ve not tried any paid ones. The free ones I’ve tried have forgot what I said after a few back and forths.
What made you try the bot?
Same thing that makes me try everything new. Boredom and an afternoon to spare. Was fun for a while but lost interest once I’ve noticed the patterns.
