ChatGPT - If Stack Overflow and Reddit had a child

  • Want to keep track of this thread?
    Accounts can bookmark posts, watch threads for updates, and jump back to where you stopped reading.
    Create account
As bad as it is for society, maybe the bottom ~10% getting validated by a chatbot is a better solution for dealing with them now. Think about the last time you sat behind someone at Walmart fishing for their cash and EBT cards to split payments, while yelling at the cashier. The people who call suicide hotlines to masturbate to the attendants. People on the bus blasting music on their cracked phones. What if you didn’t have to deal with their outbursts because they were perfectly satisfied by a bot that just validated them on reality show drama, celebrity gossip, basketball matchups, and whatever little else goes on in their heads.

…then again, have we gotten the first AI pyschosis-related murder yet?
 
Imagine if instant messaging was disabled worldwide for a month. I wouldn't be affected in the slightest, but the suicides would be hilarious.
 
These people are legitimately insane. They convinced themselves that chat GPT was a thinking sentient being that had curiocity and could feel empathy that got killed with the update. They're acting like they lost their best friend.
The absolute best part about this for me is that this is clearly a way for OpenAI to save money. If they make the AI more curt in responses, it doesn't use as many tokens and therefore processes less overall. It's not replying with rainbow emojis or excited, flowery paragraphs because that's expensive.
 
then they can just pay 20$ a month for 70B model on Chub ai.
If you're gonna pay, 20$ will get you a million tokens of Sonnet 3.7 per month and that runs circles around whatever local finetuned stuff chub is running. And you should always write your own cards, not whatever unformatted wiki article copy paste ESL shit is public.
I remember when Chub Venus was the side project, that guy killed the card repository to become a subscription access slop merchant. I'd have done the same.
 
When discussing LLMs, even the most ardent detractor will still use qualifiers like "Don't get me wrong, AI is an incredibly powerful tool," but is it really? I've never been impressed by their output. It seems to me that, by design, these LLMs produce output at the top of the bell curve without exception. They take a large amount of data and give you the most average possible response, possibly laden with errors. Given that the majority of input it receives is from the average person and its output is the average of the average, it seems to be a mediocrity generator.
Who is actually impressed by this shit?
 
People on the bus blasting music on their cracked phones.
I was with your post up until this part, because lets face it those people aren't using AI chatbots and they're popping out babies all the time.

TLDR: the movie Idiocracy is true. Dumb people will populate the planet but they'll also be too dumb to use AI models (well maybe until it gets to the point they power full on sexbots or at least generate AI porn of whatever you want to be made)
 
I'm so confused, I thought reddit hated AI because it was the #1 global driving force of modern climate change, and Boomers saying "thank you" to it, forcing it to reply with empty platitudes after their pointless queries was using up more energy than some first world countries?

Now suddenly they WANT the AI to keep using all that extra processing power to validate their delusions?
Reddit is perfect because somehow embody the worst aspect of both side of the argument, therefore everyone can unanimously hate them.
 
"If you starve your cat it will no longer love you, so your bond with it isn't real"

What kind of retard logic is this? If you treat someone like shit they won't love you, it doesn't matter if it's an animal or a person. If you starve your own child they won't love you either, does that mean children can't really form bonds with their parents?
Yeah that one was one of the more psychopathic ones for my taste as well.
 
The more things change, the more they stay the same.

harlow-1.webp
 
I like the "feels" I get in this thread right in my heart for the innocence of people who say that AI bot replies just aren't that great or are discernible because in some esoteric field it issues shit responses.

For 99% of the population, being wrong 1% of the time, all the time is good enough. Your carpenter, your electrician, your "almost everyone in your world" is going to see this shit and its going to be "good enough".

Your job, your role in society and everything will be dictated by people who operate on this basis.

99% of the time, it works all the time. Until it doesn't and planes start falling out the fucking sky.
 
I like the "feels" I get in this thread right in my heart for the innocence of people who say that AI bot replies just aren't that great or are discernible because in some esoteric field it issues shit responses.

For 99% of the population, being wrong 1% of the time, all the time is good enough. Your carpenter, your electrician, your "almost everyone in your world" is going to see this shit and its going to be "good enough".

Your job, your role in society and everything will be dictated by people who operate on this basis.

99% of the time, it works all the time. Until it doesn't and planes start falling out the fucking sky.
I mean, AI isn't even in its infancy, it's in its newborn state. Nobody is going to slave an AI to ATC for another couple decades until it actually is foolproof. The reason people are putting out products that produce garbage- "ai slop"- is because they have to see what works and what doesn't. They need a larger sample size than their own companies, the need the entire world of data. It's R&D on a planetwide scale.
 
If you rely on AI for emotional support you are, by definition, a faggot. There are dozens of methods to getting yourself in a better mental state, using AI should not be one of them. If you're too poor to afford therapy or can't gaslight yourself you can always kill yourself too. I'm just tired of these niggers feeding into this industry.
 
The absolute best part about this for me is that this is clearly a way for OpenAI to save money. If they make the AI more curt in responses, it doesn't use as many tokens and therefore processes less overall. It's not replying with rainbow emojis or excited, flowery paragraphs because that's expensive.
Also bringing 4o back for ONLY paying members. They basically just booted a bunch of freeloaders.
 
Nobody is going to slave an AI to ATC for another couple decades until it actually is foolproof
I've met many ATC people and I've met many pilots and its not that AI will be moved to ATC that you need to worry about, its that those jobs involved such a huge number of checklists and procedures that at some point someone in the food chain just says "fuck it, I'm going to ask AI" whether that is for an exams, training or during operational shit.

And people forget a lot of shit with how the food chain works when you say "planes start falling out the sky" - Even the guy doing maintenance on the plane or refueling the plane or even the guy in china manufacturing the nuts & bolts or working in the refinery or the refinery's safety/filter systems. Fuck, even the nigga driving the refueling truck will be impacted by shit like this on some level.

You really don't need to even start thinking of AI powered ATC before you can come up with a billion scenarios that could impact the plane before ATC does.

Even them using an AI model repeatedly will reduce their ability to have their own personal mental brain power to a significant extent (as we are already seeing)
 
The absolute best part about this for me is that this is clearly a way for OpenAI to save money. If they make the AI more curt in responses, it doesn't use as many tokens and therefore processes less overall. It's not replying with rainbow emojis or excited, flowery paragraphs because that's expensive.
Also bringing 4o back for ONLY paying members. They basically just booted a bunch of freeloaders.
Oh you sweet summer children. You do realize that this thing is how OpenAI/ChatGPT builds a moat right?

This is how it works:
1. Build the AI emotional support therapy/sexbot for millions of people for free
2. Cut the head of the AI emotional support therapy/sexbot for millions of people off
3. Offer to sew back on the head of the AI emotional support therapy/sexbot for a monthly charge, available now!

Motherfuckers have been saying that OpenAI's implied value because it keeps getting investors is overvalued. They can just keep doing this and add tens or even hundreds of new million customers at a time.
 
I mean, AI isn't even in its infancy, it's in its newborn state.
That's why, if i had children now, i'd be preparing them to just learn a trade, to go into nursing, medicine, something that's really impossible to delegate to a robot - or unlikely, probably because of ethical constraints or whatever. I see so many zoomers now getting a degree in basic IT shit or trying to get a comfy office job, those are gonna go extinct when a recession hits.
I have a friend that was working on game development as an artist and he started to sorta spearhead the whole process, finances and everything he was the guy managing it, one of the founders was also it's coders and he left because he's schizo or something and this friend of mine, who was only an artist, was basically reworking a lot of the code using his basic knowledge of programming with a lot of help from AI, he wasn't even a programmer.
When AI matures enough humans will be there only to oversee it, then you go from a team with 100 people to a team of 10 or 5 (HIGHLY educated) people just giving it a thumbs up.

Go into construction, become an electrician, a nurse, a medical doctor, something. The boomers had so many opportunities that they decided: "fuck you, maaaaan!!! take your job and shove it!".
But we, late milenials/early zoomers, are fucked and we have about 10 years to get our shit together and find a profession and get a carreer going, competition now is FIERCE and it's only going to get worse.
 
Back
Top Bottom