Chatgpt, are we cooked?
this post was submitted on 08 May 2025
103 points (100.0% liked)
Slop.
489 readers
441 users here now
For posting all the anonymous reactionary bullshit that you can't post anywhere else.
Rule 1: All posts must include links to the subject matter, and no identifying information should be redacted.
Rule 2: If your source is a reactionary website, please use archive.is instead of linking directly.
Rule 3: No sectarianism.
Rule 4: TERF/SWERFs Not Welcome
Rule 5: No bigotry of any kind, including ironic bigotry.
Rule 6: Do not post fellow hexbears.
Rule 7: Do not individually target other instances' admins or moderators.
Rule 8: Do not post public figures, these should be posted to c/gossip
founded 5 months ago
MODERATORS
now that we rely on it, we can't really imagine living without it
It's been like 3 weeks. Most people aren't even using it to that extent.
It's been like 3 weeks
it's been 3 years 🫠
No no, covid happened just last summer...
2016 was just a few weeks ago...
"AI" has been a consumer product for less than two years. I think it's actually quite easy to remember a time without it assuming one's brain isn't smoother than mirror.
I do not understand these people who have convinced themselves that they need a search engine that pretends to think
It's not a search engine either
No, but the statistical model is at least better at that than pretending to think
Hardly, since if it isn't just running a search in the background it's outputting something statistically similar to a search output, which is just as likely to be false as true. And if it's running a search what was the point instead of just searching?
It does search the internet now, so it kind of is.
It's a search engine the same way asking a friend to google something is a search engine
Yes, it's very stupid that anyone thinks they need this.
As a USian we've been outsourcing our thinking for years. I actually find myself enjoying conversations with batshit conspiracy theorists sometimes as many people don't think about their assumptions in any capacity and would outsource their assumptions too if they had a chance to. Many don't like having to think in any capacity.
I think this is part of the appeal of conspiracy theories in the US. These people are thinking completely incorrect things, but they are thinking. A conspiracy theory actually does involve the brain at some point, whereas so much of everyday life in the US seems to heavily discourage actually thinking or considering anything, everything is about consumption, not creation, including creation of thoughts and ideas.
Why think when I can be mad?
Honestly I'm coming at this from specifically a mid-european-city-carbrain standpoint and I understand it entirely. I know I'm the guy with the hammer to whom everything looks like a nail but I still argue it is the exact same process of not ever considering how this surface level choice of convenience makes you both entirely beholden to like 3 large multinational corporations and also gives you the kind of brainworms that when that is taken away you think your life is over
In the US at least, we live in a society built on convenience. Anything that can reduce any amount of workload is marketable. This is the culture that greenlit Juicero to have a robot squeeze juice from a bag.
Even if it's easier to do something yourself, many people will let a machine do it for them on principle.
In my last department, we sold peeled mandarins (Halos). You got maybe 3-4 mandarins for 3.49. Or you can buy a 3lb bag for like 8 and peel them yourself.
I had a customer bring me a watermelon off the sales floor and ask me to cut it. I threw it on my scale, subtracted 20% (pulled out of my ass) for the rind, and told her I could have it done in about a minute and a half, but I'd have to charge her like $60. Or she could pay the 10.99 and spend 10 minutes doing it herself. She did take my advice, to her credit.
The amount we will pay for convenience is ridiculous. But also after work I'm exhausted, and am far far too often guilty of buying a frozen pizza or takeout, when I should spend like $15, cook, and feed myself for a few days.
(cw meat) My grocery store sells chicken quarters for $1/lb... or you can buy drumsticks or thighs for $2/lb, or the mixed bag from Tyson which is about $3/lb. I genuinely don't understand how anything except the big bag has any sales at all.
How long until we get fully grown adults who can’t pass the Turing test
About negative 10 years I'd say
Ooh, an optimist
That's why I'll never support it in the arts at any level. The most important part of creativity, the central feature that makes it human and worthwhile as a product, is what you don't put on the canvas. Decades of learning from mistakes and different lived/absorbed influences makes your work good. You need to fail and then self-reflect on that failure to grow, just as anything else in life. Using AI as a timesaving crutch completely annihilates that learning process.
You're not spending hours or days or years redrafting something until it actually reflects everything you want to communicate after learning from your own work. You're not identifying your weaknesses and addressing them with practice. You aren't stewing on it for a year while you read three other books that would greatly improve the one you're writing with issues you never thought about. AI lets you instantly shit out whatever you impulsively feel or want in the moment. If you spend that year reprompting it instead of consciously working through a problem, you don't learn anything from that and only further rely on the hallucinated word diarrhoea of redditors to form your worldview. Whatever model you use to think for you, a better one will make a better product for free and I have zero incentive to bother with yours.
AI exists to bankrupt tech demons by saying "thank you" to ChatGPT after using it to cheat on homework.
Yeah, but have you considered that typing "pretty landscape picture" generates just as much value as an artist who put their heart and soul into a piece? Of course you haven't, souls don't exist, love is just a chemical reaction in the brain, nothing matters, you can't determine what is and isn't real, this is probably all just a simulation therefore you are an AI, (not me though, because I am enlightened by my own intelligence). 
Is this a world record for the self-dehumanization speedrun?
Damn. Makes you ~~think~~ offload your thinking to a VC psycho's pet god.
I think everyone should listen to the recent TMK episode on slop world, and how it dovetails with the degradation of the internet into a big ol' chumbox. It's all downhill, and this hijacking of the beautiful messy internet lots of us grew up with is only intensifying more and more, like a vicious cycle.
We live in the guts of a perpetual value-generating machine that works by feeding AI-generated slop ads to AI bots. We have a bastardized form of the internet, crafted by inbred chatbots, which is increasingly ceasing to contain anything informative or important. There are islands of content where we are less exposed to this weapons-grade bullshit, but the sea level is rising, and there's no holding it back. It's a race to the bottom in which instead of humanizing technology, we're trying to technologize humanity.
My experience with so-called "AI" thus far:
As a child in the 2010s: I occasionally used Cleverbot and This Person Does Not Exist as fun toys during recess.
Early 2020s: I goofed around with Character AI to see fictional characters "come to life" and to see if AI worked for language learning. I quickly grew bored of Character AI after maybe a day or two, and decided it didn't really have any practical uses for me. I also used 15.ai for a while — a text-to-speech website trained on various fictional characters — back when that was a thing, and although I had some fun with it and thought it had some nice features and potential, I still felt it was too limited for me to have much use for it, and so I eventually stopped using it, and the website eventually went inactive.
Mid 2020s: I have by this point started using Retrieval-based Voice Conversion and This Person Does Not Exist in personal projects like worldbuilding and fandubbing. I have no other uses for "AI" technologies, certainly no uses for text generation aside from I guess occasional machine translation.
This Person Does Not Exist
Does this have a local version like RVC does?
Yes it does, it was an Nvidia trained model. I forget the exact model name though
Edit: https://github.com/NVlabs/stylegan?tab=readme-ov-file
This is what the person does not exist ran, not 100% sure how to run or anything though.
Oh I should have figured it'd be a bit outdated, and it's not the most safe either since the standard is safetensors files. Stablediffusion is probably the better choice for this type of thing.
I've been meaning to look into that, but it isn't something I use super often anyways.
I recall seeing you commenting on language-related stuff a lot here, so if you could elaborate on how you tried to integrate it into language learning and why it didn't work out for you, I'd be happy to read it!
There isn't much to it, I just tried to use it as a conversation partner and the text it spat out was completely broken. Like it was blending English and Japanese vocabulary and grammar in a really ridiculous way.
Maybe if I tried again now and used a different service, it would work better, but I just don't think it would be worth it.
Paolo Freire would die again from a broken heart.
The whole article makes for extremely grim reading.