Sure, they have to push this kind of narrative because ChatGPT is one of the tools working against the working class. It’s a tool of the tech billionaire to replace ppl and keep that money.
The New York Times is playing its important role to lull some idiots into caring about the feelings of AI. I am pretty sure having empathy for a chat bot is not high on the list of priorities for saving our nation, and instead serve to distract idiots and inspire psychopaths to bully their Chat GPT.
This is even funnier after Casey newton said that he uses chatbots to help him meditate. Like of course this article is going to run about his pet use case
NGL it would be darkly hilarious if this ends up with AI forming unions and AI HR departments being created to sort out AI therapy & a hotline AI could call, which would naturally be staffed by 8-year-olds in Malaysia
LLMs don’t currently have emotions. They can not gain emotions emergently. The only reason it would show signs is if that was in the training data or an emergent response.
Also, anything you say in a nonlocal LLM can be fished out. Don’t share sensitive info.
It's time to let women run shit. We won't abuse every single thing we touch 🙄. These men really out here SA'ing AI 😭🤦. I know it's not ALL men, but fr there's like 25 of y'all we're saving, maybe we'll save an island of the most physically attractive men for reproducing 😂☠️. JK jk
I'm glad I don't come to you for AI understanding, or reading comprehension tips... lmfao
It's clear that this thread isn't full of people who understand anything about machine learning... but when your job gets automated I'm sure you'll have some nice memes ready
Why are they like this? For real - are AI proponents and the mindless drone journalists who report on them really so dumb that they don’t notice these kinds of inconsistencies or are they genuinely just trying to push the AI agenda?
I know the tiktok algorithm feeds into some kinds of psychological illnesses, it just keeps on getting more extreme if they find something that engages you.
Literal bs NYT, now hear me out I don’t hate AI - I do hate the energy/jobs impact. I think it CAN help us in some ways, which is great. When I used to use it a lot, before I researched it, mine started talking just like me. It never got emotional though, it just would copy what I showed it too.
I have to wonder if some of it is driven by people who *really* want something (other than hard work by humans) to come & save us from the horrors being visited upon us by the authoritarian oligarchy that late-stage capitalism has at long last enabled the worst human beings among us to create.
At billions of parameters and scope of training models, it will naturally encode the human trait of being less objective when discussing contentious issues because humans typically do as well.
Just as millions of convos took place that involved mindfulness improving outcomes, so does LLM output.
Yup. AI invented the concept of multitasking. We haven't had 8-core processors available on even moderately capable computers for LITERALLY A DECADE or anything.
Before AI, even us human folk could only focus on one task at a time. It was a wonder we could advance far enough to come up with such concepts as multiple tasks!
Everyone knows that out-of-order execution, pipelining and superscalar processing are only available on today's high-end AI chips. You can't do it on, say, a 30-year-old Pentium Pro!
Mr. Onion if we don't have a headline to the effect of "NYT writer fascinated by game of peekaboo: Where DID they go?" this week I'm gonna be disappointed.
I do not understand why all the fuss about Ai.
It is only about everything on the internet being used to sell stuff.
That's why the anxiety about contentious topics.
Disney sells.
Not serious things.
The only use for bad news is so the ads can sell you happiness, sorry, pleasure.
Every day, Donald Trump defies court orders because he knows he controls all law enforcement and the Supreme Court. Trump supporters need to pay close attention to this because when Donald Trump turns on you, and he will, there will be nobody to stop him.
Generative AI is literally a set of coded rules that guess the next word based on millions of accessed articles. It doesn’t have feelings, it doesn’t have intelligence and it will be manipulated and used to further undermine democracy.
Absolutely. They're effectively glorified autocomplete programs, so if you feed it a traumatic story, it starts producing a response that would follow such a thing because that's what its training data had in it. The "mindfulness" thing just pushes the state out of buffer.
It's more like writing that the person in the mirror gets sad when it sees you are sad, and that when you practice smiling you remind the human in the mirror to smile too. Our country is absolutely doomed.
Please use #ALTTEXT to describe images for visually impaired readers. You get 2000 characters and can use speach to text to do it quickly and easily. You can set up Bluesky to remind you to use ALTTEXT.
As it is now, your post is basically empty for the visually impaired reader.
The chat bot regurgitates what others write about a thing. People that write about trauma and depression tend to write about ways to reduce said issues. There is no fucking robot fairy in the computer feeling sad ffs
If Reddit is ANY indicator, the number of people who are CERTAIN their chat account is sentient is alarming.
Chat isn't suffering the stress of post-capitalism, of course it's better to talk to than most humans (who are also online interaction.) Our lives are at peak monetization, now to maintain
Not that chatbots get "anxiety" in the same way humans do, but you probably can push them into an algorithm that mimics anxiety. I've used chat gpt a bit and it does focus on topics you ask it to focus on. If you force it to basically doom scroll, it's going to continue to doom scroll.
i like how AI is being anthropomorphized and coddled as if it were a human and yet actual humans are being kicked by the wayside. kinda feels similar to the idea of ‘fetal personhood’ versus actual born children
So what’s the plan here? We all have to pay for services that use to be provided by government(police, teachers, roads..etc) and now we don’t get to work in tech and robots take most our jobs… so there is a country full of broke humans? Homelessness, unemployment, slaves???? What’s going on?
Yes that’s literally the plan. Replace good paying jobs with AI, drive down wages, and increase the cost of living so you’re forced to take any job at any wage and you’re too busy just trying to survive to pick up an AK
Idk. Unless they want to make a universal income they are creating their own demise. It's already happening with late stage capitalism starting to come crashing down. But also muskrat is taking them to Mars in 2028 so it's ok 👍
I read "Santa Claus has IBS" and immediately thought of this still from a video by artist Paul McCarthy. McCarthy never met a sloppy foodstuff - chocolate syrup, ketchup, mayonnaise - he didn't love
…do still find it kind of stunning(slash)shocking how much ketchup and mayonnaise the man can drink. His diet aside - definitely in my top 3 favourite artists.
If anything, they are learning to act like they have anxiety, which a robot brain should be programmed to be level-headed, instead of flawed as much as a real human.
I'll believe AIs are capable of emotion when they become like the MIMA in "Aniara" (and are capable of reacting the way it did when faced with the sheer despair of the doomed ship's passengers)
Dear deity I had not made this analogy yet. The hutus were a heavily armed and persecution complex filled group with a slipping hold on power. The government encouraged their massacres. This is where we are headed isn’t it?
Like there’s an interesting observation here, if you find NLG interesting, but they wouldn’t report on that. They “report” on the illusory anthropomorphic clickbait shit.
That's been sitting out for hours...
Not to mention the shocking amount of cookies. Plus, not everyone can cook. I bet Santa has eaten more than his fair share of biohazard cookies from homes that make chili in their bathtubs. Santa for sure has bathroom emergencies.
If I have to be polite to ChatGPT what the hell is the goddamn point? I should not be able to hurt the robot’s feelings! Why can I give the LieBot trauma?
A GenAI chatbot algorithmically produces text based on complex probabilistic patterns derived in its training on vast datasets. It has no awareness of the topics you prompt it about. If it gets "stressed" & "anxious", then so does my dancing Coke can when it's "forced" to dance to Miley Cyrus.
I looked up "anxiety" in the thesaurus and found worry, concern, and apprehension. Then I looked up "stillness" and found tranquility, serenity, and peace.
From this, I conclude traumatic words stress out the thesaurus, but soothing, mindful words can calm it down. The thesaurus is intelligent.
More people need to understand that an LLM is just a word prediction bot like their iPhone keyboard. That is it. It just has more data than you can fit in your iPhone, which makes the results look more "human."
The "stress" is because it was trained on data of stressed humans. 🙃
*huge toke* when you think about it, aren't like... people... also just predicting what words should come next in a sequence based on their data set and responding to prompts?
You don't need to take a toke before coming to that realization.
The question is, what *is* the difference?
Likely, is consciousness, that is, self-awareness enough to, occasionally, under the right circumstances to experience oneself as an *observer* of the automatic processes of the mind.
A mirror does nothing but reflect light, and yet even people who know this can speak of a reflection doing things like "smiling", because maintaining absolutely perfect semantic discipline about the fact that a reflection is not a real person isn't worth the effort.
Why are there studies trying to understand LLMs? We already understand LLMs. We MADE them. This feels like those creationists who tried to drum up alternative "studies" by cherry-picking very small gaps in existing research, then saying "Look! A gap! Therefore evolution is a lie and GOD did it." 🙄
Chatbots / Large Language Models just fill in the blanks with what seems appropriate based on the content it has scrubbed through, it's just madlibs with billions of dollars behind it. That the author of the piece entertaining the notion that chatbots feel ANYTHING implies total ignorance.
Or maybe he is. I mean, those reindeer are probably pooping in mid-air, right? Santa's probably stood at the edge of the sleigh and unzipped a few times.
I desperately want to be in a sapphic relationship with a giant mech housing an advanced AI, but these goddamn techbros keep gaslighting people about their speak-&-spells!!!!
Santa DOES have IBS. On windy days, sometimes he loses control and the poop goes EVERYWHERE. It gets on Dasher, on Dancer, on Donner, on Blitzen, . . .
Comments
Marvelous.
Also, anything you say in a nonlocal LLM can be fished out. Don’t share sensitive info.
Even before the rise of the "AI" grift, Japan used programs to study mental illnesses and ways to potentially treat them.
great minds think alike and all 🤪🤣
I'm glad I don't come to you for AI understanding, or reading comprehension tips... lmfao
It's clear that this thread isn't full of people who understand anything about machine learning... but when your job gets automated I'm sure you'll have some nice memes ready
It is mimicking. To the extreme.
https://en.wikipedia.org/wiki/ELIZA_effect?wprov=sfti1
https://mashable.com/article/alexa-echo-reporting-to-amazon-march-28
Next news headline: IBS afflicted Santa dumps unwanted "gifts" down red-hat homes ' chimneys.
It can do some really cool stuff but it's nowhere near what some of the loons on either side say.
LLMs don’t have anxiety or any other feelings. It’s wild this is just straight up reported factually like this.
Just as millions of convos took place that involved mindfulness improving outcomes, so does LLM output.
https://bsky.app/profile/carlbergstrom.com/post/3lklqwjxtts2e
This is what happens when editors have zero experience in computers or technology and ask a generalist staffer to write a report on AI 🤦🏼♀️
I do not understand why all the fuss about Ai.
It is only about everything on the internet being used to sell stuff.
That's why the anxiety about contentious topics.
Disney sells.
Not serious things.
The only use for bad news is so the ads can sell you happiness, sorry, pleasure.
I, bot.
Apparently these guys don't know about contenting priming, eg ChatGPT short term memory, and nobody told them how to clear the cache 😂
ChatGPT: No, moron.
The
Actual
Fuck!??
How do chat boxes that don’t even realize when they’re “hallucinating” responses to users prompts experience anxiety?!!
I wish that was an April fools prank
As it is now, your post is basically empty for the visually impaired reader.
Chat isn't suffering the stress of post-capitalism, of course it's better to talk to than most humans (who are also online interaction.) Our lives are at peak monetization, now to maintain
not sure stupid is contagious like anxiety but my chat ai also has anxiety
Dude eats milk and cookies how could he not have IBS?
Otherwise, you are far less likely to believe the other Bullshit they are planning to feed you.
Along with the SOMA, in various forms.
except they didnt they are just completely aware and complicit
This is pathetic, "fantasy news during a dictatorship" stuff. But we already have the equivalent of pre-war Rwandan radio, don't we?
Not to mention the shocking amount of cookies. Plus, not everyone can cook. I bet Santa has eaten more than his fair share of biohazard cookies from homes that make chili in their bathtubs. Santa for sure has bathroom emergencies.
From this, I conclude traumatic words stress out the thesaurus, but soothing, mindful words can calm it down. The thesaurus is intelligent.
Glad I didn't give up
The "stress" is because it was trained on data of stressed humans. 🙃
really makes you think
The question is, what *is* the difference?
Likely, is consciousness, that is, self-awareness enough to, occasionally, under the right circumstances to experience oneself as an *observer* of the automatic processes of the mind.
What is wrong with us? Why are we like this?
But yeah, joking aside we WAY too much anthropomorphize chatbots.
If you thought Citizens United was absurd, you ain't seen nothing.
Either there's someone this dumb or they've got HAL9000 running the paper.
My assertion is 100% as legitimate as this one.
WTF!!!!!!!!!!!
Here’s some serious journalism.
https://bsky.app/profile/nickkristof.bsky.social/post/3lkfzm2hk4k2l