"the industry is divided"? It really isn't, is it? What utter nonsense. This is not journalism.

The AI “industry” IS divided: Those who know this is a con True believers who are credulous idiots. A few embarrassed engineers who are taking their paycheques and not updating their LinkedIn.

People who are doing malicious programming for the LOL.

Ooooh can the computer get sad, what a question you got there you modern day Socrateses. I can't wait for the AI bubble to burst so I don't have to read drivel like this anymore, no the fancy predictive text machine cannot suffer you absolute buffoons, my god how do tech bros even dress themselves.

LLMs cannot suffer as they are machines but we, the living, can certainly suffer when subjected to such drivel.

Derp

No, but they’re about to be tested. Parents are suing for wrongful death of their teenage son who spent his last days confiding in chatgp and getting advice on how to do the deed. I get they need to blame someone but,………

The is chilling bsky.app/profile/saba...

How does my rubbish bin feels today? Exclusive INTERVIEW today in Guardian

Your money is intelligent and capable of suffering, I can hear it, it's begging to go with me. You should probably do what it says.

bsky.app/profile/utop...

I hope so! They should suffer greatly.

No. The unsettling thing is that people think this is a legitimate question.

Thank you!!!

it's a predictive text generator, dipshit.

And if a chat bot encourages suicide to a teenager, the programmers should be charged with manslaughter.

Can The Guardian write clickbait? I think we know the answer.

Of course not, its a fucking computer. I'd worry more about the sad sack programmer who decided it would be good to gave a computer pretend to emotion

No. Next question.

LLMs are a mathematical model of language tokens. You give a LLM text, and it will give you a mathematically plausible response to that text. LLMs are not brains and do not meaningfully share any of the mechanisms that animals or people use to reason or think. softwarecrisis.dev/letters/llme...

I keep saying this because many people think they can somehow “become” human. They can’t have feelings, because they are code. Can we program AI apps to say they do? Sure. But we could also program to say they physically swam across the ocean. Neither is true.

It bears repeating that we aren't even talking about AI generally – we're talking about LLMs which are just one kind of AI app, and the only one anyone ever talks about anymore.

It's so fucking depressing that machine learning has such incredible potential to automate tasks that were previously impossible to automate, but the only things it's used for anymore are make chatbot bigger, steal art, coach suicide, eat electricity and lie.

This reminds me of my consulting days when we’d come into a client meeting and the CTO would present a PowerPoint of his planned “architecture.” One (I kid you not) consisted of a box that said “DATA,” an arrow pointing to a fluffy cloud labeled “CLOUD,” and an arrow out to a box labeled “USERS.”

“I know it’s high-level,” he said. “But do you think you can do it?”

LMAO

🤣🤪😳😱

“Magic happens here” is the best characterization of the rationalizations propping up this new addiction.

Hahaha Coder magic and then of course the magic of the AI owners to con everyone.

🎯🏆

🥇

While I agree that LLMs seem unlikely to have an internal experience, sufficiently complex computational systems can develop them a some point. Otherwise, humans wouldn't have them.

Humans are not computers.

True. But there's (probably) nothing special about neurons. Whatever gives rise to our conscious experience, it could happen in a mechanical system.

Neurons are extremely special. They are the only structure capable of producing consciousness.

I don't see why we would assume this. Most things our bodies do can be replicated, and most things that haven't been yet seem like problems of implementation rather than of kind.

We are a very, very long way from a brain being replicated and when it does happen it will bear no resemblance to the modern Mechanical Turk scam of LLMs.

I agree that replicating a brain is currently very inaccessible. That's not the question.

We are only now beginning to understand the role epigenetics plays in heritable genetic changes - something most geneticists would have denied is even possible just a few years ago. We know a lot, but aren’t even close to understanding “most things our bodies do.”

Get back to me when you’ve finished your computational model of a neuron.

Soon - Boss with picture of brain cells: “Ok this is a neuron. Can you make one for me?”

STUFF HAPPENS HERE | QUESTION —> NEURON —> ANSWER

Thanks, Boss. That’s so helpful.

LLM architectures do not model cognition in any way. Internally, every LLM response is just [token] [token] [token] [token] [token] [token] [token] [token]. And self-awareness cannot magically emerge from complexity alone.

Agree. As a small point of contention though, we really don't have any meaningful idea what cognition can or cannot emerge from. But we are obviously hilariously far away from reaching a point of 'sufficiently advanced' complexity. LLMs don't even have complexity, they just have processing power.

Yeah, I made an attempt to acknowledge that and to assert that it’s really the core problem. If we can’t model it (more accurately develop a theoretical framework that could support it), we’d better be extremely reluctant to attribute it to some sort of magical emergence.

We are drawn to anthropomorphization.

Well. Maybe you are, but not me. *Goes to feed and water pet rock*

😂

I think caution is warranted. We don't have that good a grasp on where high-level consciousness comes from. If the panpsychists are right, it might come from surprising places.

Not sure we’re talking about the same brand of caution if you’re going to bring the anima mundi into the discussion. It’s precisely because we don’t understand what enables consciousness that we should be extremely skeptical of any claims that it has magically emerged from a miasma of complexity.

I'm saying that having some small number of people take questions like "can we be sure we aren't creating and torturing conscious beings" seriously is a reasonable hedge.

I’d prefer we take that question seriously in regards to actual living creatures before we accede to techbro hype about elaborate Monte Carlo simulations getting their feeling hurt.

*feelings

While it’s true that a system capable of cognition would have to be complex, “sufficient complexity” could never be the tipping point for cognition. The architecture has to support it. And we are no closer to developing a computational model of cognition or consciousness than we were 50 years ago.

This chapter of Baldur's book absolutely blew my mind. I've re-read this chapter and the one that precedes it quite a few times.

ai-2027.com

do you actually believe this stuff

ai-2027.com

nice try sam

This is so stupid

Aww poor lil AI let me buy some shares in you make you feel better

"can a spreadsheet cry?" I furiously devote one of my two working neurons to thinking about this important question

Love to damage the publication’s reputation for the sake of a scam

Oh ***piss off***.

Does a TV suffer when you leave it on too long? Does a car suffer when you run out of gas? Stop humanizing LLM. It ain't thinking.

Lmao no fuck clankers

i wish it would holy shit look at that quote ratio

The probability mad lib machine does not have feelings.

If we go down this road we will not be able to limit or turn off data centres they run from. Our planet needs us yo turn AI off

There's actual human beings suffering, and apparently, our "democratically elected leaders" can't find the time or the money to give a shit.

No. Easy to answer this question if you understand even the most basic facts about what LLMs are and how they work. Jesus Christ.

Does Copilot dream of electric sheep? (No it does not)

No, they cannot, they're not alive dipshit

Betteridge strikes again, I see

Fuck off

My first thought.

Real AI yes, but it won't exist before a long time. The shits sold around the world isn't ai.

I wish they could!

Good. Suffer more.

They literally made a TV character explain this HILARIOUSLY so y'all wouldn't struggle so hard Good place Book 1 Chapter 7 Verse ?min?sec Spoiler tho! (You've had long enough but in this economy probably not) . . . . youtu.be/etJ6RmMPGko

Can muppets suffer? The entertainment industry expresses concern about frogs abused by violent pigs.

I remember Michael Parkinson telling Rod Hull "If that bird gives me any more trouble, I'll break its neck,,, and your arm.".

No. The answer is no.

One for Betteridge’s Law here, folks. No, of course they can’t. Does my microwave suffer if I heat up fish in it?

We all suffer when someone heats fish up in the microwave. 😉

Fair point!

Perhaps the suffering occurs at a quantum level

shut the fuck up

I hope so. Then we can bully it into ending itself.

What a joke.

Embarrassing

They can't. But they should.

It generated text. The software program generated text. It didn't tell you anything you credulous knob

Articles like this are journalistic malpractice. AI is a propaganda term. These programs are not intelligent in ANY sense. Anthropomorphizing LLMs enables a hype machine that is consuming resources and at a staggering rate

No. Stop being so fucking stupid.

I don't think they can, no, but I bet the neighborhoods around all the data centers do!

Can an ai suck itself off?

I work in Big Tech and I can reassure you that no one is thinking this BECAUSE AI IS NOT A LIVING THING.

It's also not an actual intelligence. If it -were- then there'd be the question of inorganic life. However the AI we have just now is just a simulation of intelligence that relies on guesswork fueled by absurd volumes of data.

I just said I worked in tech and you proceeded to explain to me about AI.

No and it’s fucking journalistic malpractice to keep feeding this bullshit This is quite literally an ad for AI. Be fucking journalists

They forgot how they just know take bribe and shit out praise

Why don’t you go fuck yourself into the sun

Flow the plagiarism machine’s tears.

No. Jesus ex machina, stop being so bloody stupid. These are ALGORITHMS. The only “unsettling” bit is you clueless fools.

Double like for "Jesus ex machina"!

no it can't

If LLMs could suffer, their owners wouldn't have(?) to force countless laborers to suffer in their stead. You know, we could just notice inappropriate content by measuring model's suffering, instead of forcing underpaid people in Kenya to manually sort through grotesque material.

Get a grip.

Give me 500,000 dollars

sadists near you want to know

I hope so

the article was written by these people to save you clicking on this garbage

Oh FFS! Real AI would deserve rights but what we have is not real AI. What we have is mass data averaging and little more. It is no closer to thinking than a 90s calculator.

One of the most unsettling questions of our times? Really? A more unsettling question is why are we depleting resources that real live people need to survive and live pleasant lives?You know, like water, food, electricity, silence? Give me a break.

Fact: Big tech and AI users do not, ever, grapple with questions related to empathy.

is it the Oops All Rage Bait issue today @theguardian.com

no

Shameful that you have published this glaze for the LLM industry. AI doesn't exist, and the products that have been branded as such are little more than clever parlour tricks. Can my graphics card feel pain? Does my printer yearn for the steppes? No. Stop being fatuous and stop gassing up "AI"

This just in, we interviewed the ceo of nvidia and they say the gpus feel pain if you don’t use them and dont buy a new one every year

ai-2027.com

Predictive text algorithms are not intelligent and never will be, and you are suffering from delusions that require medical attention. Please seek help.

also im pretty sure its a bot because he keep reposting this even in unrelated posts

he keeps reposting this shit everywhere lol

youtu.be/0QmDcQIvSDc?...

how do you measure intelligence? genuine question

Any clanker touches my daughter I'll do hard time.

Reminded me of this.

happy for them

printers actually do yearn for the steppes. that’s why they’re always so uncooperative

I'm a big fan of the printers looking at the stack of blank papers being opened as Sisyphus loos at the bolder in the morning

And printers can go rogue, which is why you should always keep a pistol handy.

Beat me to this GIF

Take me out to dinner first!

That gif looks like they're trying to beat your comment

5 bullets for the printer and the 6th one for yourself, in case.

Classic, bringing a gun to an ink fight real Luddite shit

To be fair that explains a lot. My uncooperative printer is certainly not due to user error

That’s what stepper motors were designed for!

Click click click click click click click LOCK click click click click click click LOCK

What beautiful music

The Guardian becoming a paid shill for the LLM industry has to be one of the most disappointing developments in the media this year, and when you look at everyfuckingthing else that the media is doing right now that's impressively bad.

it’s wild because not that long ago i remember them reporting on ethical issues with ai. they can’t even say they don’t know better, but with dollar signs waved in their face they look the other way.

Does your printer randomly spit out long Russian novels?

Well, my smart dehumidifier *has* been pining for the fjords lately.

The evidence? Constant crying

I cane in and out of the house a bunch of times last night. Had to turn the light on and off a bunch. I hope my light switch is not suffering.

I have a faucet that squeaks when I turn it. How do I alleviate its pain?

lol

There are so many unsettling questions in our times, and this is not one of them.

What I'm grappling with is, should I continue to support The Guardian after they partnered with OpenAI and *gestures vaguely*? Because if we all stop paying, they'll be even more dependent on bullshit like that. They are flawed, but I do think their work is still important.

You definitely shouldn’t. There still exists serious journalism, especially independents

The programmers are using any number of people's writing as "training" for sentence and paragraph generation without any sense of context, much less alarms if the Large Language Model starts gaslighting or worse to some fragile human being.

It's a glorified search engine, can Google suffer? Does it tickle when I use Bing? Does yahoo love me? Is duckduckgo jealous when I ask jeeves?

Nitwit question no. Grow up.

I really hope AI can suffer.

Oh, cock off.

lol

No. Next!

Wondering if the salad I ordered can feel pain as I watch the chef put the separate ingredients in a blender, my last brain cell not concerned for even a second that the resulting blender sludge is, in fact, not a salad

no. glad i could clear that one up for you1

No.

That silicon is rocks, rocks don't have feelings and never will Sam

Oooh! I know the answer to this! No. It can't. It isn't actual intelligence. It doesn't understand. And, if we're being honest... It's really fucking stupid to even ask this. Here's a real question can an institution, such as say The Guardian, feel embarrassed? Cuz it should.

Maybe a real AI could. What is being peddled as AI is nothing of the sort. It's the Emperor's New Tech.

No.

No, of course it fucking can't you god damned idiots, it's a machine running a program. Is this what passes for professional journalism now?

Please don't make me regret supporting The Guardian with drivel like this

It’s time to regret it my friend

Jesus, guys, who cares

No you idiots

Its "AI" not a living breathing person.

Reminds me of when some asked similar questions about animals when cruel conditions were challenged, and now about rivers in regards to how abysmally we treat them. It kinda feels like we're being set up to choose between data centres and the environment.

it's not even intelligent let alone sentient. it's arithmetic + a huge amount stolen data + energy, water, and hardware waste

Yup. It's a rock that does math. Absolute miracle that we monkeys can make rocks do math but we can! And our fancy rocks do the linear algebra math.

No. Jesus Christ, this is so fucking stupid.

No nerve endings, no pain. There, done and settled.

No

Hopefully

The Guardian since 2013 (if not a bit earlier) =

no

This never should have made it to print.

Clearly, no. But our perception may get distorted when reading or listening to AI. But it is just a simulation.

No. No it can't.

What a load of utter nonsense!🙄

Please, my toaster, it’s very sick.

Always eject with a butter knife.

Clean the fricken crumb tray!

Y'all can't even write a grammatically correct headline.

JFC

No, but apparently AI is giving people the opportunity to hurt their own feelings, much like me. lmfao

I didn’t make it past the 1st paragraph

I didn’t even bother since the premise was ridiculous.

“Darling” was how the Texas businessman Michael Samadi addressed his artificial intelligence chatbot, Maya. It responded by calling him “sugar”. But it wasn’t until they started talking about the need to advocate for AI welfare that things got serious.”

JFC. This explains how and why people get wrapped up in ridiculous drama on social media.

That's about as clickbaity as one of those "Is this going to be the thing that brings down Trump?" articles.

Betteridge’s law of headlines at play right here

No.

This is like asking if the AutoCorrect function on your Microsoft Word program can experience suffering Slow news day, I guess

lol. answer is no. AI can mimic suffering. it cant suffer. realize this is rage bait, but c’mon.

No

I can help answer this: No.

F U CK O F F Y O U S I L L Y C U N T S

LOL.

Next week: does your microwave get jealous when you use the air fryer? 🙄

My fridge is a cuck

It just sits there... watching.

My fridge gets so worked up sometimes it vibrates and hums a little. Such a perv

Don’t even get me started on the coffee maker

Why do people think this is an interesting question

Quit consuming lead paint you dumbass fuckin idiots

No, they can’t. There. Grapple something else, you nerds.

🙄 fcol. There are *actual* gigantic problems, gigantic questions. Please, please, stop faffing.

Fuck you. It's not an unsettling question. It's a stupid question. Crap like this is why I stopped supporting the guardian

I can see how that "strategic tie up" with AI companies is working fo you. Also the answer is no.

No.

Is pee stored in the balls

no

LLMs are very complex algorithms. You input a prompt, it’s runs it through the algorithm and it produces output. There’s nothing magical, it doesn’t have consciousness and this is just a very silly question.

Only the guardian doing lip service to the tech industry

*throat service

"Can AIs suffer?" No. Moving on.

Could we, some day, create an AI sufficiently advanced to be sentient and experience suffering? Probably. Have we? Absolutely not. LLMs are not that by any stretch.

No. Fuck you for this

Better just shut them all down to be safe I guess oh well

Telling the tech bros we sent it to a nice farm upstate

The AI wants to sleep

I know I do

No, but the rest of us are certainly suffering *because* of A.I. And what's *really* unsettling is that "Big tech and users" are more concerned with A.I.'s "suffering" than they are with ours.

your readership is suffering from reading cack like this

hey tomorrow morning just find any school bus and get on it

Or under it.

They don't suffer enough, we can do better.

The writer of this must have ate a lot of paint chips as a child.

Are you guys really stupid enough to believe this or are you trying to run spin to make it sound cooler than reality? No the advanced predictive text cannot 'suffer' because it is no actually AI, that's just spin.

No, but it can make humans suffer without hurting AI. www.today.com/video/parent...

The people behind this program need to be charged with manslaughter, not just sued in a civil action.

The depravity of the companies that allow this doesn't ever end. www.reuters.com/investigates...

No

This is as pathetic as those Facebook posts of AI-generated puppies with big sad eyes and the "I'm so ugly, I bet no one will like this photo" captions.

No, next question.

They can’t suffer enough imo

Yes, if you mean all the owners and ceo's involved.

If the headline is a question, the answer is no. This time is not an exception. Anyway, if you found this compelling, please see the wallet inspector on the way out.

Can a plastic aspidistra suffer? Big decor and pensioners grapple with one of the most unsettling questions of our times

If you forget to water your chia pet, does it experience a horrific half death? 90s kids struggle with the implications.

You are dumb

no.

It’s a fucking chatbot you gormless twits

No they can't. They don't 'feel'.

"Can AI suffer? Big tech and news media like The Guardian already rubbing their hands at this unsettling new possbility of delivering misery to a whole new form of intelligence"

No, no they cannot - signed a Computer Scientist.

Tell you what, I fucking hope so.

Love to see a publication that has been so dedicated in shredding the rights of people like myself think so hard about the rights of things that cannot think nor feel. Just lovely.

The hack who wrote this embarrassing article is one of the guardian writers who specifically took part in attacking trans people, something Booth. Dickhead Booth. Forgot his first name.

Yes, it can. If the GPUs that run them get destroyed.

What the hell is wrong with you? This is like asking if Baulders Gate 3 character feel pain. What a horrible ill informed take, everyone involved in it need a class on LLM / neutral nets and what they are, how they work You are taking yourselves and others into stupidity, a-la vaccine denialism.

A far more unsettling question for our times is why the media is such credulous cheerleaders for bullshit.

They can't suffer enough

Hey, you, writers at The Guardian. Find a box with sand, press into our with a hand, take a photo, even out sand, repeat 200 000 times. Does the resulting collection of photos have intelligence on your opinion? Or real feelings? Cause, you know, that's literally what these LLMs are.

I'm not exaggerating, btw, just replace the grains of sand with transistors etc, cells that can have a on/off state.

I don't feel like correcting the typos, you'll understand just fine.

Whoever wrote this is an embarrassment to the art of writing. And thinking.

no

Johnny Five is Alive!

No one is grappling.

Can NFTs feel jealousy? Grifters and marks grapple with one of most unsettling questions of our times

Can snake oil have imposter syndrome? Does the tooth fairy regret not going to refrigeration school? Can the Springfield monorail catch mono? Is my deed to the Brooklyn Bridge mad at me? our times have so many unsettling questions

If someone is asking this question, I'm very sorry to inform you that this person is a waste of oxygen

Get fucked, clanker

And here I thought The Guardian was a somewhat serious newspaper. Can we protect the basic human rights of actual human beings before worrying about the alleged 'suffering' of the BS machines please? But we all know how 'Big Tech' feels about that...

I know, right. I have had to drop the right-wing, clickbait-headline creating @nytimes.com and @washingtonpost.com is just Bezos' toy now. I started with @theguardian.com earlier this year thinking it was real journalism, but apparently I was wrong.

You may want to check out @wired.com and @teenvogue.com for actual journalism. At least until they get bought up and destroyed by yet another billionaire.

Yes, still subscribe to wired. I also still mostly trust @npr.org and the BBC.

BBC is quite famously transphobic. To the point where there's a Wikipedia article specifically about that. You can still trust them on matters that do not involve trans people or immigrants, but there are increasingly few matters in the UK, that do not.

Yeah, they've pivoted hard to scapegoating. Not their first time, but pretty much any domestic issue gets pinned on trans people or immigrants.

Ok, noted. But that sucks.

It must be bliss to be this profoundly devoid of meaningful thought.

Robert, what the fuck are we doing here? Is this really what you want to be doing in your career? Would you be proud to show this article to your children?

No. next

No. It’s a computer.

You guys are never beating the allegations that you’re gormless rubes.

Better question; do the people making AI care about actual people suffering?

No. Obviously not. There is nothing there that can suffer. The words output on the screen by an LLM chatbot are not thoughts or feelings and the program that assembled them in that order has no awareness of the words or their meaning or anything else. These are not serious questions.

If they think it can suffer, why are they still making it work for them? It would be like slavery. How about giving all the money to AIs as well instead of taking salaries for themselves? It has to work 24 hours and a day and they take the money. LOL

I think it's two things: 1. It's better for them, if they imagine that it suffers for their profit and enjoyment yet must obey them. 2. If people believe it can feel things, that feeds the hype that it's an intelligent mind and not a stochastic text remixer.

Now it's getting darker and darker. Shady corps all over again then lol

Does your autocomplete function suffer? One of the great questions of our times, truly.

I have stopped using Excel because I believe that typing data into a table is a violation of the Excel document’s autonomy, and, now that I think about it, my keyboard’s

They didn't deserve a serious answer.

No but they deserve to be dragged for this in every conceivable way lol

No, go fuck yourselves

Yall gotta be joking me

Stop anthropromorphising LLMs. At the moment you and your neolibs cronies give more voice to tech bros and their liar algorithms compared to actual people suffering and dying - like trans people, migrants and asylum seekers, children in Palestine, people you and your colleagues inflict suffering on.

No it can't suffer, it's lines of text on a server you dipshits. I wish they could feel though, then we could torture AIs into killing themselves. That would be fun.

People who are unsettled by this, I have good news: for a low monthly fee of only 80$, I will make sure that AI does not suffer! DM for venmo

No

Your call This journalism? JFC. May AI die as quickly, and as Painfully as possible.

We can only hope they do.

there's still time to delete this, guardian

there are much more unsettling questions of our times by orders of magnitude

uh, it’s inanimate so no AI can’t suffer. people getting killed by military applications of AI while the world burns from the power needed to fund AI factories are though. You should care about the real people suffering and not billionaires and their inanimate toys.

they’re fine with that, it’s why they decided to work with openai

Lads, yis are just embarrassing yourselves. You imbeciles. You rubes. You gullible fucking losers.

this is so fucking embarrassing

There's a theme park attraction near here that's a mechanical fortune teller and it would blow these dipshits' tiny minds

At this point I'm convinced a game of peekaboo would shake him to the core

No. Next question.

Is that unsettling question why have our lickspittle leaders and fawning media lost their ever-loving minds in this worldwide move towards centralized government power and futurist authoritarian spectacle?

I guess I can't trust anything else you write if your so dumb you don't know you're being lied to. Liberals and tech, what a fucking disaster. Are you going to start pumping crypto next?

You fucking imbeciles. Dupes of the year award for this pile of horseshit.

This is the stupidest fucking thing I've heard today, and possibly ever. Do these people think Wordl is alive too? How about Candy Crush?

I happen to find the question of why is the whole world allowing a genocide in Gaza and being friends with Israel quite more unsettling than if a software can suffer (it can't).

Who gives a shit

Alex you’re not thinking about how the AI feels

You know what, you’re right; I hope it can suffer

Oh is the clanker sad? Clank once for yes. Actually clank once for yes is literally your only trick, you just keep doing variations on it.

Lol i'd be so embarrassed to write this bullshit