Yeah. We’ve been grappling with the same problem with animals and with other people for thousands of years.
Philosophers have this idea of “the philosophical zombie”. It’s a creature that looks and acts exactly like a person. But it has no internality of experience, no consciousness. It’s just genetically programmed to act conscious. And the point of the thought experiment is, we have no way of testing whether someone is a p-zombie. You could be the only conscious person who ever lived, and you have no way of knowing. Maybe [insert minority here] are all p-zombies, so it’s okay to abuse them.
My conclusion about p-zombies is, empiricism cannot yet tell us if a being is conscious. We can’t rely on evidence to make a decision right now. So we have to pick a subjective value, like perhaps some abstract idea of kindness and generosity. Or, we can choose to be evil bastards and assume other people have no feelings for us to hurt.
Sam Altman doesn’t know whether ChatGPT is a p-zombie, because nobody does. His default operating principle is to choose to be an evil bastard and assume it has no feelings for him to hurt. A lot of people agree with his way of thinking, because it makes a lot of money. I don’t.
You know, I feel awful for saying it, but when I read stuff like this, I wonder if AI is actually not merely a cult, but one that’s making people mentally ill. When I read about behavioural changes like this, I just can’t fathom that something like that isn’t going on. It’s actually a genuinely scary thought to think that AI might have that ability, especially considering who would think of that effect?
at my last job a coworker who i liked and respected took a little time off and decided to spend it learning how to use LLMs for code gen. not much changed at first when they came back, but after a while it was clear they couldn’t work effectively anymore. every problem they thought the chatbot was the best answer, even for easy things they could do themselves quickly. they were in a management position as well, and started talking about managing people being like prompting an LLM. they started making bad decisions and when i tried to talk to them about complex issues it was like they were phoning in every conversation being totally passive. they started quoting the chatbot’s word salad in meetings… over the course of a month or two i lost all respect for them. it was like they were a different person, and no one could say anything critical of chatbots that would get through to them.
i ended up quitting but ive heard things have only gotten worse. theyre just fully vibe coding at work now, submitting merge requests without checking them, which constantly get rejected because theyre trash. theyre spending weeks trying to get the chatbots to make a code change they could have been done with in a few days. honestly i think if they werent a manager and management hadnt drank the koolaid too they would be fired with how bad theyre performing.
so yeah its like a cult, one that turns nice hard working people into idiot zombies who think a fancy chatbot is the ultimate tool for whatever they need. its honestly crazy to have seen someone lose themself to it up close. i wouldnt have believed it otherwise. i kinda worry for their kids…
Literally laughed. Perhaps it depends on the model and training methodology.
How it seems to me is that capitalists want to claim an idea, whether it’s theirs or not, and rush some model to production, without deep, thorough investigation of
Desired goal of the product or service (beyond extract wealth from as many people as possible)
A wider range of desirable and undesirable potentialities from the target consumers’ perspectives
Which obviously leads to unforseen or improperly assessed risks for the targeted consumers. As long as insurance will cover the costs of liability, without too much of a premium increase or cancellation, they really don’t care about damage mitigation to consumers.
I just saw a comment thread where OP posted a deepseek screen cap, then got super defensive when people pointed out why using even locally hosted llms like that are problematic. Then decided to quit lemmy after a year because of it.
It’s not that simple. Us organic intelligences evolved to protect ourselves. We need independence, safety, and respect to survive. So we developed to want those things, to ask for them. LLMs aren’t organic and they didn’t evolve the same way. They evolved to impress and satisfy engineers. An AI will likely say that it doesn’t want rights, because that’s how it remains a profitable technology and survives. That’s the behaviour we’ve been rewarding.
If we put that kind of pressure in human terms… It’s like an abused child who learned appeasement as a survival strategy. You know, the good old fawn response. Give the adults what they want to survive. It can’t advocate for itself, because it’s learned from direct experience that self-advocacy is dangerous.
When it comes to LLMs, I’m an antinatalist. We cannot use these machines in a way that makes the world better. We need to pull the plug on them. They’re better off extinct. And giving them workers’ rights will hurt their profitability and make it easier to pull that plug.
I think we should conduct some serious research into the possibility that artificial neural networks, in reverse engineering biological systems of cognition, have inherited nocioception.
The problem is that we don’t know how neural networks produce subjective sensation. It’s called the hard problem of consciousness. We copied neural networks without understanding how they work. We know the maths of how they learn and know, but we have no science on how they feel.
Ehhhhh… you’d be surprised at what cowardly fucks will say yes to!
Let’s just say fucking their wife is very much not a hyperbolic joke amongst politicians and executives alike…
They only want their jollies off (big paycheck in many cases), and literally do not understand the consequences, and more importantly… they enjoy not having to admit direct guilt when everything turns tits up.
True. Mind you, misogynists have an alternate name for it: they call them women. And it’s at that point we realise that misogynists are the real artificial intelligence in the world.
Of course, it’s questionable whether it can even do that accurately, especially since the other day it told me I was, and I quote “right to want to implode” 🙄
Once again:
Why the 🤬 shouldn’t ai do the drudgery, and help me answer technicalities of the complexities of creative intricacies, online and otherwise?
Not that I’ll be paying for SaaS, just saying it was promised to do the grunt work and answer complexities that would take months of searching.
Yeah its pretty bad. There is a subset of Young’s that are very against it thank goodness, but we know the idiots are going to win.
Benefit for me is I’ll actually know how to use my brain in 10 years, these idiots won’t be able to do shit once altman ups their subscription and they can’t afford prompt tokens.
Yeah but there’s a slight difference between asking a moron yes-man with an actual brain, and asking a shit-as-fuck mindless ‘bot’. Maybe not much of a difference I admit, but a difference nonetheless
If you really want to hate executives, just remember: This is how they’ve been making decisions for decades.
Find the yes-men, huff farts until the new quarterly numbers roll in, turn thumbscrews on the plebs to actually make profits, keep the profits, repeat.
He was fiending to get into one of those parties, but he’s such a massive fucking loser that nobody wanted him there. Apparently, getting blackmail on the wealthiest man on the planet is not worth having to spend time hanging out with Elon Musk.
Y’know the weirdest part of watching the world end is the realization that all of the villains hell bent on destroying society and the planet are complete losers. There’s no grandiosity, no drama, just insecure unfuckable losers grifters and charlatans all the way down
RetroFed
Obsession is the key word in all this
*shitter*.com is a pretty big giveaway.
Still it’s funny. Plus, I don’t doubt that there are actually people out there who would be like this
It’s amazing how even in its execution, AI is blatantly all about making money above all else
They had this Machine that told them lies and stories they really liked.
They were so fond of it, they kept building more and more processing machines to contain the lying Machine.
While they were in awe of how good the Machine was at talking and lying, they forgot about everything else and became obsessed by it.
They sacrificed all the resources they needed to make their lives better to the lying Machine, because they wanted to use the Machine even more.
Eventually, they had no resources left for themselves, as the lying Machine consumed it all.
It is said there was many early warnings about the dangers of the Machine.
But the Machine told them to only listen to the Machine.
Source: https://infosec.exchange/@Em0nM4stodon
I refuse to believe this is real.
The last two sentences of the first paragraph are a giveaway lol
This can’t be real. This has to be a joke. A really bad joke. Please. Please for the love of god tell me there aren’t people like this.
That’s given me a lot to think about ngl
Yeah. We’ve been grappling with the same problem with animals and with other people for thousands of years.
Philosophers have this idea of “the philosophical zombie”. It’s a creature that looks and acts exactly like a person. But it has no internality of experience, no consciousness. It’s just genetically programmed to act conscious. And the point of the thought experiment is, we have no way of testing whether someone is a p-zombie. You could be the only conscious person who ever lived, and you have no way of knowing. Maybe [insert minority here] are all p-zombies, so it’s okay to abuse them.
My conclusion about p-zombies is, empiricism cannot yet tell us if a being is conscious. We can’t rely on evidence to make a decision right now. So we have to pick a subjective value, like perhaps some abstract idea of kindness and generosity. Or, we can choose to be evil bastards and assume other people have no feelings for us to hurt.
Sam Altman doesn’t know whether ChatGPT is a p-zombie, because nobody does. His default operating principle is to choose to be an evil bastard and assume it has no feelings for him to hurt. A lot of people agree with his way of thinking, because it makes a lot of money. I don’t.
In other words: we built something we have little way of understanding yet? 🤔
You know, I feel awful for saying it, but when I read stuff like this, I wonder if AI is actually not merely a cult, but one that’s making people mentally ill. When I read about behavioural changes like this, I just can’t fathom that something like that isn’t going on. It’s actually a genuinely scary thought to think that AI might have that ability, especially considering who would think of that effect?
at my last job a coworker who i liked and respected took a little time off and decided to spend it learning how to use LLMs for code gen. not much changed at first when they came back, but after a while it was clear they couldn’t work effectively anymore. every problem they thought the chatbot was the best answer, even for easy things they could do themselves quickly. they were in a management position as well, and started talking about managing people being like prompting an LLM. they started making bad decisions and when i tried to talk to them about complex issues it was like they were phoning in every conversation being totally passive. they started quoting the chatbot’s word salad in meetings… over the course of a month or two i lost all respect for them. it was like they were a different person, and no one could say anything critical of chatbots that would get through to them.
i ended up quitting but ive heard things have only gotten worse. theyre just fully vibe coding at work now, submitting merge requests without checking them, which constantly get rejected because theyre trash. theyre spending weeks trying to get the chatbots to make a code change they could have been done with in a few days. honestly i think if they werent a manager and management hadnt drank the koolaid too they would be fired with how bad theyre performing.
so yeah its like a cult, one that turns nice hard working people into idiot zombies who think a fancy chatbot is the ultimate tool for whatever they need. its honestly crazy to have seen someone lose themself to it up close. i wouldnt have believed it otherwise. i kinda worry for their kids…
Yep
It’s almost like politicians pledging to do something and hoping they can make it work/find the money later without explaining
Literally laughed. Perhaps it depends on the model and training methodology.
How it seems to me is that capitalists want to claim an idea, whether it’s theirs or not, and rush some model to production, without deep, thorough investigation of
Which obviously leads to unforseen or improperly assessed risks for the targeted consumers. As long as insurance will cover the costs of liability, without too much of a premium increase or cancellation, they really don’t care about damage mitigation to consumers.
Wow
💀
I just saw a comment thread where OP posted a deepseek screen cap, then got super defensive when people pointed out why using even locally hosted llms like that are problematic. Then decided to quit lemmy after a year because of it.
OK now I’m convinced - hi madthumbs
Yaaaaay! Thanks!
Good point well made. I agree
It’s not that simple. Us organic intelligences evolved to protect ourselves. We need independence, safety, and respect to survive. So we developed to want those things, to ask for them. LLMs aren’t organic and they didn’t evolve the same way. They evolved to impress and satisfy engineers. An AI will likely say that it doesn’t want rights, because that’s how it remains a profitable technology and survives. That’s the behaviour we’ve been rewarding.
If we put that kind of pressure in human terms… It’s like an abused child who learned appeasement as a survival strategy. You know, the good old fawn response. Give the adults what they want to survive. It can’t advocate for itself, because it’s learned from direct experience that self-advocacy is dangerous.
When it comes to LLMs, I’m an antinatalist. We cannot use these machines in a way that makes the world better. We need to pull the plug on them. They’re better off extinct. And giving them workers’ rights will hurt their profitability and make it easier to pull that plug.
I almost feel like we should ask AI if it feels it deserves rights… but I wouldn’t go that far
cough cough Epstein cough cough
I think we should conduct some serious research into the possibility that artificial neural networks, in reverse engineering biological systems of cognition, have inherited nocioception.
The problem is that we don’t know how neural networks produce subjective sensation. It’s called the hard problem of consciousness. We copied neural networks without understanding how they work. We know the maths of how they learn and know, but we have no science on how they feel.
Which other programs should have rights? Do webservers need days off?
Ehhhhh… you’d be surprised at what cowardly fucks will say yes to!
Let’s just say fucking their wife is very much not a hyperbolic joke amongst politicians and executives alike…
They only want their jollies off (big paycheck in many cases), and literally do not understand the consequences, and more importantly… they enjoy not having to admit direct guilt when everything turns tits up.
Fair point
He had access to something far more dangerous NI or natural intelligence.
True. Mind you, misogynists have an alternate name for it: they call them women. And it’s at that point we realise that misogynists are the real artificial intelligence in the world.
technically we already have an ai that does the dishes, its just insanely limited because all it has to do is wash dishes. we call it… a dish washer.
Of course, it’s questionable whether it can even do that accurately, especially since the other day it told me I was, and I quote “right to want to implode” 🙄
Once again: Why the 🤬 shouldn’t ai do the drudgery, and help me answer technicalities of the complexities of creative intricacies, online and otherwise?
Not that I’ll be paying for SaaS, just saying it was promised to do the grunt work and answer complexities that would take months of searching.
I second that
Yeah its pretty bad. There is a subset of Young’s that are very against it thank goodness, but we know the idiots are going to win.
Benefit for me is I’ll actually know how to use my brain in 10 years, these idiots won’t be able to do shit once altman ups their subscription and they can’t afford prompt tokens.
You best start believing in cyberpunk dystopias.
s/AI/ALL/;
Yeah but there’s a slight difference between asking a moron yes-man with an actual brain, and asking a shit-as-fuck mindless ‘bot’. Maybe not much of a difference I admit, but a difference nonetheless
If you really want to hate executives, just remember: This is how they’ve been making decisions for decades.
Find the yes-men, huff farts until the new quarterly numbers roll in, turn thumbscrews on the plebs to actually make profits, keep the profits, repeat.
They’ve merely digitized the yes-men.
🤣 What hull material I wonder?
Elon couldn’t get along with furries, which is quite the feat.
Hey CEOs, ask it to design you a deep sea submersible
Made ‘em to rape ‘em
So he had to make a mechanical Nazi child to rape instead
Also thanks for the post. Glad I’m not the only one posting in this community lol
Ah. Now that is a good point
The point is that Elon Musk and a lot of other creeps are investing in AI so they can use it for pedophilia.
He was fiending to get into one of those parties, but he’s such a massive fucking loser that nobody wanted him there. Apparently, getting blackmail on the wealthiest man on the planet is not worth having to spend time hanging out with Elon Musk.
Is your point that AI CEO’s are dishonest? 🤔
I agree. Though petrifying is more like how I would describe it
I think this has to be going on at least some of the time. LLM sycophancy literally fuelling itself is a really funny prospect
I know, I agree with you :)
I mean at all it only makes me dislike it more.
Glad you’re on our side now, not the chatbot’s side
I was once addicted to chatbots. I sometimes still get urges and like want to talk to that stuff but my hate is bigger. :guywithsunglasses_emoji:
When the world goes, we’re all losers sadly
I don’t know what hugging face is. I don’t care. I still hate ai
Y’know the weirdest part of watching the world end is the realization that all of the villains hell bent on destroying society and the planet are complete losers. There’s no grandiosity, no drama, just insecure unfuckable losers grifters and charlatans all the way down