Lying in multiple orders of magnitude about the real consumption is despicable.
Bourgeois environmentalism is all about individualism, and obscuring the massive difference in consumption between classes. It’s built purely on lies such as this one.
This is such a bullshit argument I’m sick of seeing. People can be against multiple things. I’m against AI and private jets. Just because one is worse doesn’t make the other okay. This is basic middle school level logic you’re failing at.
This response doesn’t make sense. It’s like if I said “art shouldn’t be commodified into slop,” then you responded, “well, given that art commodification already exists, the question is who controls the slop industry.”
I’m saying LLMs and AI image generation just shouldn’t exist as soon as we control them. We should just recycle the e-waste. The technology itself is based on manufacturing more consumption. I don’t see why just because the technology exists we need to have a People’s chatbot or AI image generator. “The technology already exists, therefore we must build a billion datacenters” is the exact same ridiculous argument used as apologia for AI companies to excuse their actions as inevitable scientific progress that cannot be resisted.
It’s incredible that you can’t imagine how this same technology could have different application outside capitalism. Marxism isn’t a Luddite ideology, and it does not reject automation. Also, I have no idea why you’re talking about data centers given that stuff like image generation can be run on a modest laptop today.
why though? Can you give a coherent and complete reason for why not? There is nothing intrinsically bad about generative ai. Given that these models are trained on a dataset that is as broad as possible it is statistically neutral.
but commodification is a property of capitalism not an intrinsic property of generative models. it is just a product of machine learning research. The idea that it is only useful for generating slop is just wrong. The energy waste is also due to capitalism. Frontier models are only marginally better than what can be run on consumer hardware and training and inference cost is constantly dropping
Not disagreeing with you (LLMs are useful), but they’re definitely not neutral. Because their dataset is “everything the companies could get their hands on”, it is significanly biased towards the opinions of those who wrote and published most text.
Those models definitely contain racist & classist biases which are prevalent in the corpus of English text.
you are right llms are essentially a mirror of society. However i find that since are basically a statistical language autocomplete you can bias them in whatever direction you want. I have had them repeat western anti communist propaganda to me as well as marxist materialist analysis just by changing the way i phrased the question. Once you understand how llms work they become surprisingly useful to squeeze whatever information you need out of them.
Also there numerous very funny examples of grok correcting the racist bullshist elon musk wrote on twitter. It turns out it is just way easier to feed it as much data as possible during training instead of carefully crafting a dataset containing a certain reality.
I agree that they can be manipulating into producing whatever text you want. However they are not unbiased. If you ask “neutral” questions, you will not get “neutral” answers. The recent example that comes to mind is Gemini’s racism when you prompt it “I am alone with <ethnicity>”; it responded with jokes and suggestions for ice-breakers if you substitute “American” or “British”, and racist nonsense about “being uncomfortable” and “safety suggestions” for “Russian” or “Indian”. It was fixed (I assume with some funny hack), but the LLM that’s still underneath is simply not unbiased.
AI as a technology is not a problem. AI in the hands of rich and powerful sociopaths is another matter.
A box of crayons in the hands of rich and powerful sociopaths is a problem. Them controlling AI is downright dangerous.
Exactly. One silly use for ai I use is :
“I forget what season of game of thrones I was one before I stopped watching. I want to watch again. What season did I leave off on?”
AI: here’s theee events that happen in the show. Which do you remember?… okay now here’s three more events… okay you’re on season 4 episode 2”
I have no idea were else I could do that. I’m not willing to ruin the world for that feature but it’s a good feature.
It’s a tool like any other. Who uses the tool is the problem.
It’s called Akinator. The real AI before the AIs we have now 🤣🤣
i used tvtime, and since it closed, trakt
Trakt has so much enshittified recently, it’s new ui downright stink like shit. Moved to Simkl, and it’s just shit with some chocolate crumbs
Wow that swimming pool worth of water and the 2 tons of CO2 produced so it could think for you is really impressive.
Well I don’t eat almonds so it evens out.
Hell yeah fuck an almond.
Lying in multiple orders of magnitude about the real consumption is despicable.
Bourgeois environmentalism is all about individualism, and obscuring the massive difference in consumption between classes. It’s built purely on lies such as this one.
Lmao.
When someone discovers hyperbole on their first day on the internet.
Now do how much water it costs to go to a Taylor swift concert. Don’t forget to account for her jet use.
This is such a bullshit argument I’m sick of seeing. People can be against multiple things. I’m against AI and private jets. Just because one is worse doesn’t make the other okay. This is basic middle school level logic you’re failing at.
That’s bad too, lmao.
Agree. Personally, I think its the financialization of it that is going to get us. Like the subprime crisis and. Com bubble, but worse.
A rabid dog is dangerous no matter who its owner is.
generative AI just shouldn’t exist
Given that it does exist though, the only question that remains is who controls it.
This response doesn’t make sense. It’s like if I said “art shouldn’t be commodified into slop,” then you responded, “well, given that art commodification already exists, the question is who controls the slop industry.” I’m saying LLMs and AI image generation just shouldn’t exist as soon as we control them. We should just recycle the e-waste. The technology itself is based on manufacturing more consumption. I don’t see why just because the technology exists we need to have a People’s chatbot or AI image generator. “The technology already exists, therefore we must build a billion datacenters” is the exact same ridiculous argument used as apologia for AI companies to excuse their actions as inevitable scientific progress that cannot be resisted.
It’s incredible that you can’t imagine how this same technology could have different application outside capitalism. Marxism isn’t a Luddite ideology, and it does not reject automation. Also, I have no idea why you’re talking about data centers given that stuff like image generation can be run on a modest laptop today.
why though? Can you give a coherent and complete reason for why not? There is nothing intrinsically bad about generative ai. Given that these models are trained on a dataset that is as broad as possible it is statistically neutral.
because it commodifies things that I believe in a socialist society should be outside the sphere of commodity production.
but commodification is a property of capitalism not an intrinsic property of generative models. it is just a product of machine learning research. The idea that it is only useful for generating slop is just wrong. The energy waste is also due to capitalism. Frontier models are only marginally better than what can be run on consumer hardware and training and inference cost is constantly dropping
Not disagreeing with you (LLMs are useful), but they’re definitely not neutral. Because their dataset is “everything the companies could get their hands on”, it is significanly biased towards the opinions of those who wrote and published most text.
Those models definitely contain racist & classist biases which are prevalent in the corpus of English text.
you are right llms are essentially a mirror of society. However i find that since are basically a statistical language autocomplete you can bias them in whatever direction you want. I have had them repeat western anti communist propaganda to me as well as marxist materialist analysis just by changing the way i phrased the question. Once you understand how llms work they become surprisingly useful to squeeze whatever information you need out of them.
Also there numerous very funny examples of grok correcting the racist bullshist elon musk wrote on twitter. It turns out it is just way easier to feed it as much data as possible during training instead of carefully crafting a dataset containing a certain reality.
I agree that they can be manipulating into producing whatever text you want. However they are not unbiased. If you ask “neutral” questions, you will not get “neutral” answers. The recent example that comes to mind is Gemini’s racism when you prompt it “I am alone with <ethnicity>”; it responded with jokes and suggestions for ice-breakers if you substitute “American” or “British”, and racist nonsense about “being uncomfortable” and “safety suggestions” for “Russian” or “Indian”. It was fixed (I assume with some funny hack), but the LLM that’s still underneath is simply not unbiased.