Would AI Be Hated As Much If It Wasn't As Resource Intensive?
If LLMs used very little RAM and processing power, was light weight even for Mythos or Sol, would people be as opposed? Like no data centers, no memory shortages, like it was able to run on a calculator with the same performance.
67 replies
LLM the technology is interesting and impressive. the way companies are shoving it into everything and trying to sell it as a replacement for employees is awful. The way the AI companies are behaving is awful.
AI is great! it just needs a little work on the IP theft, the environmental impact, the dopamine addiction, the cornering of the component market, the reliability, the hidden system prompts, the non-determinism, the push to have everything in the cloud, the enshittification, the empowerment of the billionaires, the reduction in user’s cognitive abilities, the lack of real-world use cases, the reliability, the spying on humans intimate thoughts, the overblown marketing, the bubble impact on the world economy, and the reliability. Other than that, it’s just fine! Trust me, bro!
It's great! Except for the several teen suicides
The same can be said about LITERALLY every technology ever made. Everything you listed is real, problems that need to be addressed, and the companies responsible for exacerbating them held accountable.
A technologies worst implementations are not representative of its usefulness.
Unless you want to ban the entire field of biology as well for the atrocities committed associated with it.
Yeah transformer models are cool.
But you cannot separate the industry from the tech in this case. The reason it’s so bad is because it’s being misused and it’s everywhere, even where it shouldn’t be.
No, that's not true. Universities have been making their own models. Private individuals have been making their own models. The most famous ones are, of course, purely corporate. But claiming that the tech is inseparable from the corporation is demonstrably incorrect.
Evo 2 was recently used in a revolutionary antibiotic research paper, which is entirely open source and trained and public data. Is that also corporate?
This is pretty much my thoughts as well. My unsolicited rant:
The technology is good and anyone who says it's "just autocorrect" are wrong. Having said that, I agree that LLMs are useful really for nothing more than low-risk, personal research and/or signposting to other sources. It's in no way a replacement for a human (unless your definition of what a human does is that it's just a thing that produces words at you - in which case, you're probably a middle manager - very myopic, regardless). It's also not appropriate for anything financial, professional, or confidential. Here, the first L in LLM is a hindrance. Local, small language models are more appropriate and cost-effective here.
Anyway. I don't "use AI" like some people but over the last year I've used my access to Proton's Lumo (already included in my subscription) a handful of times and been pleased with the results. The way that I see it is that all the LLM libraries are trained almost entirely on Reddit or websites that stole their content from Reddit anyway, so use it like you would Reddit: "LGBT horror movies before 2000," "Summarize and cite criticisms of Camille Paglia," "Was Bioshock 2 well-received at launch?" All very low-stakes. Doesn't really matter. Could've been a Reddit thread, but it's likely that any response there would've been a middleman for an AI answer anyway.
Anyway. More: Call it cope but I've been developing a theory/outlook that the internet as a thing is basically dead and that the Pandora's Box of LLM tech, which can produce a basically infinite amount of what was once the internet's currency (read: information, but really: text), has killed it. The Fediverse is utopian but I don't think anyone here really believes it will ever be anything more than a cluster of fringe communities. Having said that, I justify good LLM use (like mine, lucky that) as being the "New Game Plus" mode of the internet. Everything is already there, everything's unlocked. There is no challenge. It's kind of cheating and you're not really supposed to play the game like that. But at the same time, there is no going back to the real beginning, I just have the option of nostalgically re-creating it by going to old.reddit or postfixing before:2022 to all my
GoogleDDG searches. Fun, and game-ish in it's own right, but also artificially limiting for no reasons beyond that./rant
I was already sold on it, you don't have to keep buttering me up on the idea.
When everything is so searchable and indexable, obscurity is almost a feature
Except a large part of the reason that people used reddit is that you tended to get decent information. The big/good answers would often be some dude who is waay to into football having a 6 post page about who the GOAT player is.
Now, AI is basically so prolific, that a lot of what AI is being trained on...is other AI. Its slop badly reproducing slop, and the answers that AI gives you are either wrong, or are driven largely by somebody making 50 AI chatbots to swarm reddit with their opinion (or more often, to sell their product).
A lot of people like it this way but I really don't. What I wished happened during the last few reddit exodus was what happened back in around 2010 when digg users ran away to back-then-new reddit. The good thing about the fediverse is that people can actually choose instances they like that defedrates from content they don't like. I'm a hypocrite if I say I don't agree with that.
But somehow something must have changed with internet users between the digg-to-reddit exodus to newreddit-to-fediverse exoduses (weird that my auto-complete says this is the correct plural form) and not a lot of people want to contribute for making content. I'm pointing at myself as well so sometimes I do ponder why I don't be the change I want.
Gist of it is I want reddit circa 2010-2015 on the threadiverse specifically (as I don't use the micro-blogging side) but it's having trouble attaining good people that are willing to contribute time and energy to make worthwhile content.
No because dead internet theory and it literally makes humans dumber.
Lower reading comprehension. Less able to write or even explain something.
Lower attention.
Lower proficiency in just about anything ai can do. Lower critical thinking.
There is a ton of research on this.
Ai also tells people they are always right and gives them bad advice. Sometimes fatal advice.
I think you are jumping to conclusions. Many of these things started happening before AI. This shit has been happening far longer than the past few 2-3 years.
Reading proficiency and critical thinking started their decline when tech was introduced in the classroom.
Now the constant affirmation of one’s thoughts and essentially echo chamber I can’t dispute.
Don’t get me wrong. I dislike AI and it’s proliferation into the world as much as the next person, but we can’t blame it on our societies failing schools or rather lack of true learning.
I'm not saying it's all new problems.
Yeah you make a good point. Tech in the classrooms started the decline. Not to mention covid caused higher reliance on that tech.
And then primary schools have 1 teacher for 30 kids.
And they no longer hold kids back. Everyone passes rather they should or shouldn't.
When I was a kid, the social shame of being held back was a strong motivator. Not just for the kid but the parents to put in more time making sure the kid caught up.
And then even after I finished, I heard that mandatory summer school was being used to get kids caught up.
But idk how common that is around the country.
But to come back to AI. There is a lot of research showing using llms literally reduces reading comprehension and critical thinking even in a short time of 1 semester.
Fair point. People don’t have to put in nearly as much effort to convey their point if they can use AI and say clean this up, make it more intelligible, etc.
The problem is it’s designed to bullshit you. Why would I trust a foundational problem?
My buddy works for the government in contracts and he's mid level. Anyways he has like 5 people directly under him, and they're younger, were low 40s they're like upper 20s. He said sometimes he gets things written for multimillion dollar contracts that are fucking crazy because they use AI sometimes. He's like "I don't care if you do, but for the love of God read it once or twice".
It's also designed to eliminate your job and leave you for dead.
Yeah except the parasitic sociopathic class is already trying to find ways of getting rid of most of us from the for profit margins. Every employee takes money from the profit.
Any good thing is ruined once it's co-opted by capitalism.
AI and LLMs (In my very very personal and completely subjective opinion) is the coolest shit humanity has ever invented and under the right system would've been unanimously celebrated.
But noooooooo, capitalism had to go and fucking turn a miraculous tool for human creativity and knowledge into a relentless driver of layoffs, corporate slop and copyright theft because some greedy fucks want more money for themselves than they know what to do with.
But, well, that's just, like, my opinion, man.
Yes, because of the way its used, the way its presented, and how bad it is.
Different people would hate it for different reasons tho
The most common use for AI seems to be Nazi/Fascist/Republican propaganda. Basically anywhere you go is an endless number of bots pushing to defend neo-nazis, tell people that Trump totally isnt a pedophile, make posts about "If he had just complied he would have lived" etc etc etc.
The subset of people who recognize this, would hate AI because its really only good as a propaganda mill.
The next set would be because of the way its being pushed to replace workers, AI doesnt actually make the remaining workers jobs easier, and it doesnt do work more efficiently, its just another excuse to cut headcounts, make a shittier service and charge more for it.
Finally is because its A) Very often wrong and B) Being used to replace actual critical thinking.
Its a growing problem, especially in academia. Students are coming into college or high school with absolutely no idea how to do actual research, how to write a paper in a logical way, how to follow basic logic chains etc. And there is so much bad information being produced by AI now that its becoming harder and harder to find actual good information. For every doctor telling you to vaccinate kids, there are a dozen AI driven videos that have some person being confidently wrong about everything to do with vaccines.
Are there still students who didnt cheat their way through high school with AI? yes.
But they are becoming more and more rare.
And given how wrong AI usually is (AI is very good at being confidently wrong), this means that we have a whole generation growing up who cant tell the difference between "Measles can kill you if you dont get vaccinated" and "Shooting gay people is a fun weekend activity"
It would probably be hated even more as all of those points would be magnitudes higher in volume than they are today. It’s what is wanted by corporations, a cheap surveillance tool that closely monitors everyone to keep everyone in line. There are multiple dimensions to this issue. I think AI should have stayed in R&D for a while longer. It is so half baked, but unfortunately trusted as an “expert” by people who may not be familiar with hunting down primary trustworthy sources. Though that last point is nothing new. I think it should have never been used for things like scanning private/personal files or to accuse people of crimes they didn’t commit. These companies are buying their own bullshit and are trusting/betting on AI to reduce the most costly component of their company, the labor force. It’s all just smoke and mirrors though, always has been. It’s not like these things are gods among us, but they are sold that way. In the end, not matter what, with the current state of AI or if they were cheaper to use computationally, the bubble will burst. At least when that happens, maybe some of the left overs can be repurposed for actual useful things for society, but I don’t have my hopes up at the moment.
There are 3 reasons people hate AI right now:
So switching to local AI fixes problem 3 (mostly). But we also need the hype bubble to pop so number 2 goes away. Unfortunately the first problem will be hardest to fix and right now requires the public backlash as new norms are being formed and bad actors need to be called out.
That said, I think we're about 1 to 3 years away from the bubble popping and everything starting to balance out to where AI is treated more like just a tool rather than a genie in a box. The hype will fade out and one or two of the big players will die, but at this point I don't think generative AI use is going away
The trouble is that intellectual property laws are so far behind technology that there's basically no way to implement this technology without fucking over the creators of IP.
I don't understand why so many of the top comments omit this part
At work (big tech, ~100k employees) we had a panel about IP and AI. The guy presenting unironically said something along the lines of "it is of course a concern, but frankly it is something we will just ignore for now and see how the technology develops"
I took that as "we do not give a fuck if it creates AGI, and if it crashes and burns it will be the bankrupcy lawyer's problem.
so, yea, most people do not care
The resources used are but one of the problems with the modern uses of AI. Others include:
There are also legitimate and useful ways to use AI, and many of them could be very valuable, but that's not what the industry or society is focusing on right now.
AI would probably be more acceptable if all of these issues were addressed.
This song says it perfectly:
It's taking all the interesting and fun creative work, leaving people to make sure all the t's are crossed and the i's are dotted.
Where I think the bulk of the hate comes from is less the tech, but more the political and business elite pushing this tech.
These people are going full mask off trying to delete those they deem the lower class, so much so they are blinded in how much they require people outside the LLMs to maintain society.
LLMs as they are now, are basically the Golden Calf scenario all over again; it is ungrateful people build a statue to claim superiority and mock that which put them in good places.
Qwen 3.6 27B has similar capabilities to Claude Sonnet 4.6 for coding, but uses about as much power for inference as playing a game on the gaming PC running it, compared to Sonnet that probably runs on like 8 B200s in a data center. If people were mostly just running open weight models on gaming PCs with no tech bros involved, I don’t think we’d be seeing the same level of outrage.
There would still be many questions around ethical training and output reliability (depending on the task being asked), but yeah the major problem is that they’re using it to steal our natural and electronic resources and sell them back to us for something we don’t really need, that doesn’t really do what they’re convincing normies it does. The problem is that if you don’t understand what it’s actually doing, it appears very impressive.
In other words, it hits real hard if you’re stupid
Absolutely not, I hope.
The impact on the minds and the constant push for the thing for all the wrong reasons are enough for me to keep hating it.
And they always add it to everything, even where is not needed/wanted. It is a cool tech marketed and implemented in a bad way.
Resource intensive is one thing, theft is another. Pay all the artists for using their art and we'll begin to talk about whether destroying the planet further is worth it.
I would still hate it for being jammed into everythibg it doesn't need to be and being a shittier version of searches that used to work and being completely unreliable.
So slightly less.
It would still be anti-labor and a force multiplier for confident stupidity.
The main problem for me is that it's so confident in it's tone, like that idiot friend who just makes stuff up, but totally believes it. So there's no clue (like you would get with a normal person) that the thing has no clue if what it just wrote is true or not. I could bear it more if it said "I'm not sure, but..." But then if did that then none of the idiot CEOs would buy it, so here we are...
It is functioning in similar vain to idol statues placating and validating the idolmaker. In fact, that's my go to term for it because it's ridiculous for they practically deify this tech.
These 'AIs' are still dependant on human input and human information (not knowledge) and the fact they follow a business model and not academical, means they are always seeking to earn a customer's approval as opposed to teaching a student what they need to know.
The hardware and power shortages piss me off in particular because I'm a computers guy. But I also have friends who are writers and digital artists so I hate it for how it impacts them too. So 'as much'? No, but it'd still be hated.
edit: also I play TTRPGs so i hate it for polluting image aggregators too
I was at a scientific conference a while ago and someone said "did you know there's a piece of software that can check if any of your papers have been scraped by AI?" So everyone checked. And everyone's papers had been scraped. Everyone's.
There are still more major problems and risks than I can count on my hands
The question has already been asked yesterday. It makes people lazy and stupid, that's my main issue.
And it makes people who were already stupid think they're smart.
Nothing quite like arguing against a second-hand LLM conversation in your area of expertise, being proxied by someone who doesn't understand what either you, or it are saying.
"AI" is the wrong term, and I hate that it's been implanted. These are prediction models.
Prediction models can be extremely beneficial. Think of weather forecasting, for example. Being able to predict power usage across a grid to prevent failures and increase efficiency and properly deploy resources. Even LLMs can be useful to give you a starting point for discovering information that you are short on. (Don't trust implicitly, double check, but you'll have new search terms to use.)
Remember, too, that the publicly available models are relatively weak. Specialized models with well-curated training data are going to be much better at doing their specific prediction. And we're still in the beginning of this era. It's going to be crazy in ten years.
With regard to resource intensity, it'sy understanding that creation of images and video is what demands the most computing power. And that kind of usage sucks all the way around.
I’m very anti-AI but it’s mostly because of the impact it’s having on the capitalistic west. People are going nutty mode over it, and it’s unhealthy across a huge variety of variables.
However from a purely pragmatic take on usefulness, yours is fairly based.
Same here. The way it's being deployed and used and trusted by certain portions of the general public is troubling. There's a great potential for good, but it doesn't seem like it's going that way.
Generating text uses just as much energy..
Often for less time and fewer tokens, though. Not that it makes things any better…
I suspect yes. Even if you take these considerations out of the equation, AI is largely still just a glorified plagiarism machine being shoved down our throats.
There is benefit in models trained to perform specialized tasks, like assisting in medical imaging interpretation. But we knew this long before the “My LLM is god and deserves all the resources of industry and nature!” era.
It will always be resource intensive since companies are in a dick measuring contest of who can sink the most resources into it. If it was lighter they word simply do more of it
If it were truly actually accurate maybe not... but the issue is it is a syncophancy machine being obtusely oversold as capable of multitudes of things it cannot intrinsically do accurately or as well as a person with critical thinking and is being trained on fallible data sets that include human error and now hallucination from LLMs.
If it actually were capable of doing everything the people selling you on AI are trying to get you to believe it would be an incredible tool, but it isnt, and it isnt being treated as a tool.
I think it also does us a disservice as human beings because we are at risk of losing the ability to do our own problem-solving and process of logic bedause it can be offloaded to something else and that imho can manifest as a learned helplessness.
Yeah.....we're running out of those people.
Yes. Cause the resource intensity isn't the main problem people have with it.
It's the theft.
No, it would be less hated. Maybe only detested instead of loathed.
LLM and AI are two circles in a Venn diagram.
it’s one segment of the problem but not all of it by far
I think its hated because its about building dystopia. We see the datacenters coming up now that are the core infrastructure for global high tech dystopia and people still think it's about chat bots.
History books will wonder why people didn't stop it.
Slop is still slop, and if it was cheaper and less intensive in those ways it would simply be producing way more slop.
Would be even more. The main harm of AI comes from malicious actors being able to ruin the internet forever with practically undetectable bot spam costing cents to run. Also taking jobs in art, where automation was never possible as efficiency is not what art is valued for.
it would reduce it quite a bit. the whole house of cards economic thing not being a thing would help to.
I know I shouldn't be so snarky here, but your question is basically "would AI still be hated if AI wasn't AI?" Or "what if we had magic?"
My opinion: honestly, no. People hate LLMs because of the slop they produce on the internet and their use as a tool for surveillance, not even for the outrageous cost it takes to produce it and the financial criminality to keep it going.
Ed Zitron is the biggest voice in "our sane fringe" and yet he really doesn't focus on these immediate consumer-side effects because in the grand scheme they're inconsequential and actually (as in: mechanicistically) temporary.
There are some (uninformed) peopke who think the data centers are the only bad thing about the LLM Bubble, but anyone who's actually put the thought in knows that making less hurdles to proliferation is a bad thing.
I read the first few sentences, but already I disagree before your thought train had a chance to leave the station.
Think of your phone. Your phone might have AI, but mostly thats just connecting to the data center.
But imagine if it didn't. Imagine if the AI ran directly from your cell phone. Suddenly you have no need for data centers.
But it's not magic. In 1985 Nintendo released the NES to America. Previously released in Japan as the Famicom.
In 1985 if I asked you "what if there was a device like on star trek, wheere you could call someone, and it's a calander, and a camera, and a calculator, and a step counter, and a source of news, information, and a flashlight??? Oh, plus it can play NES games right on the device!"
You would say in 1985 thats magic. Except it's not magic. It's just tech that didn't exist YET.
Today it's called literally any cheap cell phone.
Now we know time will always move forward. I'm fairly confident that within 10 years AI run locally CAN exist. It probably can exist right now. The reason it doesn't and won't is because 5 years ago the tech just wasn't there yet.
So they bought a lot of land. They bought a lot of ram. They spent a lot of long term money on this early model of AI from 5 years ago.
So in 10 years, you'd expect there to be more fully functional local run AI. But here's why that won't happen.
When these rich assholes buy land, and all the buildings to be built and paid for by them, they feel obligated to use it. Even if NOT using it weilds better results.
Studies have shown that workers working from home result in more productivity, more work done, and more profit. Yet employers keep the big 12 story office with outdated windows xp computers running legacy software. Why? Because they already paid for it. So, "Come back to work dickhead!" says the boss.
Otherwise he'd have a big office, thats already paid for, but its empty.
So every data center that gets built should be considered permanent, even though from a tech standpoint it doesn't need to be.
So what they're simply asking is, what if the AI didn't drain towns of water and electricity? What if it just ran on your phone?
That's fine to disagree with me but I think it's a pretty generous reading of OP's phrasing to take away that all she's asking is "What if AI ran on your phone?" Having said that: you're still making some huge leaps here that do approach sci-fi if not magic outright. Firstly quantum computing as a field has stalled. To compound on this, transnational technological collaboration is also slowing. What this means is that it's unlikely we'll see another spurt of insane growth in mobile tech that will rival a $24,000,000 rack of Vera Rubins any time soon. The processor tech we have today seems pretty well plateaued.
Locally run AI can exist, and has for a long time now. It's very feasible, and also resource-intensive, especially on hardware that wasn't designed for it. The issue of data centers also isn't really removed here, just mitigated. First and most pedantic: data centers have been a thing since the internet was created. When you pull up a video on YouTube, that data is being served through a data center, it's only against the insane demands of AI generation that our existing supply of data centers isn't enough. YouTube doesn't need to generate that video every time you watch it, just store one copy and fetch it for you whenever you ask. The second point is that locally run AI still needs to be trained. Even offline models need basic updates to prevent model drift and keep accurate knowledge (to be useful). This training data needs to come from... a data center. Admittedly, in possibly smaller amounts than full-access to an online generalized LLM.
Anyway, another point: there was a lot of trial-and-error to get to the cheap smartphones of today from 1985. 41 years of it. In retrospect, most people would say it was worth it (most). Will people say the same about AI? It's doubtful. This is for the exact reasons I said above: the criticisms of AI come from when it works right! They're criticisms against the product itself, not like those against early smartphones ("It's too expensive!" "The battery dies too fast.") It's almost a given that efficiency will improve, but is this a good thing? More slop and more surveillance for cheaper? I don't think so.
If it wasn't terrible for the environment and the data it was trained on were ethically sourced, is hate it less. I still would use it or trust it, but at least I could live in a world with it.
It's designed to be scalable. You can already run basic models on home hardware, but available resources determine speed and power.
If you could get today's results on 1/10 the RAM/electricity, that would be good, but the big companies would just scale up to use the same resources anyway.
It wouldn't be hated as much if it wasn't forced upon us; if we had the chance to decide whether or not we actually wanted to use it.
Well if that was the case tech bros would just strap 1000000 calculators together, call it mega AI and proceed to claim it is the next best thing that ever happened to humanity and billionaires would fund them. It is these opportunistic vultures that enshittify everything and make them anti-human.
If the AI today always proceeded with the mind set of openness and locality it could still produce very useful tools for a fraction of the environmental impact it causes now but they instead, as always, have chosen thr most aggressive possible method because it can be hyped more.