Post

people are always saying anti-AI arguments don’t understand the full nuances of the technology. which is correct. but only because it stole everyone’s work, exists solely to get you fired, has made the entire internet unusable and has been forced on everyone without consent

1.6K reposts5.9K likes

Excuse you, i understood the full nuances of it when it stole my translation job: clients were happy to pay less for a shittier product because the corpo overlords deemed it sufficient.

M2 SSDs are basically RAM so they are now exorbitant which has had a knock-on effect of allowing manufacturers to raise prices on spinning metal HDDs so now an 8TB external costs more than a 12TB external did last year.

i've been needing to buy a new laptop for years and feel like i missed my chance bc now everything is 1) expensive af, and 2) loaded down with shitty ai "features" that i neither need nor want

From the beginning all the marketing from AI companies has been "this will put everyone out of work and replace all art and human interaction." Can anyone really be surprised that people's opinions on it are driven by emotion instead of understanding of the tech?

You're just not considering the nuances of the orphan grinder, it's about to revolutionize agriculture. Do you really want US agriculture to fall behind? Yeah it's pushing orphanages out of business but they're just an outdated technology now that we have the grinder.

It only gets more infuriating the more research you *do* bother with... Every tidbit about the technology itself makes you think about how useful a specific tool could be, if only the *entire conversation* had been conducted completely differently lol

It's also not even a single product, it's a marketing blitz for mostly LLMs but also lots of other machine learning applications. You can even appreciate LLMs and predictive models and neural networks and GANs and whatever for their own merits (I mean I don't, but you can) and still hate "AI"

no, the LLMs and image gens are the shitty, dead-end tech that all the problems are about. the other ML applications are completely different techs that they are conflating genAi with in order to manufacture consent for genAi.

Believing this will paint you into a rhetorical corner where you have to defend obviously false statements, I don't recommend it. Not as absurd as their claims that agi has arrived and LLMs can do everything, but not especially realistic either.

except what i said is 100% true. genAI is shit technology that will never be viable, and its also completely different tech then the "ai tools" that have been around for years that they trying to conflate genAi with.

AI is not the problem. AI in the hands of cultist billionaires plotting world domination while immersed in a Harry Potter inspired cult polycule is the problem.

The tech is an enabler, for sure, but I think the real culprit is the perverse incentives of our economic system that rewards the most heinous behavior.

Also, there are real scientists and engineers who do outreach and understand the importance of outreach to the public. We are having funding stripped from us. These tech bros, flush with money, do not value that, have never valued that and in fact exhibit general disdain for the public every day.

So why should the public be putting in effort to understand their so-called 'nuanced' arguments? In my field of study, if you can't explain yourself to other people, there are consequences. Either loss of funding or not being taken seriously.

Another scientist, Fauci, is being excoriated by the government and made into some sort of example for the rest of us who dare do our best to be truthful. Where is this scrutiny for the AI bros? Oh that's right, they are billionaires so they can say and do literally whatever they want.

The thing is there isn't much nuance to the technology. Every possible application would be better served by a specifically designed tool made by those who will be using (like everything else!) but these techghouls want an omnitool. They want the all-purpose hammer.

I’m being genuine here, sorry not trying to annoy! It’s just all the tools I use daily to build things now are only possible (I think?) with tech like an LLM and I get huge value from them.

Understanding the "nuances" also doesnt materially change most of the issues people have with it. People can get a bit overzealous in some ways, but ultimately the broad strokes hold firm. A jughead can yell about "curing cancer" or whatever, but the orphans are still being put in the grinder.

the colloquial one. this is a post about how average people are feeling and reacting to this thing that showed up one day and made everyone's lives worse for no reason not really a breakdown of the actual processes at play, which if i want to be pedantic is less theft and more primitive accumulation

I usually think of prim accumulation as the process by which things become capital. To become capital requires people, both the state and the people-proper to respect the rules of bourgeois ownership, the legal fictions that say "this capitalist owns this capital".

Now, clearly, there is a commercial character to AI output that "my posts I wrote" did not previously have, but my posts were still, legally, my property, and I could have chosen to extend their commercial character by putting them behind a paywall. Did this change?

makes sense. im very pedantic about definitions and stuff. its definitely making a lot of things worse but the whole "theft" framing bugs me

Feel like a lot more people would be quite willing to learn how AI actually functions and what practical applications it could have if it wasnt constantly being used in the worst ways imaginable by the worst people imaginable

It doesn't help that every time I give a specific function, area or use any leeway, I learn it's fucking shit at that as well. My learning process has been 'ok, I'm sure there must be a use... Oh, but that's not it.' every time. Put it away until it's fixed and useful to humanity.

Before the AI boom there was a YT channel that was like 'I taught a neural network to play Super Mario'. The bot trained on its own gameplay with positive and negatively reinforcing stimulus. More interesting than "my bot datamoshed a gagillion stolen arts to produce generic trash 24/7".

Yeah that's the video! Though I think I imagined it being more detailed than it actually was. Maybe got mixed up with a different unrelated video or someone else's video on the project.

that we were sold a Utopian version of AI benefits and ended up with something entirely different. are the uses for AI that i like? yes (science). but that's not how it's been sold to me. instead i'm being forced to consume actual garbage and the cost of a scorched earth is too high

I tried to capture more of the nuances bsky.app/profile/symb... 🧵

Linotype Pilgrim@symbo1ics.bsky.social

the fraud, circular financing, and ponzi nature the misallocation of capital the layoffs the destruction of learning outcomes the destruction of countless millions of books and other media the waste of single purpose semiconductors and electronics the pointless tax, water, and land use concessions

It’s not the consumer’s job to sell a product to themselves. They failed to adequately sell anything good about AI to the average person so people are reacting based on their experience, which is also often not very good. It’s not only not rocket science, it’s marketing 101.

"use it, or get left behind!" is the mantra used to deflect from the technology being used in a destructive way...... but it really be some late stage capitalism bullshit. why should any good willed people be 'left behind' simply cus they aint using a corporate robot?? wtf morality is that???

Enslaves third world workers to help refine the language models, guzzles water and rare materials like there's no tomorrow, makes consumer electronics and computers like 3x more expensive, monopolizes a shitload of money which could be put to use for a better world... And I'm not even finished

I hade functional understanding of ML and how LLMs work. I'm not anti-AI - it's a predictable advancement in data structures and algorithms. I'm anti-"ignoring the effects of rapid technological progress in a society already crippled by massive wealth inequality".

on some level, you have to accept that yes, people are reacting emotionally without doing detailed research but really you can’t blame them or stop them. this was some bullshit not a single one of us had to think about three years ago and now it’s making shit worse in like most daily interactions

Indeed, I spend a lot of time researching both the good and bad just to be sure that I know the standing of the tech, and hoo boy the pit they've dug's so deep that it bounced off China and back several times.

Also in my experience people assume the technology is way more complex than it actually is because they need to imagine the computer as a smart person who is also extremely dumb (for arcane reasons), instead of just imagining someone who is extremely dumb.

it's the same reaction we had to nfts. suddenly I had to know what the word fungible and blockchain meant because a bunch of weirdos decided it was THE MOST IMPORTANT THING while not really being able to make an articulate case for it i just wish AI shit would flame out as quickly as those did

Yeah, I've been trying my best to focus my own language towards explicitly calling out LLMs and image generation specifically in order to make sure that they can't come up with some absurd argument to block me out, because AI is way too much of an umbrella term

What I love about the "detailed research" whine is none of these guys understand, for example, what goes into actually constructing a massive building or just what the requirements of a 1.6GW gas plant that's supposedly going to power their fantasy building actually ARE.

Not to mention the fact that it completely obliterates any local economy near these data centers. These structures always bring down the value of properties nearby, which the only buyers of these properties will be data center companies.

This turns into a death loop for municipalities, since there will be fewer people living in the region over time. Business owners will move away from the region and likely take their businesses with them. Those who remain will eventually have to close up shop because nobody will be buying shit.

also. the thing people hate is not ‘the technology’ it is the tech industry. they hate move fast and break things. they hate having everything they say tracked and processed while companies do whatever they want. they hate every single ideological value that animates silicon valley

There is also plenty to hate about 'the technology' tbf. The theft and plagiarism of peoples work, churning out bullshit and its vast energy and water requirements all come to mind. What are its good points?

The "technology" is just the underlying math. That's fine, and could be useful. The stuff you're describing is the way the companies pushing this chose to go about it. They could have done this ethically, but they all believe they're in a race to be the last company on earth.

Not to mention the fact that everyone hates the fact that the technology itself is responsible for countless deaths, either through coercion or militaries using it in place of analysts. The United States almost started a damn war with China just over a damn hallucination over some cargo recently!

The United States started a war by targeting a girls school because, in part, it was nobody's job to verify the targeting data being coughed up. The nightmare scenario has already arrived and people have just kinda shrugged.

Pretty annoying how out of all these nuances discussions some dullards have especially from the left leaning or left wing ones they seem not to fucking care at all that this has pretty much destroyed the electronics market and barely anyone can afford new PC's and phones now

Exactly! Why do these folks expect a “rational response” to be to look at this thing that has tangibly made their life worse and more expensive, and go “we SHOULD start weighing the benefits” like fuck no! That conversation can’t logically happen right now

And anyone who has more smoke for “people being annoying about being anti AI” are doing that thing where they’re feeling mildly irritated at what is honestly a very normal emotional response to the scale of this shit, and somehow going “yeah the irritated ppl are at fault here” like what

The entire industry has changed user prompts from 'Yes/No' to 'Yes/Maybe Later' and it's just a poignant micro example of the direction the industry has taken towards pushing everything upon users whether they want it or not, whether it's useful or not, whether it's functional, safe, or even harmful

i mean, i hate the technology too, but i get your point. it actually makes it worse for me as they are doing all this with a tech that is fundamentally shit. its not even good tech that's being shoved down our throats. not that it matters in the end. it stole my work & of many other artists. fuck'em

As it turns out “your complaints don’t matter, it’s happening if you like it or not” was not an argument to make people receptive to the product

It’s also cray because the tech guys are coming at us with a combination of rape talk, negging, being high as kites, apolcolypse, I mean so little of what they are saying is rational or logical so why is the burden on us? Sick of all of this privatize profit but socialize responsibility and cost.

it's interesting cause markiplier of all people is putting in the effort to fully understand AI so that people can actually do something about all the negative aspects of it

I understand the core concepts of how it works, well enough to write a rudimenatary LLM. i'm not anti-AI, I just recognize that LLM-AI's basic premise is incorrect. It's a linguistic simulator. It's the thinnest surface layer of the ocean of intelligence, sapience, sentience. It's too wasteful.

I *cannot* stand that weird amount of finger wagging from people who aren't even that hard into genAi. It's acting as if people opposed to the rapid, destructive application of it were throwing the first stone in this stupid fight.

"You just don't understand it" I *understand* it wouldn't exist without the massive theft and that's everything I need to know about it to viscerally hate it on top of everything else.

I think it's really telling that people point out the many downstream harms surrounding this technology and are patronized that they aren't an 'expert in nuance.' Yet any discussion of clean tech is immediately met with cynics discussing in broad strokes any drawback, real or imagined.

Exactly, their argument reads like "the technology of the death ray Is actually really useful and can be used to treat cancer, you are losing the nuance of the death ray tech"

And even then, I still find that most of the time those pro-AI people understand the nuances of the tech less than those who criticize it, all they know is what they've heard from the marketing, no matter how obviously BS it is

Exactly. I’ve yet to talk to someone online about AI who understood the tech and wasn’t completely invested into it already. It’s hard to discuss the pros and cons when they have finances at stake.

And it gets so damn miserable asking for help online as well. I've posited a lot of questions about certain pieces of code I'm trying to understand, and the most common response I've been getting is just people telling me to ask ChatGPT or Claude.

I'm sorry, no. I want to be able to properly understand the arcane language our ancestors developed so I can get the damn rune machine to do its job on the first try, not speak with the digital yes man and want to blow my brains out.

"I understand that the mass spectrometer is a fantastic piece of research equipment with many interesting uses. What I don't understand is why every appliance in my house and workplace has been replaced with one without my prior knowledge or consent."

Also the **boosters** of AI technology are wandering around saying it will kill all of humanity! If several high-ranking Ford executives came out and said “our trucks go very fast but have a 10% chance of exploding in a fireball of death” no one would be confused when F-150 sales tanked.

I don't understand the full nuance of the chemical composition of a flashball round, but I still hate these when they're shot at eye level in the middle of a protest. Also, I'm pretty sure I understand the nuance of "AI" tech better than most of the bros forcing it everywhere.

The people that say they understand AI don't. They all think it's this magical computer thing and if you try to explain to them it's just a probability algorithm they tell you you are wrong. It's so bad that in some places in my industry they have training programs on dealing with non adopters.

I'm not saying that people won't do stupid things, just that knowing in detail how something works isn't required to recognize negative effects it causes

To extend the analogy, the chainsaw is currently working its way through our skull and it would be useful if we understood it well enough to turn it off and get it out

AI couldn't exist without exploitation of the global south and AI couldn't exist without child labour. And it's used for genocide. 3 reasons which stand alone why AI shouldn't exist.

imagine being so thin-skinned that you refuse to learn about the beautiful design of this no-name brand space heater simply because it also burns your house down every now and then. you just don't get it. in the future we'll be so warm we don't even notice our houses being on fire!

If I understand it better, will it benefit me financially, or will I just nod solemnly in the soup line, confident in the belief that that thing is out there doing my job better than I ever could?

i also really, *REALLY* hate the rhetoric some people apply that you simply have to accept it in tech and programming spaces, that if you don't you'll be left behind

The tech field is just a land of misery at this point, where you're not allowed to have cognizant peers to help guide you. I'm just certain at this point that the technology has enabled so many idiots to crowd the scene that you can't find a seasoned professional who can actually debug anything.

I always ask them why should we strive for mediocrity? We should we rush to put out flawed and defective slop? Usually met with scoffing noises and derision but no real logical answers

Some of the models are good enough to not output slop. The problem is that those models are proprietary and so resource-intensive to run that nobody else can even try, even if the model is leaked. The size a model needs to be to have a noticeably improved amount of intelligence keeps getting (c)

… bigger and bigger, the larger and larger they get. In other words, you get diminishing returns: with an 8B model, a 10B model isn't noticeably better.. But a 13B model is. With a 13B model, an 18B model isn't noticeably better.. But a 24B model is. And so on. As the size gets bigger, to make (c)

Said more simply, that argument, “you just don’t understand it”, is gaslighting. It’s a reflection of the pundit’s refusal to accept that you simply dislike it. It insinuates that you are uneducated, and therefore automatically dismisses your opinion.

I'm sure it's a great boon to the spreadsheet people. And their work does have value(not being sarcastic). But you trot out Tilly Norwood. You say it can do art even better than humans even though it lacks a soul or a concept of meaning. All I see is the internet getting shittier.

new inventions and making people jobless have been a thing for centuries, i don't think people will care about this argument point anymore, believe me, i tried multiple times harsh environment and the benefit of convenience are a no-joke combo, they keep saying it makes their poor life easier

I understand the technology perfectly well, I've been interested in neural networks since carykh posted his videos with evolution simulators. That's exactly why I'm able to see past the marketing and realize I'm roleplaying with a giant spreadsheet imitating a person.