DISCLAIMER: Before I start this piece, I’d like to highlight that no issue is black-and-white and that I tried to implement as many nuances as possible into this piece. I change my mind slightly to one or the other side of this issue essentially weekly because it IS at the end of the day an ethical dilemma just the way certain other lifestyle choices like flying or eating animal products is. This piece was made to make you think, not for me to claim I’m right. I want people to create solutions that are systemic and I firmly believe that can only happen by engaging with nuanced takes about complicated ethical topics like this.
So before I start to explain to you why the anti-AI movement will likely fail if it keeps going like this, I have to cover a few things real quick. I myself think it is certainly an issue that many everyday people use AI like it’s all-knowing. I myself am someone that only uses LLMs for any answers after I tried several ways of searching for something on DuckDuckGo or for truly complex tasks that otherwise would take hours. I myself would/could never defend AI to the extent tech bros do. I feel like this has to be prefaced because I’m about to play major devil’s advocate. And I’m playing devil’s advocate from the position of a systemic thinker who very much understands why people (not companies) elect to use genAI. And, obviously, this entire essay is about generative AI, not the predictive models used in medicine and other industries to improve life quality.
So let’s start analyzing this…
Stay with me here! How often have you seen an anti-AI comment somewhere that could not back up their claim with at least a vague mention of a study? I don’t mean a linked study, I mean something a simple as “someone researched this specific part of the server farms and found that xyz is true” or mention a journalistic piece about local effects. I find this even more egrigious when I see it in articles and videos about the topic. People keep referring to water waste but they keep not citing their sources. Because if they did, they would quickly come to find out that there is a LOT of complexity to those. There is the complexity of different sources not agreeing on how much water is wasted. There is the complexity of what gets counted and what doesn’t because while cooling of server farms is certainly water-dependent, the power that the servers are ran with is often cooled closed-circuit. So we have an issue of what gets counted where. Doesn’t mean server farms aren’t significantly impacting certain areas, it just means that for a lot of that stuff we don’t have the clearest data. So it’s easy to potentially over or even underexaggerate the waste of water.
The whole water use claim is also something that underlines the lack of research many anti-AI people online do in the sense that, systemically speaking, the core issue isn’t the water use, it’s the decades of all types of companies not doing anything for water recuperation infrastructure (looking at you Nestle and Co.). Water isn’t eternally lost if you have infrastructure for it to have a healthy balance to at least some extent. I’m no environmental scientist, but the issue clearly is sooo much deeper than just server farms needing drinking water for cooling.
Listen, I get it, I like being correct too as a very rigidly thinking autistic person. HOWEVER, it truly is mindblowing how nuance-less the anti-AI movement often is past its well-researched core advocates. The point blank “you are a bad person” type of statements hurled around are concerning. You aren’t thinking from a systemic or holistic standpoint if you don’t grasp that there IS nuance in AI usage. You can deny it as a valid accommodation for disabled people all you want simlply because some disabled people live fine without it, but you’ll still entirely ignore the people that it indeed helps and accommodates. It’s also ignoring the amount of companies who are making people dependent on using genAI on the job and the amount of jobs now demanding some level of genAI knowledge for employability. I’m all for you shitting on Fortune 500 companies for making it a systemic necessity, what I’m not about to hear is that something is “always” bad. Moral absolutism isn’t helping anyone, basically ever.
Many anti-AI people will backpedal and claim their blanket statements don’t include people who get accommodation from genAI or are required to use it on the job but the shame still hits the individuals harder than the companies that are ACTUALLY the ones using it at an unnecessarily large scale. It is simply a fact that generative AI has positives for individuals under capitalism – a system the rewards productivity. Shaming individuals more than shaming elites and companies is just not helpful because we’ve already done this with other big issues and sure the shame works, but it works way too slowly for the goal you’re trying to reach and leaves out so many realities of life (looking at the veganism crowd specifically).
It’s a classic at this point that people will claim that using genAI makes you dumber. And I cannot overstate how much you lack the critical thinking you defend so fervently in the “don’t use AI” debate if you think you could securly show sustained cognitive decline like that without a longitudinal study that spans at least a decade. We have more proof for social media increasing rates of anxiety and depression than we have proof for genAI making us dumber. Devon Price wrote an excellent piece going into what the research everyone points to is actually saying. Hint: A lot of people need to learn that “cognitive offloading” doesn’t mean you get dumber, it just means you might with time use some parts of your brain less if there is no need/motivation to use them. And frankly, think about things that were normal 20 years ago that you also don’t do anymore and therefore don’t use your brain for it anymore. Might not be the best comparison but it gets the point across that it isn’t about people getting more stupid.
Also going back to the whole water usage thing: Plenty of people will claim numbers for water waste but we haven’t even agreed on a range of water waste of genAI usage and the parameters used. There isn’t a consensus so you’d have to point to your source. Like, you legitimately should link a study when you claim a specific number because boy oh boy do different sources say entirely different numbers. And boy oh boy do people not think about how their usage of other server-heavy things might impact the environment (ever thought about what running Netflix in the background does?). The anti-AI crowd REALLY needs some awareness about what they DON’T know yet (at least scientifically speaking).
As I mentioned earlier, moral absolutism and its black-and-white thinking are prevalent in the anti-AI movement. Don’t get me wrong, I entirely understand why this is the case because with each day this world gets closer to water wars actually becoming a thing. What many of these people don’t seem to grasp is that their friend Betty using AI isn’t what they should be having their rage focus on. The rage should be directed at the entire business models of most businesses of the elites that have planetary harm as the goal instead of just the collateral at this rate.
You cannot be this mad at genAI usage without organizing in some way in my opinion. If you aren’t at least reading up about different ways to change the way society runs I’ll assume you’re just in this movement to be mad about something without having to put in any effort. Being anti-AI and not leftist is frankly kind of embarassing in my opinion (it’s hollow!!!) and I do see a massive amount of liberals having zero self awareness about how their rage will lead nowhere without systemic change. Currently a large part of the movement is the grandpa yelling at clouds meme. You can shame people all you want, if the system sticks around the issue sticks around (again: looking at the veganism crowd specifically).
A lot of antis I’ve crossed cannot even discuss the nuance between image generation and the use of LLMs. I’m sorry but that’s so pathetic to me personally because not seeing how an LLM actually has great usecases is a sign that you did not use any critical or systemic thinking, you just wanna be angry. And I know I’m repeating myself with that but it truly is so obvious for a certain type of person within the anti-AI movement.
I feel like I’ve led up to this the entire piece so far, but the anti-AI movement massively lacks systemic thinking (the sociology / social work kind) & systems thinking (the engineering kind). I know many of the leading voices in the movement are well aware of the systemic connections of the genAI issues, but the people outside of the direct core of that movement at large act, frankly, dumb as rocks. Which is infuriating to see as someone that understands anti-AI as a movement but sees the nuances because of my systemic thinking. The fact you see anti-AI people fighting against anti-animal-agriculture people left and right when the topic comes up is a great indicator of that (it happens so much more than you’d think). Because, for once, the annoying vegans (sorry, but also not sorry) are correct. And I say that as someone that is “only vegetarian.”
The vegans have had this exact issue for decades now (and they already HAVE a scientific consensus on their side ffs!). The same “you either go full vegan or you’re horrible” type statements. And as I said earlier in the piece, this shame works but it works too slow if your entire rest of the movement for the most part has only three protesting tactics that are all pacifistic. The vegan movement has literally the SAME systemic enemies as the anti-AI movement because the enemy is the people in power and with money. It’s just in this specific case the water waste animal agriculture creates is so much larger than what AI does that it feels like a fever dream when anti-AI people claim “that’s different” because it really isn’t. I’d argue my 12+ years of being vegetarian minus the times I use genAI to accommodate myself still equal out to a net positive when it comes to preventing water waste. I’m not saying yield to the vegans, I’m saying look at the failures of the vegan movement so far and learn from it because it’s more similar than everyone likes to admit (because that would mean those militant annoying vegans were right, listen I hate it too because I also hate when that crowd doesn’t have any nuance, lacking nuance always fucking sucks).
Again: This comes back to the bottom line of “you can’t look at the symptom, you have to fight the root cause.” And, fam, the root cause isn’t individuals using genAI, it’s not a department of a corportation using genAi even. It’s systems making the individuals feel like that’s easier than finding community and learning stuff in their free time. It’s the system of capitalism and hyperindividualism and anti-intellectualism and I could keep adding words here but you get the idea.
If you want to change something you need to learn it first. That includes the positives that your “enemies” will rattle down each time. The anti-AI movement, if it wants to be taken serious, needs to actually make sure even the most casual anti-AI people understand what the fuck they are talking about. That means you better make sure far and wide people grasp what we have research on and what is still hanging in the air. That also means you sit your butt down and learn about the reasonings everyday people have for using genAI. And I don’t mean the simple first answer of “it makes my life easier” but learning what needs get fulfilled and what issues they actually have at the core. I promise that alone will make you a radical anti-capitalist if you weren’t one before because you will hear how much of this shit is a system that leads to the usage of the poison aka genAI.
The anti-AI crowd needs better messaging and understanding of psychology. Why? Because right now anyone that isn’t an actual scientist or leader within that movement looks like a kid throwing a tantrum. That’s another reason why I drew the veganism parallel. You don’t get people that way nearly fast enough. Yes, you will get some with shame and some with shock through documentaries, but you won’t ever make it an effective movement if you don’t create better messaging that gets carried forward (accurately) by even the most random idiot in a comment section somewhere. Because believe it or not, those temper tantrum randos in comment section represent your movement the most right now and they are more annoying that helpful (again, I’m saying this as someone who leans towards the anti-AI crowd on the ethical topics).
Learn to live with the dichotomy of “it’s bad, but it’s helpful” that a majority of people live with about genAI. Majority I’d say are well aware of how bad genAI is for the planet but they see the usefulness of it way more because it’s right in front of them and something they can FEEL. Not everyone is a point blank hater of it. It doesn’t work to simply go “I cannot believe how anyone would need this” or the more holier-than-thou takes that go around in especially artist circles. That isn’t fucking effective on a systemic level because you’re the equivalent of the nerd group that still had slide-up phones when everyone else already had smartphones (and, again, I’m saying this as a hater of image generation with genAI).
You cannot get out of this without 1) educating yourself, 2) educating others and 3) building alternatives. The first point I’ve covered now, the educating others part I covered to some extent, but educating others means so much more than citing sources. What I NEVER see discussed are simple question matrixes people can use to determine if something is worth getting the answer to without using AI. Because let’s face it, in some cases the hours it might need to get a certain answer by calling a place and being in 50mins of a waitline is a net negative for the average person (plus they’ll often speak to an ai there as well). Yeah yeah, “reject convenience” and all that, but come on you won’t get a middle class person to reject convenience on that big of a level). You cannot deny the power of being able to backwards search an answer (also a skill to teach especially the older crowd leaning more and more on genAI without being prepared for it at all).
I also barely ever see discussed that LLMs can be ran locally on a computer. I’d say a decent amount of people nowadays have desktop computers capable of running genAI locally to use up less water (because let’s be real, we won’t ever backpedal on the entire internet being unethically crawled and fed into the machines, you cannot undo the unfortunate usefulness of that dark af grimoire of knowledge imho). Make running this shit locally easier if you’re anti-AI but a CompSci or Engineering major. Harm reduction could be done so much faster if you join that fight from the right side and finally admit that you won’t get people to stop using it unless all governments ban it (not gonna happen, even if capitalism dies you’ll have people who’ll want the solarpunk type of future includign genAI in some form so only a further point in favor of localized LLMs). Make the less harmful option easier. Make the vegetarianism to the veganism if you will. Like, seriously, I’m not bad with tech stuff, but running an LLM locally still looks way too complicated with the amount of extra steps still necessary.
Actually work as a collective to make more encyclopedias for what people use genAI for most in their own lives (understanding paperwork, medical terms, legalese, etc.) and push the online glossaries that already exist back up in the search engine rankings. Like, for fucks sake, some of you need to simply sit down, study the Google search algorithm and then figure out how to make important knowledge bases show up more easily on the first page again. In all languages. And some others of you should be standing up way louder for the payment of what were wage slaves even before genAI made them redundant. That goes for you reading this too if you really want your fellow humans back as customer support. You don’t get to sit there, point a finger, be angry and think that is enough (join a union btw).
All I’m saying is: There are ways to reduce the harm if you actually start critically thinking and stop ignoring the psychology of humans at large. You can’t out-angry an exhausted working class seeing a tool that’ll make them slightly less exhausted.
I wouldn’t use genAI to help me study and look for research if we had an improved higher education system and full-sentence search function for finding studies and academic books. Hell, let’s tie that in with some other shit. GenAI does so much shit that I still barely use it for but want to when it comes to work-related stuff because the useful non-AI alternatives cost money in a subscription model. That’s either solved through anti-subscription movements (meh, not fully sensical as long as we’re in capitalism) or through making students and workers more financially stable (I could be quoting the communist manifesto here but I think that would be too on the nose).
We need more power to the people again for anything to be fair again without ruining the planet. If you are anti-AI but not absolutely lividly anti-capitalism then I’m shocked. The issue then becomes the policing of other anti-capitalists. I focus my work on other things. I’m vegetarian to reduce my impact on the planet while upholding a small community that focuses on preventing violence against woman in a specific community space. I very much do work that is a puzzle piece in moving towards a better world and future, but tend to occasionally use genAI in my life. Meanwhile you might be omnivore, not use genAI at all but drive a car for 2h daily for your commute while trying to boycott almost everything on the BDS list. We can’t all do everything and we gotta stop shitting on other individuals(!) who are mindful of genAI use while gently(!) educating others about overuse. The big issue lays with big megacorporations, lobbyists and governments aka the entire economic and political system.
I feel like I’m talking in a circle at this point so let me just end it with this sentence: You cannot reduce a system to its individuals and you have a treasure trove or knowledge you could gain from the failures and wins of the vegan movement in the 2000s and 2010s. USE THE ACTUAL KNOWLEDGE AND SCIENCE WE HAVE ON HUMANS AND SYSTEMS TO SOLVE THIS FFS!
PS: The boring ass take that “there is no good reason to use genAI” or “nobody needs genAI for anything, stop using it” are making you look stupid as fuck. They might at the core be true, but it just sounds like ableist as fuck statements, because there are more than enough ways genAI has helped people with different disabilites to cope without needing to spend thousands on other accommodations. Again: You cannot tell people to stop using something point blank, you HAVE to understand the needs that are being fulfilled and build alternatives. That’s literally the age old topic of anti-capitalist leftists: You cannot strike down a system without making people self-sustained away from that system (there’s a reason the Black Panthers were seen as a risk by the government at their height). Just because AI is so much newer doesn’t mean that notion doesn’t still apply here.
Anyway, I hope my two edits of this got all my points clear. I warn you that the block button is used freely on here if you get insulting or willfully misinterpret the message of my writing. I do enourage civil discussion.
If you want to support my writing I prefer people using Ko-Fi for one-time and monthly financial support since the payment processor Substack uses is not following GDPR guidelines for the country I live in correctly.
Through Ko-Fi you can do one-time donations AND subscription payments. Using an external subscription service means that all my Substack posts are able to stay free to everyone.
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.