In March 2024, I wrote a post declaring “There Is No Ethical Use of AI”.
My point was not that you should not use AI. Personally, I try never to use it, but it’s getting more and more difficult as so many electronic products now don’t give you much option. As I said in the post, we are all enmeshed with unethical behaviors and tools every day, inescapably. Welcome to life outside of utopia! I am not at all a utopian; techno-utopian thinking has gotten us into a lot of the current mess. Purity tests will undo us all. Rather, I was arguing against the idea that there are ethically neutral or even ethically positive ways to use this technology. Though plenty of people want to argue that AI is, or might be, an ethical good, I think that’s delusional.
During the year-and-a-few-months since I first wrote “There Is No Ethical Use of AI”, a tidal wave of material about the technology has crashed to our shores, pro and con, cheerleading and doomsaying, shallow and deep. More material than anybody could possibly keep up with. (“A chatbot could!” sez somebody who wants to see if they can inspire me toward homicide.) And the technology has quickly progressed in a lot of different ways.
Still, my fundamental points in saying there is no ethical use of AI remain, as far as I can tell, strong and valid. This technology provides few unique benefits but it uses massive amounts of energy, it’s founded on copyright infringement and contempt for artists’ and writers’ labor, it is destroying many people’s jobs, is has ramped up misinformation/disinformation by orders of magnitude, it is a major accelerant for hypercapitalism, it is complicit with the worst excesses of police and the military … and more!
As I tell people who ask me for a quick explanation for why I loathe AI: it is destroying everything good in the world while making a small group of the worst people on the planet obscenely rich.
The early resistance to AI because of its copyright infringement seems to have cooled a bit, except among writers and artists. The general public doesn’t much understand or even care about copyright or about creative labor — it’s difficult to convince people that artists ought to be paid for their work, never mind that artists ought to have any control over how their work gets used. Decades, if not centuries, of devaluing artistic labor have led us to here. I myself have a thousand complaints about copyright law and the ways it inhibits creative community, but I am also fiercely protective of its basic existence because there is no other defense against exploitation. As much as I love Creative Commons licenses (the majority of my own writing has been CC licensed), the folks at Creative Commons themselves argue that CC licenses are useless against AI. (I am deeply frustrated by CC’s capitulation to AI. They’re trying to gild the rotten, poisonous lily with a new tool. Bleccch.)
One of the things that most infuriates me about the AI evangelists is their absolute disdain for art and creativity, a disdain baked into the product from Day 1. Big tech’s sneering at culture and creativity is a topic about which I haven’t yet found exactly the sort of essay I’d like to read, so perhaps in the coming days or weeks I’ll try to write it myself. Plenty of writing addresses parts of it, however (see the last section below).
Here, I’m just going to give a bit of an overview of topics that have arisen since I wrote “There Is No Ethical Use of AI” last year, then add some resources specifically about AI, culture, art, writing, publishing, and copyright.
Environmental and Health Impacts
The environmental and health impact of AI — and, particularly, of the data centers that these tools require — has become a more complicated topic in some ways and clearer in others. MIT Technology Review published a significant study of AI and energy use, which got lots of response. The most interesting skeptical response I saw was from Andy Masly, who wrote that the data actually suggest that text-based LLMs don’t use a significantly greater amount of energy than other internet apps but that AI video generation does seem to use a wasteful amount of energy. (Masly wrote this in a Substack newsletter, and these days I try not to link to Substack newsletters, given the company’s open empowerment of Nazis, but the information is valuable.)
Myself, I think it’s important not to put too much weight on individual users’ energy consumption via AI; rather, we can look at what AI (and Bitcoin) is doing to energy production and consumption generally — and that means data centers. Honestly, we should all cut down on our use of the internet generally, given, for instance, what it’s like to live next to a data center. Last October, the great podcast Tech Won’t Save Us did a 4-part series on data centers and AI. In February, the Luddite communists at Goldman Sachs estimated that because of AI use, “global power demand from data centers will increase 50% by 2027 and by as much as 165% by the end of the decade (compared with 2023)”. That same month, a report from researchers at UC Riverside and CalTech estimated that data center pollution has led to $5.4 billion in public health costs over the last 5 years. In May, Bloomberg published a damning report titled “AI Is Draining Water from Areas that Need It Most” and PBS covered the environmental impact of data centers — increased fossil fuel use, increased emissions, increased disease, increased global warming. Yesterday, the New York Times reported that “Big Tech’s A.I. Data Centers Are Driving Up Electricity Bills for Everyone”.
Of course, it’s possible — even likely — that AI’s energy needs will decline, for instance via similar innovations as launched the Chinese LLM DeepSeek. But that’s not what the energy trend lines are showing for the next few years, and during that time a tremendous, irreparable amount of damage could be done.
There isn’t a lot of precedence in history for energy demand declining. Since the industrial revolution, all we’ve done is add more and more sources of energy to try to feed an insatiable appetite. Sure, there are sustainable-ish alternative forms of energy, but we’re still using more fossil fuels every year. Until we reduce the demand for energy generally, there’s really no hope for any of us.
Slop, Fascism, Militarism
Meanwhile, the cultish forces of hypercapital are enshittifying the internet with infinite amounts of AI slop (a term which now has its own Wikipedia page. As does, for that matter, “enshittification“. Welcome to the future!).
There’s been some interesting thinking by Roland Meyer about how AI slop represents a fascist aesthetics. Gareth Watkins has also proposed a similar idea in his short essay “AI: The New Aesthetics of Fascism”.
The best piece I’ve read about all this, though, is “Deadly Slop: Artificial Intelligence on the Battlefield” by Sophia Goodfriend in The Baffler, which does an impressive job of showing the links between technocapitalism and militarism, a tremendously important topic when tech is so deeply complicit with policing, surveillance, military action, and genocide around the globe.
This Is Your Brain On AI
Another topic that has popped up everywhere recently is AI’s effect on users’ brains and learning. An MIT Media Lab publication, “Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task”, got a lot of attention for demonstrating how AI makes users stupid. (Okay, it’s more nuanced than that…) This follows on reports such as “AI Tools in Society: Impacts on Cognitive Offloading and the Future of Critical Thinking” that confirm what ought to be a fairly common-sense insight: thinking and learning require effort, require trial and error, require reiterative reflection … all the things AI tools evaporate for users. But it is a selling point of those tools that they abolish the work of thinking.
There has also been research into the psychological effects of AI, especially people’s (sometimes erotic) attachment to chatbots. Once again, researcher at MIT’s Media Lab have done useful research, finding that “Exploratory analyses revealed that those with stronger emotional attachment tendencies and higher trust in the AI chatbot tended to experience greater loneliness and emotional dependence, respectively.” The results are not all negative, including in that study. (Scientific American published an overview of the psychological pros and cons of “AI companions” in May.) There is emerging (and, at the moment, somewhat contradictory) research on whether “AI addiction” is widespread enough to be concerning, but there is plenty of anecdotal evidence for disturbing psychological effects, especially for heavy users of AI tools.
Of the psychology-based critiques of AI, I’ve most appreciated “The LLMentalist Effect: how chat-based Large Language Models replicate the mechanisms of a psychic’s con” by Baldur Bjarnason, published back in 2023, but I just saw it recently. I’m endlessly fascinated by con artists, fraudsters, psychics, etc., so the framing here was catnip to me, but I think it gets at an important point — AI’s lure is built on pretty old principles, ones that have been used to manipulate and trick people for centuries. See also The AI Con.
And then of course there’s education, which is being destroyed by all sorts of forces at the moment, of which AI is but one — though AI is also inextricable from the forces of capitalism and neoliberalism that promote anti-intellectual posturing and have made so many teachers’ and students’ lives miserable and turned so much schooling into shallow exercises and psychological torture. (For reference, see anything by Audrey Watters.) Just as I was finishing writing this post, I saw an announcement of a new website for teachers: Against AI. A great start!
I must also mention here my colleague Martha Burtis’s excellent AI Challenge module, a phenomenal nonpartisan first-stop for any educator; it includes material on teaching about, teaching with, and teaching against AI.
Resources: “AI”, Copyright, Culture
I was inspired to write this post by an excellent collection of AI-critical resources from Cate Denial (whose book A Pedagogy of Kindness is well worth seeking out). That collection is beautifully organized and selected, and it’s a great starting place for anyone who wonders why some of us think we ought to be highly skeptical of everything having to do with what gets called “AI”.
On Blue Sky, I noted that there’s a category of critique missing — that of artists and writers who are concerned about the AI companies’ devaluation of culture and art. This is not Cate Denial’s fault by any means, nor even a criticism of the resource, which is admirably focused. It’s just an observation. After she responded wondering if I knew of any good resources, I decided to look around to see what I could find, as I would myself like something I can point interested folks to. As I noted above, I haven’t found the one single grand unified essay on Big Tech’s disdain for culture, art, and humanity that I would like to read. But I did find (or revisit) plenty of useful essays, research articles, podcast episodes, etc.
Here, then, a few items not cited above, nor included in my original “There Is No Ethical Use of AI” post, or in Denial’s collection. All date from this year unless otherwise noted.
“Against AI: An Open Letter From Writers to Publishers”
“OpenAI and Google’s dark new campaign to dismantle artists’ protections” by Brian Merchant (ugh, Substack, but I so appreciate Brian Merchant’s work that I’m ignoring for the moment that he’s publishing with a Nazi-enabling tool)
“Theft Is Not Fair Use” by David Carson
“Shutterstock Has Inked a Deal With DALL-E’s Creator to Offer A.I.-Generated Stock Images. Not All Artists Are Rejoicing” by Richard Whiddington (2022)
“Is AI Art Stealing from Artists?” by Kyle Chayka (2023)
“The Unbelievable Scale of AI’s Pirated-Books Problem” by Alex Reisner
“AI Can’t Gaslight Me If I Write by Hand” by Deb Werrlein
“The Authority of Error” by Katia Grubisic
“Let the Machines Win (Because They Can’t)” by Shimanto Reza
“Translation, AI, and the Political Weight of Words” by Daniel Saldaña París, trans. Christina MacSweeney (2024)
“How Small-Town Public Libraries Enrich the Generative Research Process” by Nick Fuller Googins
“Neither Governed nor Free: On art in the age of generative AI” by James Duesterberg (2023)
“What a Year in a Robotics Lab Taught Me About A.I. and Painting” by Gretta Louw
“What Google Zero Means for Journalism” (Tech Won’t Save Us podcast) by Paris Marx and Matt Pearce
“Every AI Copyright Lawsuit in the US, Visualized” by Kate Knibbs (2024, updated May 2025)
“The AI Input Class: Constitutional Urgency and Fair Licensing in AI Copyright Class Actions” by Xuan-Thao Nguyen & Elizabeth Porter
“AI, Copyright, and the Law: The Ongoing Battle Over Intellectual Property Rights” by Negar Bondari
“Generative AI and Copyright Law: Current Trends in Litigation and Legislation” by Jim Rosenfeld, Bianca Chamusco, and Kathleen Farley
“In Copyright ‘Wins’ for Anthropic and Meta, Judges Leave Ample Room for Future Defeats” by Eryk Salvaggio
Discover more from Matthew Cheney
Subscribe to get the latest posts sent to your email.
