Luddites ludditing. (via)
As a card-carrying AI-skeptic I found it quite disappointing to read the Association of Research Library survey that reported an entirely neutral-to-positive opinion on generative technology among research librarians, with not one participant taking a negative position on the matter. (Lo & Vitale, 2023) It is shocking to me that so many information professionals take this technology so optimistically and uncritically when it is quite clear that the recent proliferation of large language models like ChatGPT has done more to muddle our information access than fortify it. Google results have gotten objectively worse since they’ve begun to implement this technology, both due to the company’s internal and sort of clunky “AI overview” function, which now frequently supersedes all human-generated search results (Titcomb, 2024) and due to the rapid proliferation of LLM-generated slop content that has just devoured the internet in recent years (Rogers, 2024). The slop wave has even found its way into public libraries already via ebook lending platforms like Hoopla (Maiberg, 2025).
And even if we put the sea of slop currently being created by LLMs to the side, if we take these tech goobers at good faith and assume that there will one day be a fully functioning version of this technology that doesn’t, for instance, recommend that people eat rocks daily (Titcom, 2024) what would that even offer us? A slightly more efficient reference encyclopedia? The fundamental conceptual flaw of LLMs logic is how it assumes truths lies in averages, that if you just chew up all of the information in the world you will find knowledge at its exact center. This is a depressingly STEM-centric way to think about the world, as if knowledge was a mere mathematical equation and not an intangible human phenomenon that sparks at the intersection of experience and understanding. Yes, you can get a rough encyclopedia-level overview of whales if you ask ChatGPT but if you want to understand whales, whaling and the overarching human condition that informs these things while also having a rich aesthetic experience you’re going to read Moby Dick. No matter how “accurate” these LLMs get they will never be able to synthesize the effect of a true genius/weirdo airing out their particular human obsessions. (This is part of the reason why visual AI art is consistently so boring [Jennings, 2024] and getting worse [Taxxon, 2024]. All the greatest creators in human history were outliers, not median-dwellers.)
I feel a real loss today when I Google whatever random cultural junk I am Googling and end up with a bunch of clearly ChatGPT generated blog posts, all written in the same wordy stilted prose designed to keep on page while saying as little as possible, all seemingly built by feeding off of one another. Fifteen or twenty years ago those same searches would’ve likely yielded a few passionate, experts sharing their knowledge in an interesting (if usually chaotic) fashion. Some of them might’ve even ended up being my friends (I would even prefer a living, breathing enemy over a robot at this point, too.) We need creative, obsessed, idiosyncratic human voices to foster information distribution, to convert facts into knowledge and build social infrastructure to cultivate wisdom.
And information degradation is just one of the many damning facets of these types of technology. I haven’t even addressed the environmental damage it is causing (Todorovic, 2024). Or the labor concerns (Demirici, 2024). Or the racism it propagates (Elliott, 2025).Or its potential for accelerating war profiteering (Brenes & Hartung, 2024). Or the fact that a lot of this magic tech is just mechanical turk-type sleight of hand exploiting third world workers (Vertesi, 2024). Or that it is likely making users dumber (Mihov, 2025).
This stuff is all bad and I don’t know why it doesn’t inspire more skepticism among library and information professionals. Especially now that some of the same tech robber barons who have spent the past few years cramming it down or throats are gutting our government towards an explicitly fascist end. I think conscientious librarians should be taking a hard line position against this technology instead of remaining ~cautiously optimistic~ about this clearly toxic technology in the name of cursed efficiency or out of fear of not seeming like luddites. (The original luddites were right, by the way [O’Shea, 2023].)
The only AI we respect, playing HORSE with a human-generated claymation model. (2004)
References:
Brenes M,. & Hartung, W. D. (2024, June 2). A.I. won’t transform war. It’ll only make venture capitalists richer. The New Republic. https://newrepublic.com/article/182145/ai-weapons-make-venture-capitalists-palantir-richer
Demirici, O., Hannane J. & Zhu X. (2024, November 11). Research: How gen AI is already impacting the labor market. Harvard Business Review. https://hbr.org/2024/11/research-how-gen-ai-is-already-impacting-the-labor-market
Elliott, F. (2025, March 4). It only took a day for LA Times’ new AI tool to sympathize with the KKK. SFGate. https://www.sfgate.com/la/article/la-times-ai-tool-sympathizes-kkk-20202315.php
Jennings, R. (2024, May 23). Why AI art will always kind of suck. Vox. https://www.vox.com/culture/351041/ai-art-chatgpt-dall-e-sora-suno-human-creativity
Lo, L. S. & Vitale, C. H. (2023 May 9). Quick Poll Results: ARL Member Representatives on Generative AI in Libraries. Association of Research Libraries. https://www.arl.org/blog/quick-poll-results-arl-member-representatives-on-generative-ai-in-libraries/
Maiberg, E. (2025, February 4). AI generated slop is already in your public library. 404 Media. https://www.404media.co/ai-generated-slop-is-already-in-your-public-library-3/
Mihov, D. (2025, February 11). AI is making you dumber, Microsoft researchers say. Forbes. https://www.forbes.com/sites/dimitarmixmihov/2025/02/11/ai-is-making-you-dumber-microsoft-researchers-say/
O’Shea D. T. (2023, July 10). The Luddites were onto something. Jacobin. https://jacobin.com/2023/07/luddites-machine-breaking-capitalism-technology-climate-change
Rogers, R. (2024, July 2). Google search ranks AI spam over original reporting in news results. Wired. https://www.wired.com/story/google-search-ai-spam-original-reporting-news-results/
Taxxon, P. (2024, June 1). Untitled [Tumblr Post]. Retreived from https://patricia-taxxon.tumblr.com/post/753935866543702016/this-is-on-openais-official-demonstration-and-it
Titcomb, J. (2024, May 28). How Google’s malfunctioning AI risks ruining the internet. The Telegraph. https://www.telegraph.co.uk/business/2024/05/28/google-malfunctioning-ai-risks-ruining-internet/
Todorovic, I. (2024, September 4). ChatGPT consumes enough power in one year to charge over three million electric cars. Balkan Green Energy News. https://balkangreenenergynews.com/chatgpt-consumes-enough-power-in-one-year-to-charge-over-three-million-electric-cars/
Vertesi, J. (2024, April 4) Don’t be fooled: Much “AI” is just outsourcing, redux. Tech Policy. https://www.techpolicy.press/dont-be-fooled-much-ai-is-just-outsourcing-redux/
@bagel thanks for sharing your thoughts on AI and all the citations once again. I think I fall a little bit more toward focusing on the positive sides and I would also suggest the library 2.0 conference online that’s upcoming because Chad Mairn who is helping organize it and some really interesting ideas about how artificial intelligence and prompt engineering may impact and change our work for the better. I would also mention natural language searching, as well as another aspect that could further benefit from ethical and conscientious use of artificial intelligence.
https://www.library20.com/ai-and-libraries-3
thanks! i’ll check it out. i try to remain more critical than cynical with this stuff, even if my original post doesn’t read like that.
@bagel I enjoyed reading your thoughts. 😀
hi @bagel I super appreciate your post about AI, I’ve also been alarmed by the number of information profesionals embracing AI. I was pretty shocked that last semester in INFO 232, one of the assignments was discussion based this question “In spite of the fear around AI (Artificial Intelligence) it is here to stay. How will AI integrate into and advance our society through public library services? Give examples.” It felt like the profesor was implying that if we aren’t embracing AI, we are anti-innovation.
I share all of your concerns and I am proud to say that my local library is not engaging with AI services and is taking active steps to avoid AI. We have been exploring alternative options to Google Search to avoid AI results and currently working on moving away from using Google services like Google Docs. Our team cares about climate change and we know about the environmental impact of AI, especially its devastating use of water. I think that our library’s choices to research and use more ethical and environmental technology options is innovative. We have programs on 3D printing and coding for children. We value technology, but we understand the harm that AI is causing in our current system.
side note- I only recently learned about the Luddites from one of my fav podcasts and wanted to share it
https://truthout.org/audio/what-todays-workers-can-learn-from-machine-breaking-luddites/