Oct 17, 2026, 1:48 PM@theluddite also I just want to be able to hate on AI but still have my space free of it when I need to and to document my opinions. For obvious reasons I think it really sucks the joy out of so much that yes I want to fight it (and I sadly work in an AI dept) but if everything is about it, they really do win.
@olivia You're *definitely* one of the people who does have something to say about AI! Too many people at 4s are following the trend then getting lost in a cloud of empty abstraction, which is basically the opposite of what you do! I'm always glad of your unpacking of a theory or model. If anything, you're the intervention some of them needed.
Oct 17, 2026, 7:34 PM@theluddite it's heartbreaking 🪻🌻💙 I also really appreciate your work
@olivia Thank you ❤️❤️ It means a lot from one of the coolest people here!
Oct 17, 2026, 8:00 PM@theluddite I link to that ridiculous replika nonsense often to show that when one digs one finds these people are lost and the conflicts of interest fractals. I cited you in a preprint too! Fingers crossed published soon, but still in revision.
@olivia omg that's so awesome! It's been extremely frustrating watching that paper gain citations even after our matters arising was published.
I also have some good information about the researcher's relationship to replika but I haven't been able to verify it with public records or find someone willing to speak without asking me not to publish about it. Ill find it eventually. Great project for any journalists following 😀
Oct 17, 2026, 9:02 PM@theluddite definitely please check if this is an okay way to cite you
Guest, O. & van Rooij, I. (2025). Critical Artificial Intelligence Literacy for Psychologists. PsyArXiv. https://doi.org/10.31234/osf.io/dkrgj_v1
@olivia yeah definitely! If you want to use my name it's Alejandro Javier Ruiz Iglesias and my co-author/illustrator is Julia Witte Zimmerman. At first I imagined that it might hurt my ability to find work so i didnt put it on the blog but I honestly can't imagine going back to writing software at this point. Plus i'm already publicly associated with it on my personal site and such so the secret is well out of the bag! I should probably update the site....
@olivia Good god you are spitting fucking fire in this preprint. It's an absolute joy to be included in it, and I have shared it with my lab!
I could not agree more with your stance against clickwork (e.g., for clickwork, https://theluddite.org/post/ai-poetry.html) and the LLM-as-data-dispenser phenomenon in the social sciences. More than half the talks at IC2S2 this year were, if you boil it down, people doing statistics to LLM output, usually asking it to quantify bespoke abstractions...
For example, here's one quantifying how civil rights a bill is (https://doi.org/10.1038/s41467-026-73607-x). Here's another that asks whether tweets contain a conspiracy theory (https://arxiv.org/abs/2602.11962). Fucking bonkers. Both these practices are high on my list of potential near-future projects.
Oh man I'm so pleased to have read this! Congrats again on a great piece of work.
Oct 18, 2026, 7:07 AM@theluddite thank you so much and @Iris and I will add your names! I will double check with you how to do it properly, but does it make sense for you to also include a "cite as" on your blog post itself? 🫶
@olivia That's a great idea. I will do exactly that. I started the blog as a (healthy?) outlet for my frustration after having been part of the tech industry for some 10 years (at the time... it was almost 15 by the time I left). I never expected anyone to read it and I haven't really caught up to that fact. I still manually maintain the subscriber list in an encrypted csv file and send out the emails with a script.
Oct 18, 2026, 12:43 PM@theluddite you rock