2016: Amber Heard exposes Johnny Depp abuse 2018: Depp in Fantastic Beasts 2 1999: Marilyn Manson admits rape threats & ex-gf murder plan 2020: JK Rowling sends Manson flowers 2008: Jeffrey Epstein convicted of CSA 2018: JK Rowling invites Epstein to Cursed Child A PATTERN

Guys I promise we do not need any more opinion pieces arguing that it’s okay to want to be thin

Please do not let all the focus on the funding environment in academia distract us from the real victims of these crimes and their enablers. Please do not look away from the misogyny. Please notice the gender differential. Please pay attention to how women are talked about in these emails.

relics from more peaceful online times: several years ago @devezer.bsky.social tweeted about Jazz24 (www.jazz24.org) and to this day it's been improving my work mood :)

In response to announcement that he will be a new CBS contributor, Andrew Huberman tweeted “I’ve always gone issue-by-issue on health & science, spoken to experts w/a range of takes & presented where I see the center of (data) mass pointed. I’ll do the same with @CBSNews”. In Sept he posted this:

Spain to offer legal status to half a million undocumented migrants 🎁: www.bloomberg.com/news/article...

Does the field of philosophy have a code of professional ethics? If not, I think y'all need to get on that, stat. A short 🧵>> www.businessinsider.com/anthropics-p...

immigration is a great thing, we need to say that - & take the fight to the nativists my latest for @liberalcurrents.com www.liberalcurrents.com/we-need-to-g...

✨ Updated preprint ✨ Iris van Rooij & Olivia Guest (2026). Combining Psychology with Artificial Intelligence: What Could Possibly Go Wrong? PsyArXiv osf.io/preprints/psyarxiv/aue4m_v2 @olivia.science Our aim is to make these ideas accessible for a.o. psych students. Hope we succeeded 🙂

"Sooner will the sun die out than that we will create AGI" I'm almost speechless, that's so clear. So much easier than trying to explain O notation and P/NP to someone non-technical.

🧵I'd love to get some opinions on ethical STT models (Speech To Text) for transcribing audio. As part of the Politechs project, we're trying to build an ethical podcast production stack, and one piece of which is machine transcription producing a rough transcript which a person can clean up.

I'm also curious about this. Speech to text is one of those things that was already pretty well developed technology before it was infected with the LLMs-do-everything approach, and I know some transcription software now inserts nonexistent text as a result. Are there still good tools?

In what's now become an annual tradition, I've written another piece for The Gist about how and why I hate A.I. On a bubble about to burst, a future rotted to its core, and the limits of the Mrs Doyle Theory of artificial intelligence. www.thegist.ie/guest-gist-2...

⚠️WOKE CONTENT⚠️ 🧵From the 1950s to the 1970s, Black inmates at Holmesburg Prison were subjected to unethical experiments overseen by a man who was America's version of nazi Dr. Mengele. (cont) #ProudBlue #ResistanceRoots #USDemocracy #Voices4Victory www.grunge.com/925197/chill...

HEARD THERE WAS ANOTHER YEAR COMING STOP WHAT THE HELL STOP FUCKING BULLSHIT STOP STOP IT STOP

I’m planning to delete my Instagram but I do enjoy using it to discover bits of queer culture (magazines, clothing lines, campaigns). Does anyone have any ideas on where to digitally discover these things in one place besides another big social platform.

genuine question: is there any large language model not built on the backs of exploited, traumatized data workers? #AI #ArtificialIntelligence #LLM #MachineLearning #StatsSky #RStats #GenAI

Okay. Since I posted this yesterday, so many people have tried to convince me that we "won" against crypto and it's dead. Because they don't hear about it anymore. I really need y'all to pay attention. Crypto is still huge. Not only is it not dead, it continues to cause massive harm.

My statement on teaching (looking for permanent teaching-centric gigs) docs.google.com/document/d/1... It really really is bad in the CS side, and I've dealt with too many faculty who are all gaga over it...

‪It has a name now 😜 Many thanks to Ken for agreeing to put his good name to my...artwork. The image is in the public domain (CC 0), but citations to the linked documents are warmly welcomed. ✅ zenodo.org/records/1808... ✅ pubmed.ncbi.nlm.nih.gov/24452418/

all this gaff around llms is like seeing slime molds do interesting things with solving mazes and just assuming if u put the slime mold in the right bucket and show it enough mazes one day it can be your friend and usher in the technological rapture

No, bro, I'm just asking if you have any model of cognition that the transformer would fit. Like even a vague but realistic shot. Something more than "look, we called them neurons so brain, we stack enough computational power together and then it ✨🌈🎉✨🔮🪄🎺, just one more datacenter we promise"

New post! There was a lot of innovation in medicine and biomedical research this year, and I've tried to summarize the biggest ones in this blogpost. Medical breakthroughs in 2025. Plus a serious note at the end. www.scientificdiscovery.dev/p/medical-br...

calling all #AI haters (complimentary): I participated in a project evaluating LLM-supported math tutoring and would love some criticism/references/whatever you've got that might influence follow-up projects. technical report here (abstract in picture): storage.googleapis.com/deepmind-med...

Partner and I started reading Mrs Dalloway aloud to each other last night. First Virginia Woolf for both. What a ride! Loving the adhd-frenzy tirades, the punctuation, the zooming in and out from small physical details to innermost thoughts to backhanded remarks about other characters #LiteratureSky

We end on "Machine Yearning for a Better Present" because why can't we dream? Why accept that universities are not places of learning? Nothing, except industry and their paid shills amongst us, force us to accept this & this force is not one of reason, but one of regressive values & profit. 10/n

which could, in principle, have been correctly labeled. If there cannot be accurate labels to train on, then a predictive project is hopeless, UNLESS your project is actually to impose labels rather than predict / recognize them. /end?

The perhaps deeper reflection: if emotions are not something we can distinctly recognize, then any "emotion recognition/detection" (e.g. "emotion AI") work is worse than just wrong: it is an IMPOSITION of an emotion upon the subject, arguably more damaging than just wrongly labeling something /4

Finally started listening to Feldman-Barrett's "How Emotions Are Made" and so far it's excellent. Great storytelling that systematically shows how the classical view of emotions as fixed, recognisable, universal categories is deeply flawed.

Read both Careless People (META) and (most of) Empire of AI (OpenAI) recently and this is very clear from both. Major safety issues are repeatedly disregarded to put profit or internal politics ahead.

If you're so against borders why do you have SKIN??

The most pathetic (and shockingly common) angry comment on our video about abolishing borders is "oh yeah? Well why don't you invite refugees into your house you hypocrite?"

But say it with me: resource scarcity is manufactured, resource scarcity is manufactured, resource scarcity is manufactured...

Think about refactoring. Refactoring code is taking code, and changing *only* how it pieces together. It does the same thing, but it now does it with such as more reuseable parts, less hardcoding, etc. It does the same thing as before, but it is now more maintainable. Easier to modify later.

the book in question --- not quite about fathering though maybe it can help you plan your children's parties www.priyaparker.com/book-art-of-...

Já não estamos a falar apenas de petróleo sob sanções ou de risco ambiental, embora isso continue a ser grave. Estamos a falar de navios potencialmente integrados em estratégias de guerra híbrida, a circular livremente em águas portuguesas

thinking of my upcoming 30th birthday parties, once again turning to Priya Parker's "Why We Father" to help me think

Perante isto, questionei formalmente as autoridades portuárias portuguesas. Questionei o Porto de Lisboa, o Porto de Sines, o Porto de Setúbal e o Porto de Viana do Castelo.

Maybe useful 1/2 bsky.app/profile/oliv...

EAs love malaria nets because it's supposedly the intervention where we have the "most statistical evidence" of it's effectiveness. One kind of silly fact about this tho is that if you read Duflo's actual paper, the choice to do the experiments on bed nets as an intervention is pretty much arbitrary

We do not have to offer and I believe indeed should not offer any benefit of the doubt: Anything sold as "AI" should be presumed ineffective, exploitative, environmentally ruinous, bigotry-amplifying and otherwise until proven otherwise. /fin

The take that tech bros get away with doing bad things because of academic critiques of AI hype is wild

"Women must speak. Do it loudly, firmly, without diffidence, and we must also learn to blast ignorance from a position of control, without reproducing men's bitterness and arrogance, their rudeness and contempt. We are not men; we don’t need to be men. We must not be men."

I've seen an argument that govs should tax automation as they tax labor -- when you pay someone a salary, the government gets a cut, and we should do the same when a company benefits from the labor of a machine that could have been a person or people. not sure on implementation but interesting

I agree that AI literacy is necessary. Students should learn what AI is. This is very different from accepting that students are going to use LLMs to write texts and code and helping them to do that

paper exists, are we supposed to ban hiding notes up your sleeves during exams? calculators with memory exist, are we supposed to ban storing formulas you must memorize? human speech exists, are we supposed to ban talking to each other during exams?

Reflecting on what I see here and elsewhere. "Despite appearances of social progress, women are not expected to speak to men, let alone discuss matters deemed their own business. The world of science and knowledge in general is one such protected domain. Women are not ⬇" cal-r.org/mondragon/S9...

"When women speak, men often take it as a challenge, frequently overreacting with a melodramatic display of erudition, in an attempt to conceal, under a veil of rational intellect, what is a palpable emotional response."