Eddie Antonio Santos @eddieantonio.ca · 27/09/2026Reason #47 for why you too should maintain a Python package called "crimes": 011
Reposted by Eddie Antonio Santosleon @leyawn.bsky.social · 24/09/2026ceo of burger king: our burgers have gone rogue again and infected people with e coli 5256191273
Reposted by Eddie Antonio SantosIris van Rooij 💭 @irisvanrooij.bsky.social · 19/09/2026All we can expect when trying to scale up AI systems is diminishing returns and devastating environmental damage. Calls for `slowing down’ can conveniently hide the diminishing returns, while still convincing people that the tech is “all powerful”. 514756
Reposted by Eddie Antonio SantosDr. Lindsay Stallones Marshall (she/her) @lindstorian.bsky.social · 19/09/2026People designed genAI so it always replicates our existing injustices. To take an already underrepresented group & just replace them w/ chatbots trained on centuries of anti-Indigenous writing, esp when reporting immediately affects their communities, is revolting. There MUST be real consequences. 05525
Reposted by Eddie Antonio SantosMichael Kölling @kolling.uk · 10/09/2026Looking for new funders. BlueJ, Greenfoot and Strype were funded for a long time by Oracle. This has ended, and we are looking for new funding. If you work at a company that wants to do something good, we would love to talk to you. Please repost. bluej.org greenfoot.org strype.org 118
Reposted by Eddie Antonio SantosGoose, The Shenaniganizer @goosethearchivist.bsky.social · 07/08/2026This is where you turn it off. Gmail -> Settings -> Data Privacy (turn it all off) -> Google Workspace Smart Features (its a submenu that leads to this page) 2119109
Reposted by Eddie Antonio SantosDr David S Anderson @dsaarchaeology.bsky.social · 10/07/2026 3348242175
Reposted by Eddie Antonio SantosEmily M. Bender @emilymbender.bsky.social · 07/07/2026I have a foolproof way to make sure that your papers contain absolutely no fake references . . . . . read the papers you cite. 201038216
Reposted by Eddie Antonio Santosblair @blairaf.com · 02/06/2026I'm looking for a new PhD student to come join me at UAlberta & Amii to study trans perspectives on AI policy & the future of AI as part of my new lab. The position is fully funded and offers access to a range of supports - more details in the poster below! 37263
Reposted by Eddie Antonio SantosLeah McElrath @leahmcelrath.bsky.social · 21/05/2026I'm finding this is a huge issue in Google Al search summaries. When I trace back references, they often don't support the argument the Al summary cited them as supporting. Google search has been degrading for years, but it's often a straight up misinformation purveyor now, so be careful with it. 1521380
Reposted by Eddie Antonio SantosJames Finnie-Ansley @jamesansley.com · 06/04/2026I’m always looking for more digital stores to add to drm-free.info. If you know of any drm-free digital media/software vendors please feel free to contribute to the project by making an issue or pull request (repo linked in site footer), or let me know of any stores I’ve missed and I can add them.drm-free.infodrm-free.infoA list of DRM-free vendors for music, books, and other media 021
Reposted by Eddie Antonio SantosE.L. Meszaros @iastranger.bsky.social · 25/03/2026It was such a treat to be included in this! Also time is a social construct. 083
Reposted by Eddie Antonio SantosHila Peleg @hilacodes.bsky.social · 18/03/2026We're running a user study, maybe you can help! We're studying the effect of new ways to show code diffs to programmers, and looking for participants for a 1-1.5hr virtual user study. In the study you will be shown several diffs and asked questions about them. 12121
Reposted by Eddie Antonio SantosRoberto Faria @betofaria.bsky.social · 06/03/2026To my Canadian friends, an acquaintance started a petition to allow US transgender citizens to claim refugee status here. www.ourcommons.ca/petitions/en... If you’re Canadian please sign it. If you’re not, please share ❤️ourcommons.caPetition e-7195 - Petitions 13919897
Reposted by Eddie Antonio SantosCoHSAT - Coalition for Healthy Streets and Active Travel (Oxon) @cohsat.bsky.social · 03/03/2026In shock news, motorists have switched to public transport to avoid paying the Oxford Congestion Charge. A spokesperson* said: "This is exactly what was meant to happen, and it did." 271053251
Reposted by Eddie Antonio SantosWalter Rowland Butler @wrrb.bsky.social · 15/02/2026Violets are grue Roses, demonic Si enil txen siht Boustrophedonic. 38821
Reposted by Eddie Antonio Santosderek guy @dieworkwear.bsky.social · 26/01/2026stop using AI to do your research. it hallucinates too often. if you want an answer to something, post something arrogant on the appropriate subreddit. something like: "this item performs 10% better than everything else. only idiots deny this." this will bait nerds into doing your research for you. 144176533560
Reposted by Eddie Antonio SantosTimnit Gebru @timnitgebru.blacksky.app · 22/12/2025Wildly different things, tasks, techniques, subspecialties being lumped into "AI" and then being conflated with each other, doesn't help. Different types of models vs the techniques to train them vs the tasks they are supposed to accomplish, all under "AI". 13821284
Reposted by Eddie Antonio SantosEndangered Languages Project @e-l-p.bsky.social · 22/12/2025Language champions from 18 countries will be joining the 2026 Ready to Revitalize program - and we have just TEN DAYS left to raise funds for their grassroots #languagerevitalization projects! 🌱 Will you support the revitalization of Indigenous and endangered languages around the world?givebutter.comRaise to RevitalizeSupport Global Indigenous Language Revitalization 057
Reposted by Eddie Antonio Santosfig @sunteafig.bsky.social · 09/12/2025Yes. Yes. Yes. Students use LLMs because they have never, not just now, had enough time & space for learning. And they are frequently missing knowledge about how the various resources work. We just skipped right over that to "they're going to use it anyway so we have to integrate it into pedagogy." 214033
Reposted by Eddie Antonio SantosBridge Chase @bchasemadethis.bsky.social · 03/12/2025This season, I am fundraising for a program called Ready to Revitalize: a unique course run through @e-l-p.bsky.social for grassroots #language champions around the world. It only costs $500 to sponsor a participant. Help us support language revitalization! #langsky givebutter.com/elp2025/chas...givebutter.comRaise to RevitalizeSupport Global Indigenous Language Revitalization 043
Reposted by Eddie Antonio SantosAnthony Moser @anthonymoser.com · 19/11/2025LLMs destroy context. You used to be able to tell a crank website from a good source partly by its style. But now we have a tool that puts careful research, incoherent garbage, and intentional shitposting through a kind of grammatical instagram filter, presenting them all with the same credibility 121445
Reposted by Eddie Antonio SantosKirill Osenkov @kirillosenkov.com · 15/11/2025engineers are basically professional understanders of complex systems. When someone I respect sends a PR I trust that it reflects a reasonably complete model of the system in their head. AI doesn't have it, it has an educated guess with varying levels of plausibility. 19711
Reposted by Eddie Antonio SantosKatie Croonenberghs Art @kamakru.bsky.social · 27/10/2025Turn off Gemini in Google! It's been turned on by default. Go to your Drive, click the gear icon, click settings, go to "manage apps" and uncheck that nasty "use by default" box. www.zdnet.com/article/how-...zdnet.comHow to turn off Gemini in your Gmail, Docs, Photos, and more - it's easy to opt outIt's a little hidden, but there is a way to remove Gemini from your favorite Google services. 4997697
Reposted by Eddie Antonio SantosAonghus Ó hAlmhain @aonghusoha.bsky.social · 13/10/2025Bréaga a chreideann ríomhchláraitheoirí faoi teangacha, le samplaí. Ó @lexiconista.bsky.social www.lexiconista.com/falsehoods-a...lexiconista.comFalsehoods programmers believe about languagesThis is what we have to put up with in the software localisation industry. 2115
Reposted by Eddie Antonio SantosGreg Egan @gregegansf.bsky.social · 16/10/2025“Well-meaning” people are using machine translation to write Wikipedia articles in languages that they don’t speak themselves, accelerating the degeneration in quality of the web corpus for several languages with relatively few native speakers. www.technologyreview.com/2025/09/25/1...technologyreview.comHow AI and Wikipedia have sent vulnerable languages into a doom spiralMachine translators have made it easier than ever to create error-plagued Wikipedia articles in obscure languages. What happens when AI models get trained on junk pages? 6387228
Reposted by Eddie Antonio SantosRobert Bickers @bickers.bsky.social · 09/10/2025I've written about the life and just some of the achievements of my former classmate Andy West aka 魏安, Ando, and BabelStone @babelstone.co.uk www.theguardian.com/technology/2...theguardian.comAndy West obituaryOther lives: Expert in encoding scripts for languages 12211
Reposted by Eddie Antonio SantosLarry Hunter @proflhunter.bsky.social · 24/09/2025Meta-analysis of 262 articles showing robust links between the vitality of Indigenous languages and health or wellness. The mental health benefits are particularly notable. Language programs help Indigenous communities thrive. www.sciencedirect.com/science/arti...sciencedirect.comLanguage improves health and wellbeing in Indigenous communities: A scoping reviewIndigenous languages in Australia, Canada, New Zealand and the United States are endangered due to colonial policies which promote English language do… 03717
Reposted by Eddie Antonio SantosLauren Gawne @superlinguo.bsky.social · 10/09/2025Happy Unicode 17.0 Day! Includes seven new emoji that are rolling out on your phones and computers in the coming year: treasure chest, trombone, avalanche, big foot, bulging face, fight cloud, orca @jenniferdaniel.bsky.social has the lowdown: jenniferdaniel.substack.com/p/tomorrows-... 36822
Reposted by Eddie Antonio SantosOlivia Guest · Ολίβια Γκεστ @olivia.science · 06/09/2025Finally! 🤩 Our position piece: Against the Uncritical Adoption of 'AI' Technologies in Academia: doi.org/10.5281/zeno... We unpick the tech industry’s marketing, hype, & harm; and we argue for safeguarding higher education, critical thinking, expertise, academic freedom, & scientific integrity. 1/n 11341512055
Eddie Antonio Santos @eddieantonio.ca · 05/09/2025Introduced my non-CS majors to floating point numbers yesterday by showing off .1 + .2, then showing them 0.30000000000000004.com and I think I lost a lot of them. But at the same time, I hate telling that lie that "floats are just numbers with a decimal". Does anybody have experience with this?0.30000000000000004.comFloating Point Math 120
Reposted by Eddie Antonio SantosStephen Chrisomalis @schrisomalis.bsky.social · 05/09/2025Having gotten through the first two weeks of classes, it's hard to convey just how radical my pedagogy has become, especially post-pandemic. Sure, AI and learning outcomes and other intrusions, but more just rethinking some of the cop shit that goes on in classrooms (including mine, in the past). 181
Reposted by Eddie Antonio SantosTechnology Connections @techconnectify.bsky.social · 02/09/2025*cracks knuckles* okay so you know how the media has a problem with taking snippets of scientific studies out of context and writing articles which mislead the public? now imagine a computer program is doing that when you ask it questions about scientific literature but also it makes stuff up. 804086734
Reposted by Eddie Antonio SantosYoav Goldberg @yoavgo.bsky.social · 27/08/2025When reading AI reasoning text (aka CoT), we (humans) form a narrative about the underlying computation process, which we take as a transparent explanation of model behavior. But what if our narratives are wrong? We measure that and find it usually is. Now on arXiv: arxiv.org/abs/2508.16599arxiv.orgHumans Perceive Wrong Narratives from AI Reasoning TextsA new generation of AI models generates step-by-step reasoning text before producing an answer. This text appears to offer a human-readable window into their computation process, and is increasingly r... 48623
Reposted by Eddie Antonio SantosBen Williamson @benpatrickwill.bsky.social · 22/08/2025Let's be clear: all the AI in edtech stuff (Canvas, Blackboard, Grammarly, Duolingo, OpenAI, the rest) is all about the "lock-in" - locking in students as subscribers or institutions via long-term license agreements for value generation. Is there good pedagogic evidence for it? No. 8686233
Reposted by Eddie Antonio SantosEndangered Languages Project @e-l-p.bsky.social · 09/05/2025Did you know that ELP offers a FREE Language Revitalization Mentors program? Anyone who’s working to revitalize their language, wanting to get started, or curious about language revitalization, can talk one-on-one with experienced language revitalizers and scholars. endangeredlanguages.com/mentors 11611
Reposted by Eddie Antonio SantosEvan Greer @evangreer.bsky.social · 02/04/2025"Age verification" laws are actually "upload your ID or get your face scanned to access every website, ending anonymity and associating your identity with everything you do online" laws and if more people understood that they would not be down for this authoritarian nonsense 3573810715040
Reposted by Eddie Antonio SantosFlorent Moncomble @fmoncomble.tract-linguistes.org · 20/08/2025When you struggle to find the equivalent of a word or phrase in another language, do you sometimes find yourself opening Wikipedia, searching for it then navigating to the "languages" section to get its title in the other language? Well, I do, and I got tired of it, hence: #linguistics #xl8 #xl9ncorpustools.prendrelangue.frWikiXl8Look up Wikipedia pages across languages 810542
Reposted by Eddie Antonio SantosFaine Greenwood @faineg.com · 08/08/2025It really does feel like a lot of formerly reasonable people have been infected by some sort of terrifying brain parasite when it comes to unthinking and total acceptance of using AI tools 572151381
Reposted by Eddie Antonio SantosEmily M. Bender @emilymbender.bsky.social · 06/08/2025Lit review query for #NLP folks: What are your favorite works reflecting on evaluation, especially of (UD) parsing and (FST or other) morphological analyzers? 122
Reposted by Eddie Antonio SantosMike Cook @mtrc.bsky.social · 11/07/2025Pretty interesting study! This result is going to get misinterpreted in a bunch of different ways over the next few days, so I thought I'd share a few thoughts with you now to contextualise some of the stuff in the paper. 415859
Reposted by Eddie Antonio SantosEndangered Languages Project @e-l-p.bsky.social · 30/06/2025Are you able to use your language online in all the ways you want to? Do you have keyboards you can easily type your language in? If the answer is "no", ELP Governance Council member Dr. Craig Cornelius can help for FREE! Contact the Keyboard Creation Project:translationcommons.orgKeyboard Creation Project - Translation CommonsKeyboard Creation Project Translation Commons has assembled a group of volunteers from industry and academia to assist Indigenous communities to digitize their language. One of the first necessary ste... 074
Reposted by Eddie Antonio SantosChanda Prescod-Weinstein 🌌 @chanda.blacksky.app · 20/06/2025And look, science communication is hard. I have a monthly column and have written 2.5 books. Every piece involves agonizing over how to use accessible language while not misrepresenting the science. But the AI convo has been sloppy and there are corporate actors want the public to be miseducated 413619
Reposted by Eddie Antonio SantosProf. Nicole Holliday @mixedlinguist.bsky.social · 17/06/2025A major danger of LLMs is that humans are SO predisposed to attribute knowledge to any entity that uses natural language fluently. We cannot imagine that a machine that outputs natural-seeming speech/text doesn't have cognition. Brilliantly articulated by @emilymbender.bsky.social et al. (2021). 341351481
![Text generated by an LM is not grounded in communicative intent, any model of the world, or any model of the reader’s state of mind. It can’t have been, because the training data never included sharing thoughts with a listener, nor does the machine have the ability to do that. This can seem counter-intuitive given the increasingly fluent qualities of automatically generated text, but we have to account for the fact that our perception of natural language text, regardless of how it was generated, is mediated by our own linguistic competence and our predisposition to interpret communicative acts as conveying coherent meaning and intent, whether or not they do [89, 140].”](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:uj2g526vwfe3qnpuskjexeyp/bafkreiappchidtcrrumnegykz4uq6chvznuu6fg4pbutw5y7yjaab2iczu)
