The mirage of AI deregulation
One of the most interventionist approaches to technology governance in the United States in a generation has cloaked itself in the language of deregulation. In early December 2025, President Donald Tr...
๐งต Trump administration AI policy is widely described as deregulatory. This description is misleading. What's happening is not the absence of governance but its rearrangement--intensive state intervention operating through mechanisms we don't typically call regulation. www.science.org/doi/10.1126/...
15.01.2026 19:23
๐ 592
๐ 321
๐ฌ 9
๐ 37
New paper from team @aial.ie! aial.ie/research/gpa...
EU's AI Act Article 53(1)(d) is an obligation for GPAI model providers to publicly provide a 'summary' on their modelโs training data. The team assessed published summaries along 6 dimensions & found that all big providers failed on all 6.
1/
05.03.2026 18:04
๐ 114
๐ 69
๐ฌ 2
๐ 3
This was a fun and really wide-ranging conversation! Thanks @davidvitale.bsky.social and @cccwarwick.bsky.social for hosting us!
05.03.2026 16:39
๐ 2
๐ 0
๐ฌ 1
๐ 0
Honestly I am surprised this most basic of regulations has taken so long to enact. I cannot understand why regulated professions everywhere, including in Canada, are not pushing for similar.
03.03.2026 19:12
๐ 1
๐ 0
๐ฌ 0
๐ 0
Opinion: OpenAI has shown it cannot be trusted. Canada needs nationalized, public AI
OpenAI, the company behind ChatGPT, lobbied Ottawa for business. All the while it hid its knowledge of Tumbler Ridge shooter
I understand the sovereignty push of Canadaโs tech industry (and other industries), but we need to grapple with a more fundamental question: What kind of tech do we want, sovereign or otherwise?
As Bianca Wylie noted, sovereign bad tech is still bad tech.
www.theglobeandmail.com/business/com...
02.03.2026 12:58
๐ 10
๐ 6
๐ฌ 1
๐ 1
#LunarEclipse FAQ:
Q: Is it safe to look at?
A: Absolutely! Unlike a solar eclipse, there's no UNsafe time to look at a lunar eclipse.
Q: Do I need a telescope?
A: Nope! Best view is with naked eye. Just look up!
Q: Does the eclipse portend terror & doom?
A: Not at all! The news has that covered.
02.03.2026 19:22
๐ 1291
๐ 405
๐ฌ 29
๐ 16
Monochrome photograph featuring two women climbing on a rock face with ropes both wearing long dresses and hats
Rock climbers Lucy Smith and Pauline Rankin of the Ladiesโ Scottish Climbing Club, 1908, Salisbury Crags, Scotlandย #WomensArt
Welcome to #WomensHistoryMonth
01.03.2026 03:53
๐ 743
๐ 161
๐ฌ 11
๐ 16
All books from CMU Press | booksellers.ca | Purchase print books and eBooks online
A network of over 100 bookstores, more than 500,000 print books or eBooks, unparalleled expertise. It is possible to buy local on the Web!
It's been a long time coming. Canadian independent bookstores have banded together to create an online book buying source so you can avoid the behemoth in the room. (Here are the titles published by CMU Press, which I run.)
Spread the word #booksky!
www.booksellers.ca/publisher/cmu-press-18021
27.02.2026 02:23
๐ 401
๐ 198
๐ฌ 10
๐ 27
"Our product kills kids" makes for a hell of an advert.
01.03.2026 12:53
๐ 23
๐ 8
๐ฌ 0
๐ 0
โUCLA neuroscientist Maryanne Wolf has shown that deep reading, defined as sustained immersion in a text, builds the cognitive circuits required for critical analysis, empathy and perspective-taking in ways that skimming, scrolling and short-form video simply cannot.โ
28.02.2026 12:46
๐ 949
๐ 368
๐ฌ 5
๐ 15
This is one of the key reasons that I have a no generative AI policy in each of my classes. I do not want students, many of whom are already stressed and vulnerable, slipping into this predicament. It is also why I design my classes for students to talk to & work with each other to reduce isolation.
28.02.2026 19:27
๐ 98
๐ 24
๐ฌ 3
๐ 1
Well, it wasn't only that the administration was demanding that they be able to use the AI models to autonomously kill people; they also want to use AI to assemble and deploy a comprehensive surveillance database on the behavior of all Americans. What's to worry about?
28.02.2026 16:45
๐ 113
๐ 17
๐ฌ 6
๐ 1
you think you've seen the whole movie, and then the alt text reveals you haven't even finished act two
27.02.2026 19:02
๐ 3248
๐ 595
๐ฌ 40
๐ 8
this is extremely dangerous.
27.02.2026 20:02
๐ 580
๐ 320
๐ฌ 15
๐ 11
Men studying a dragon
ENDOWED PROFESSORSHIP IN HISTORY OF SCIENCE
My home department at Uppsala University is advertising an endowed professorship in the history of science. This is the best position in the field in Sweden, and probably in all of Scandinavia. Apply before 30 April 2026. www.uu.se/en/about-uu/...
27.02.2026 19:20
๐ 179
๐ 147
๐ฌ 2
๐ 6
๐๐๐
27.02.2026 11:22
๐ 1
๐ 0
๐ฌ 0
๐ 0
โIn the end, you feel blankโ: Indiaโs female workers watching hours of abusive content to train AI
Women in rural communities describe trauma of moderating violent and pornographic content for global tech companies
when people talk about AI used in moderation of social media posts, you must remember there is always the human-in-the-loop, often a woman, often in the Global South, always dehumanised and forgotten
1/n
www.theguardian.com/global-devel...
24.02.2026 16:33
๐ 548
๐ 348
๐ฌ 3
๐ 23
The layers of a floppy disk are exploded so that each is visible.
schematic illustration of a cathode ray tube, with each element labeled.
I often find myself wanting to explain to students how things like floppy disks work. Great series of explainers with really helpful illustrations. Just gotta remember where this is next time I need it! www.makingsoftware.com
25.02.2026 16:40
๐ 58
๐ 14
๐ฌ 6
๐ 5
I've shared similar thoughts with reporters this week. The "reporting to police" question overshadows other tightly woven, vital issues here. Focusing on OpenAI also conveniently redirects attention *away* from Canadian govts (fed and prov) when they have many creative regulatory options at hand.
26.02.2026 17:41
๐ 3
๐ 2
๐ฌ 0
๐ 0
OpenAI's ban of Canada school shooting suspect's account raises scrutiny of other online activity
OpenAI's admission that it banned the ChatGPT account of mass shooting suspect Jesse Van Rootselaar months before the 18-year-old allegedly killed eight people and herself is drawing scrutiny to her p...
I was asked for thoughts on Open AI / Tumbler Ridge and sent 4-5 paragraphs of comment. I think the current discourse around whether they should have reported to the police, even though I agree 100% w/ negative sentiment twds AI cos, puts us on a dangerous precipice in terms of human rights. ๐งต 1/??
25.02.2026 20:48
๐ 22
๐ 12
๐ฌ 1
๐ 2
Generative AI is not just a tool for learning. It shapes how students think
Framing this technology as a helpful โtoolโ disguises how much it guides every element of critical thinking that academics seek to cultivate, says James Garvey
โInsisting on technical precision when it matters can help us resist the easy idea that AI is a helper, ready to assist. Instead we must keep our focus on the uncomfortable thought that these systems partly and covertly structure what we say, think and learn.โ
24.02.2026 03:10
๐ 95
๐ 42
๐ฌ 3
๐ 2
This App Warns You if Someone Is Wearing Smart Glasses Nearby
The creator of Nearby Glasses made the app after reading 404 Media's coverage of how people are using Meta's Ray-Bans smartglasses to film people without their knowledge or consent. โI consider it to...
The creator of Nearby Glasses made the app after reading 404 Media's coverage of how people are using Meta's Ray-Bans smartglasses to film people without their knowledge or consent. โI consider it to be a tiny part of resistance against surveillance tech.โ
24.02.2026 16:15
๐ 1096
๐ 503
๐ฌ 17
๐ 35
It is really amazing to see something that you helped create and put out into the world, take on a life of its own. There are at least four independently organized upcoming micro-consultations for the #PCAI! Linking them below, if you're in the area, want to join, or as examples to follow:
23.02.2026 22:18
๐ 16
๐ 7
๐ฌ 1
๐ 0
>they're going to vibecode cobol
so, who's stocking up on canned food and shotguns?
23.02.2026 21:34
๐ 1261
๐ 249
๐ฌ 53
๐ 33
Ontario's Curam fiasco was the subject of a provincial audit and even an external report by KPMG. But somehow Curam has failed forward into a new federal contract, putting more vulnerable people's wellbeing on the line. The digital dystopia continues.
24.02.2026 01:32
๐ 1
๐ 0
๐ฌ 0
๐ 0
Un syndicat craint que le logiciel Cรบram soit le prochain Phรฉnix
85ย 000 demandes de prestations sont en retard de traitement.
What's old is new again: My jaw dropped when I saw that Curam, the system that wreaked havoc in Ontario's social assistance program *in 2014* is now doing the same in Canada's old age security program. 85 000 benefits requests delayed as of late Jan! ici.radio-canada.ca/nouvelle/223...
24.02.2026 01:28
๐ 0
๐ 0
๐ฌ 1
๐ 0
This is a good thread & I hope people reflect on how it relates to what happened in 2020-2022, and the ongoing pandemicโhow most people, at the Democratsโ urging, decided to throw anyone who might need airborne virus mitigations in the trash๐ฎ. And in doing so helped our slide into more brutality:
23.02.2026 02:47
๐ 274
๐ 99
๐ฌ 5
๐ 5
This is the most incredible footage of blue whales Iโve ever seen
22.02.2026 08:55
๐ 6913
๐ 2264
๐ฌ 98
๐ 417