This is. . .an editorial? Or an opinion piece—if so, by whom?
06.03.2026 17:16
👍 0
🔁 0
💬 0
📌 0
This is. . .an editorial? Or an opinion piece—if so, by whom?
Israel, the country famously held to account for respecting international law. Canada, the country that allowed Netanyahu, wanted by the ICC for war crimes, fly through its airspace. Sure.
Labour are the party of transphobes and other bigots.
2. Carney tells us that what he says is shaped by concerns about upsetting Trump. Not transparency with the Canadian people, not upholding principles, not defending international law. Appeasing Trump.
ici.radio-canada.ca/rci/en/news/...
The idea that researchers would sit down for 100+ more interviews in order to then have AI agents summarize and quantify them relies on the assumption that 100+ more people in vulnerable positions will agree to spending more of their time with researchers so as to be interviewed. 2/
Even the nominally ‘centrist’ anti-migrant rhetoric from the likes of Alexander Kustov, dressed up in ‘reason’ and ‘logic’ falls flat (here’s an example from a recent insipid piece of his, framed as countering ‘misinformation’). This is the denial that other ‘ordinary people’ *are humans like you*
14. Publication lag makes Al capability critiques obsolete by the time they come out. Here is a problem that almost nobody in the debate acknowledges: academic and book publication timelines are structurally incompatible with Al's rate of improvement. When someone cites a 2025 paper (initiated in 2024) documenting GPT-4's hallucination rate to argue against using Al in March 2026, they are citing evidence about a system that no longer exists. It is like citing a 2005 study on flip phone imitations to argue against smartphones. That's probably why the new "Al Con" book is so bad-clearly outdated before it even hit shelves.
Benthams substack They then spend a significant amount of time smearing IQ by discussing its supposedly racist origins. This is obviously irrelevant to assessing whether it's a good idea-Planned Parenthood was invented by a eugenicist, but this tells us little about whether it's a good institution. They never discuss the extreme predictive success of IQ, which correlates with a number of important life outcomes. General intelligence is simply the ability to reason well across a number of cognitive domains. Al already has some degree of general intelligence, and its capacities are only going to increase.
To put a very fine point on it: it is extremely bad that Alexander Kustov relies on a racist circle of bloggers when asserting his very negative opinion of Emily Bender and Alex Hanna’s book.
this is a muscle you will have to train because kustov is just being honest about what i suspect many other people are now doing. someone that is lazy enough to do hate click content is lazy enough to have a computer generate said content.
its really three sins at once
1. hate-click post
2. computer-generated
3. computer-generated hate click post is about computer-generated content
I wrote about how the US and Israel carried out an attack on an Iranian girls school on par with the OKC bombing and US media relegated it to a back page story. No stand alone evening news segments, no front page stories, it made A11 in the NYT then everyone moved on.
These are war crimes, violations of the Geneva Conventions.
Increasingly I think we should turn the whole Trump cabinet over to the ICC to be tried at The Hague.
Let’s face it: queer studies—in all its incarnations—has always been an unruly participant in academic spaces. We arrive with blushes and leather strap-ons, coy and provocative. We dare to take our intimate practices as grounds from which to theorize geopolitics.
#drafting
Erm, 1996? Or 1997, can’t remember.
Update: the TCFA is in shambles with over a third of their members resigning in protest. Taking a stand is contagious and effective. Your voice matters. How you use it matters.
Toronto Film Critics Association in Crisis Over Censored Speech share.google/hIZNG9yHhJ1w...
As Rachel points out on her Instagram, notice that the guy at the front has a Totenkopf on his quarter zip.
That's why Maine Democrats should not let anyone displaying Nazi symbols near the United States Senate
Update. Seeing how others have been blocked, I’m not exactly surprised. But I am a bit amused because the racist’s review of Bender & Hanna also included this complaint: “Nearly everything in the book is poorly reasoned or misleading...The authors rarely saw the need to respond to objections.”
Bad news thread. Or, actually reading your sources will always be superior to letting your Claude agent do it for you.
In other words, this sentiment that says genAI agents will revolutionize each and every research project—don't be left behind! don't delay the inevitable!—and so every researcher should be using them starting two years ago, is really misplaced. Knowledge-making thrives by a diversity of methods. 6/
So, a research team with a trusted track record who are not only asking to conduct qualitative interviews but also promising to really sit with the material, re-read it, think about it, be available for further consultation when analysis reaches manuscript format, they have the advantage. 5/
Some people might say yes to all that. But I would suspect more will not. People are already over-surveyed without having a corresponding feeling that the time they contribute to another survey will have any effect, let alone a positive one, on society at large or even just their own daily world. 4/
What would make them agree? (Aside from payment.) That what they say will be automatically transcripted & put through equally automated analysis? That they won't be able to tell whether anything they said made it into any paper? That they can't tell the purpose of this study apart from any other? 3/
The idea that researchers would sit down for 100+ more interviews in order to then have AI agents summarize and quantify them relies on the assumption that 100+ more people in vulnerable positions will agree to spending more of their time with researchers so as to be interviewed. 2/
If this is your response to the comment that qualitative interviews with people who have experienced drug addition are much helped by use of agentive AI, then allow me to point out a few ways in which you might not have thought this response through. 1/
Evan Soloman's plan to "regulate" AI is to go to each company individually and beg them to do the bare minimum to keep Canadians safe.
Why even have laws, this method is so much simpler.
If AI slop is causing the mayor of one of the biggest cities in Canada to lie that his political opponent is the Santa Clause of cocaine then the technology clearly needs to be regulated
Yes, irony galore.
He blocks me for ACTUALLY reading his writing and his sourcing.
But it's us who, <checks notes>, "just need to wake up," need to develop "better norms in all this uncharted territory," it's us who are "not all right."
Unfortunately, as I discovered this morning, that skill issue affects us in another, perhaps more consequential way.
bsky.app/profile/katj...
Whereas anon accounts are the kind that people use for a particular purpose, then throw away. Most abuse I've encountered has come from anon accounts. Some from pseudonymous accounts too, but the point is that someone not putting their legal name on it is not reason enough to dismiss what they say.
Like, I'd be upset if one of you told me that I'd behaved badly about something. I'd definitely not shitpost about it, LOL. I'd take it on board.
We all know that there are very legitimate and extremely obvious reasons for people to protect their lives and livelihoods by being pseudonymous.
Yes, that's how I think about it, too. There are pseudonymous accounts with whom I've interacted for years, where we know a lot about how each other thinks, despite the fact that certain parts of their identity are not available to me. There's accountability that comes with that.