Blair is wrong about the US under Trump. It isn't seeking a stronger European partner - it's hostile to the EU and international law more generally. Both are vital European interests. institute.global/insights/pol...

First, no, everything we do as software engineers is about avoiding unmaintainable code as much as possible. The work is always there.

I could not love @edzitron.com’s latest newsletter edition more. We should shout this from the rooftops. www.wheresyoured.at/the-revenge-...

UC Berkeley Law creates new AI policy that bans AI for "conceptualizing, outlining, drafting, revising, translating, or editing any work submitted for credit" or exams because "thinking remains the sine qua non of good lawyering (and of a quality legal education)." That's what I'm talking about.

sometimes I wonder if people in other fields have any clue what we go through as women in computational fields... (graph from: www.npr.org/sections/mon...)

Speedrun xkcd.com/3246/

"AI will survive, but the speculative bubble surrounding it is a sign of a deeper structural problem — the cost of which, when finally realized, will fall most heavily on the working class." interesting analysis theconversation.com/the-ai-bubbl...

REMINDER that applications for Carpentries Instructor Trainer Training are now open until 15 May at midnight Anywhere on Earth! Apply today!

"Performative productivity" sounds like something an AI article would be about, despite this one not being. I find the notion of token usage or agent sprawl as a proxy for meaningful work highly questionable. www.lbc.co.uk/article/perf...

Our Instructor Trainer Training program provides next-level pedagogical training for individuals seeking to play a leadership role in advancing computational and data skills training at their institutions. Within The Carpentries, they are pivotal to the growth and health of our Instructor community.

The applications for Carpentries Instructor Trainer Training are now open! We will be hosting our Instructor Trainer Training starting in July. The deadline to apply is 15 May at midnight Anywhere on Earth. To apply, please read our latest blog post: carpentries.org/blog/2026/04...

I really have to emphasize that this is how generative ai/llms do *everything*

The trap is simple: AI makes code generation cheap, but it does not make correctness cheap. Once that asymmetry exists, every software engineer using LLMs to generate code is forced into a binary that guarantees madness. (3)

Not once in the 80s, 90s or 00s do I recall a politician urging the public to embrace email, mobile phones, texting, two factor authentication, online banking, air fryers, or to replace all their cassette collection with a CD collection. So forgive me if I smell a rat.

Marathon diplomatic talks that could have been an email

"Rather than letting ourselves be distracted by machines of mediocrity that have been built to capture our preferences and sell them back to us, we should focus our efforts on better understanding and celebrating the astonishing creativity and flexibility of human sense-making." #AI #LLMs

I like how wikipedia uses the infobox for cars here

Is nothing safe from AI ensloppening?

@b-list.org has real good points of the "even if it worked" variety here

There's a lot being shared right now about China's incredible shift towards renewable energy - and rightly so. But this, from the UK, is equally astonishing & encouraging. It's not just that change can happen. It's that it *is* happening. (Via climateactapp.substack.com/p/your-daily...)

We’re introducing our new Partnership program to help organisations build sustainable training capacity. Join one of our info sessions to learn more: 👉 21 April 2026: 16:30 UTC: pretix.carpentries.org/community-ev... 👉 22 April 2026: 09:00 UTC: pretix.carpentries.org/community-ev... #OpenScience

Our impact starts with people. Instructor Trainers help prepare new Carpentries Instructors using evidence-based teaching practices. Those instructors go on to teach workshops that reach hundreds of learners across the world. 🌍 That’s the multiplier effect behind our community. 🌟 Read more:

"The real threat is a slow, comfortable drift toward not understanding what you're doing. Not a dramatic collapse. Not Skynet. Just a generation of researchers who can produce results but can't produce understanding."

Similar, like a toilet is similar to a drinking fountain.

That Claude Code leak proves that "agentic AI" is just the AI-LLM sticky taped to itself dozens of times in dozens of ways in the hope it sort of works. Many of us thought "reasoning" and "tools" were just that. neuromatch.social/@jonny/11632...

I had to check this is real. It's real. The product that this month you *have to opt out of in GitHub* is for "entertainment purposes".

"if the students can use a LLM to do the task it was never really teaching critical thinking to begin with" is a very dumb take, sorry. I can bring a forklift to the gym and bench press more than Arnold. The point is not to move the weights.

Spain has closed its airspace to US planes involved in attacks on Iran. Meanwhile, the UK is letting US bombers land, refuel and re-arm. Complicity is always a choice — and our government has chosen to drag Britain into an illegal war.

Latest painting.. Oil and acrylic on canvas

No one could have foreseen this. www.science.org/doi/10.1126/...

New: Wikipedia has officially banned AI-generated content. After months of heated debate and previous attempts to restrict the use of LLMs, volunteer editors accepted a new policy that prohibits using them to create articles for the online encyclopedia www.404media.co/wikipedia-ba...

The tide turns back 🌈

kind of seems like the people saying data center energy use is overstated were full of shit

"Race science" aside, which is like calling homeopathy "materials science", it really is impossible to separate the eugenics from modern AI capitalism.

Bless the librarians of Aberystwyth University for sending this blog out to their students: "Wait a Minute…What If I Don’t Want to Use AI?" #LibrarySky

This case exceeds mere copyright. Britannica also claims OpenAI violates the Lanham Act (trademark law) by generating hallucinations falsely attributed to the publisher—effectively AI-libeling the complainant. techcrunch.com/2026/03/16/m... #GenAi #copyright

Gripping on to that desk, waiting for the Yummy TumTum Man to bring him his fifteenth Biggest Best Boy Big Mac of the day on the little aeroplane fork wheeeeeee

lol

Staggering just how quickly using LLMs in place of thinking makes you worse at thinking.

i need the “llms are might be conscious” folx to read this

Yet another pathetically destructive and counterintuitive policy by Labour just to appeal to sad little bigots.

I remain fascinated by people who are 100% all-in on using LLMs at their jobs (and telling everyone that they're doing it) and who appear to have never paused to consider "what value am I personally adding in here, at this point, and why do I think my employers won't eventually figure this out"

This is empirically backed btw, even just blocking internet on your phone improves mental health more than anti-depressants.

The House of Lords Digital & Communications Committee just published their report on AI, copyright & the creative industries, and their conclusions could not be clearer. 🧵 1/5

Output 👏 from 👏 generative 👏 AI 👏 is 👏 not 👏 a 👏 creative 👏 product 👏

On the train home, and am next to a group of five 20ish loud multicultural lads drinking beers on the way to the football. They're talking about tarot, Basquiat, starting drama classes, your true creative self, respectful Tinder dates, and reading each other their astrology signs. Nature is healing.