About to start my tenth Hackney Half. Running for WWF. www.justgiving.com/fundraising/... (12th half over all races.)

Would you like some free #agile project management for your team? Adina from our RSE team has received funding from the #DisCouRSE network+ to deliver hands on training. Applications are now open!

“Which Vampire: The Masquerade clan is each Eurovision act?” Is a fun game.

Slightly increase a band. Counting buzzards.

Excellent analysis of how to think about AI as a social technology (rather than as an individual intelligent agent)

knightcolumbia.org/content/ai-a... Now up at @knightcolumbia.org - Cosma Shalizi and myself on "AI as Social Technology." It says that AI models are really a technology of social mediation that allow humans to connect with thoughts of other humans. Short-ish thread with major points follows.

When researching for my inaugural, I needed to find my old PhD thesis. I discovered it was not open access. The default is closed unless you sign a form. It is now open access www.repository.cam.ac.uk/items/41dc86... If you like open science, you might enjoy investigating if you need to do this.

Amount of bureaucracy is set by tolerance, not by need. There will always be a new process and advocates for its worth. Adoption is governed by capacity. Technology speeds each piece of waste up. Making space for more. AI adopters looking to save need consciously limit new process adoption.

I know the fire and hope has gone out of the centre left, but who would like to help me imagine a world where AI is a cooperative capability? A national inference service for hardware, with open source models running on it? A national training effort for at least one public open source model?

What is the word for “I was hoping for a ❤️ but I only got a 👍”？

What is the legal copyright status of transclusion? If I do <img href=“http://somone_elses_website/a.png> on my page, have I copied it?

King's College London are seeking a Lecturer in Applied Mathematics (Quantitative Modelling) to "develop an independent research programme in applied mathematics, mathematical modelling, or a closely related area" www.jobs.ac.uk/job/DRG116/l...

Sycophancy… alignment… overfitting on what feedback pleases each user summed over users vs what the system needs as a whole. It rhymes with zeroth law vs first and second laws in Asimov. Obvious but I just connected it.

Everyone's building AI science agents. The claims are extraordinary. But when we test whether these systems can actually do science, recent top ones still fail challenges that human scientists can solve the majority of the time. 🧵

In “wake, watch, wonder” by Robert J Sawyer, one of my favourite optimistic AI novel series, there’s a morning where everyone wakes up and all the spam has gone, and that’s when the world learns of the existence of the benevolent AI. Reminded of this by Glasswing, but vulnerabilities not spam!

I’m reflecting a lot on the perils of overthinking things.

I’ve been looking for a way to explain to folks what is happening, and why “the AI” isn’t always the big bad all the time. Sure mega tech is trying to ruin it, but real wholesome work is happening in AI. What Jay and team, along with many others are building for us will make our network stronger.

an app is just a lens over the open network of data, and you can personalize what your lens looks like

This is incredibly relevant for many of us in academia

"The AI Grad Student": A Harvard professor describes working with Claude. Early on, he describes misconduct that would cause any student to be terminated: "It faked results, hoping I wouldn't notice." But he ends the essay with "Now I'm doing 100% of my research with LLMs". Am I losing my mind?

Arriving in Edinburgh for a workshop on research software engineering and generative AI, funded via Schmidt Sciences. A real chance to begin to set the practices and safeguards for our profession to use this technology with joy, wisely and well. Which is not how many use it now!

Our CEO @parlysue.bsky.social will be speaking at #Anthropy26 next week at the Eden Project in Cornwall. 📅 25 March at 5:30pm Speaking alongside fellow leaders on community-led sustainability Tickets: anthropy.uk/national-gat... @anthropyuk.bsky.social #Anthropy #Anthropy26 #UnitedFuture

ARCHER2, the UK national supercomputing service, has delivered an eightfold return on investment to the economy according to independent analysis. Operated by EPCC, it has enabled more than £4.2bn in benefits to the UK economy since launching in 2021: edin.ac/47JGMHB

Which UK HE has the least obstructive and self-harming professional culture? I’m reminded (via supplier relationships) how different corporate cultures in the same sector can be. Is it us? I do hope it’s not us.

🇨🇭 In Ticino, the open Swiss model Apertus powers in-house AI translation for the Cantonal Administration — boosting multilingual services with local, transparent control. A great example of sovereign AI in action bit.ly/4rzU0Og @epfl-ai-center.bsky.social @eth-ai-center.bsky.social

Taking into account distributions of human reaction times and bullet velocities, what is the probability distribution over the percentage of people that would die in a Circular Firijg Squad, as a function of distance between participants?

The cultural observation game at Sadler’s Wells East when West Ham are playing is hilarious. “Are you here for the dance or the football?”…

"Shock! Shock!" is how the "Father of the Analysis of Algorithms" Don Knuth described an interaction with Claude, proving an open problem he worked on for weeks #HPC and #AI compute enables the unthinkable and begins to automate the scientific method and human (?) progress 🚀 Full buff.ly/olgx1sX

OpenSAFELY is open from today! Huge thanks to all who supported this vast collaboration: whole population GP data; in a productive platform; innovative privacy protections; unprecedented support from professions, privacy campaigners; &c Now it's over to users! www.bennett.ox.ac.uk/blog/2026/02...

I made an embarrassing malapropism while talking to an LLM and now I can’t deal with the asocial anxiety. Help me.

Suppose there are fewer entry level coder jobs because these are done by AI, but we still need more advanced engineers. In this world “pedagogical programming” : writing software designed primarily to *teach humans* becomes Huge. This is not yet a model training objective.

Is there a name for the phenomenon whereby management communities are unable to hear scientific advice which is inconvenient? Eg goodheart’s law to someone who invents a poorly designed new metric …missing not at random to a person who wants to use survey data with low response to make a decision.

I’m running the Hackney half for charity again, this time WWF. www.justgiving.com/fundraising/...

Fiction: “I have instructed my AI to remove all unnecessary politesse and verbiage from all incoming messages. There will thus be no manipulative value to you in not getting to the point. Just say what you want without the schmooze.”

Delighted to join today’s UCL Policy Lab Roundtable on AI, the future of Work and the Role of the University. There was a broad panel of participants with distinctive & fascinating perspectives. HE mist be relevant to student needs, but what is ‘relevant’ increasingly is difficult to predict

We have funded 10 ambitious Challenge-led Awards through the @UCL Grand Challenge Data Empowered Societies, University College London www.ucl.ac.uk/grand-challe...

This is a classic format. The last time I saw this, the subject was signal-to-noise ratio. It is interesting to consider that math is the signal, and 'magic' is the noise.

At the SC25 trusted research environments BOF. Sadaf Alam presenting our work on Federated Research Infrastructures through Data Governance Extension (FRIDGE). With @bristoluni.bsky.social @ucl-arc.bsky.social Turing Institute and Cambridge

Come to UK Universities booth at #SC25 to find out more about the UK's #DRI programme and pick up some merchandise!

Extremely impressed by what U Illinois have achieved with in house LLM services using their local HPC investments and open source models and tools. This is what I’ve been advocating for UCL and they’ve proved it can be done.

Please visit us later this week at the UK Universities booth to learn more about our research and activities.

It’s fascinating how hard we had to work to get computers to be bad at maths.

Causing this to occur has been an ambition of mine since 2019: We did it! A joint uk universities booth at Supercomputing! Congrats to the team.

Suppose cost of software does fall 20x because of LLMs. (jury is out, but let’s explore). What does that do? It might make bespoke software for each org competitive with platform solutions. The value of better org-sol fit might be bigger than cost of build. Makes IT orgs builders not buyers.

Arsenal WFC vs Chelsea. #coyg

Uk hpc, what is your contingency planning around US ATC issues for SC25? I’m wondering if I can rebook through Toronto as that would be more robust?

We have people booked to bring broken food processors, favourite jumpers with damaged zips, and ailing plants. Don't miss this opportunity to repair your stuff with the help of 40 technicians and makers. Repair Café | Sustainable UCL - share.google/2qFlWYPHRPLs...

arxiv.org/pdf/2510.22254 “Ten Simple rules for Ai assisted coding in science” is very good. poldracklab.org/10sr_ai_assi... .

I am the cowpox of grumpy professors. Work with me so it’s safe to work with the real ones.

Data is the new sewage.