Apertus 1.5 is out! 🚀 ✅ Multimodal capabilities ✅ Stronger reasoning ✅ A roadmap of regular releases Read more: apertus-ai.org/news Models: huggingface.co/swiss-ai @cscsch.bsky.social @eth-ai-center.bsky.social @icepfl.bsky.social @mjaggi.bsky.social @abosselut.bsky.social

Who is writing about rescuing tech-for-good from technosolutionism? I’m looking for insights on how we use the wise insights from the anti-technosolutionist analysis while maintaining a space for better informed enthusiastic tech-for-good communities and keep the space for nerd voluntarism.

I’m trying to train colleagues to say “how will we evaluate benefits of this project” not “how will we measure benefits of this project” - rigorous formal qualitative approaches are much better than unnecessarily quant ‘measurement’.

It’s lovely to have someone who speaks like I do^W used to before I changed my accent to fit in running the country.

Squee I’m in a disinformation seminar with David Buckingham and I’m reading “The End of Information” right now.

One of the things I found useful about last nights Kirstein lecture by Katie Winkle was the idea of adding metadata about what part of the futures cone you’re working in to each project. This creates space for designing for preferable futures alongside work that is useful in the plausible.

Excited - as every year- for this years’ Kirstein Lecture: Katie Winkle on “Designing Robots. Designing Society: Opportunity or Automating Inequality?”

If I had a penny for every time I’ve thought “the computer is wrong… oh wait no it’s my fault”, I’d have… about the same amount of money I do now, because that’s work I’ve been paid for.

At the UCL festival of digital research. Hearing from our data centre partner, Datavita, about the fully renewable energy and very low water use of the Lanarkshire AI growth zone. UCL is at the heart of the UK’s London AI startup cluster and is proud to partner with Scotland for clean HPC and AI.

When I send you ambiguous instructions or an opaque email thread that you have to carefully interpret to find out what I want you to do, I am paying you the complement of believing you are smart enough to solve the puzzle. If you get a careful step by step task breakdown - feel insulted.

Fear of losing a job you hate is a perversity emerging from the unsuitability of our economic system to respond to technological change. With confidence in education and protection while we retrain, change could be welcomed. If change is indignity and disempowerment we are all anti-tech.

At a managementy away day thing. Discussion assumes that staff capability is fixed and limited. Arranging resources well. Which is odd for an education provider. Human capacity is not limited: we can and should help each other learn and grow. *That’s* how you do more with less. By becoming more.

Presented the draft of the LLM assisted coding policy for ARC to our team. It went actually fairly well considering.

Next time I come to Swindon for a UKRI meeting I might make a T-shirt with a picture of four buckets on it. 🪣🪣🪣🪣

In the grim darkness of the third millennium, playing with fascist superhuman toy soldiers is memetic defence. @warhammer40k.bsky.social

Patrick Wyman@patrickwyman.bsky.social · 2mo ago

You know who isn't getting radicalized? Guys who do ships in a bottle. Embrace model-building and avoid the pitfalls of right-wing extremism

You know who isn't getting radicalized? Guys who do ships in a bottle. Embrace model-building and avoid the pitfalls of right-wing extremism

lauren@lauren.rotatingsandwiches.com · 6mo ago

every time i see a white guy in their fifties radicalized by the twitter algorithm into racism and extremism i think about how it's a total failure of the youtube algorithm to instead radicalize him into home brewing, elaborate A/V setups, or DIY solar

Has anyone built a reduced capacity programming assistant ai (system prompt to limit capability) for pedagogic purposes? “Build me X algorithm” would have it decline, but step by step instructions which showed the algo was understood would each be implemented, assisting with syntax and cruft.

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!

Agile Project Management Training for dRTPs

Learn practical Agile project management skills to improve how you plan, deliver, and manage research projects in digital research teams.

buff.ly

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

Slightly increase a band. Counting buzzards.

Kate Bevan@katebevan.com · 3mo ago

Slightly increase a band Simon and Garfunkel and auntie

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

Henry Farrell@himself.bsky.social · 3mo ago

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.

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.

AI as Social Technology

knightcolumbia.org

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...

Lecturer in Applied Mathematics (Quantitative Modelling) at King's College London

Recruiting now: Lecturer in Applied Mathematics (Quantitative Modelling) on jobs.ac.uk. Click for details and explore more academic job opportunities on the top job board

jobs.ac.uk

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.