Nice to get an Ada double header appearance today‘s The AI Shift - thanks @jburnmurdoch.ft.com and @sarahoconnorft.ft.com www.ft.com/content/852c...

Today we publish 📏𝐌𝐄𝐀𝐒𝐔𝐑𝐈𝐍𝐆 𝐔𝐏📏, a new @adalovelaceinst.bsky.social briefing on AI and public sector productivity. We find: ⚖️The link between measurement and real-world outcomes is unclear 📐Workers and citizens are absent from research design ⏲️Industry shapes what gets researched Read more:

Two weeks today! Sign up for the latest @transformgovtalks.bsky.social with: @laragroves.bsky.social, @adalovelaceinst.bsky.social on www.adalovelaceinstitute.org/report/scrib... Owen from the ICO on ico.org.uk/about-the-ic... In person (central London) or online 👇

'it’s interesting to compare how hard it was to differentiate a web browser with how hard it is today to differentiate a chatbot' Fab piece from @thebenedictevans.bsky.social asking hard Qs of OpenAI when its tech is similar to other labs and has no network effects www.ben-evans.com/benedictevan...

not to be a massive bore but in these times of A) received inter-gender polarisation and b) AI sloppified media/textual offerings it does feel insane to keep chiding men for... reading books www.newyorker.com/culture/the-...

🎄All we want for Christmas is for AI regulation to work for people and society. Our new polling shows the UK public are concerned that tech companies’ needs are being prioritised over their own when it comes to AI regulation. Read the full findings ⬇️ www.adalovelaceinstitute.org/policy-brief...

academia friends, anyone got experience using Open Review as a Programme Chair? help needed please!

In the intro issue of New Left Review, Stuart Hall noted, “The humanist strengths of socialism—which are the foundations for a genuinely popular socialist movement—must be developed in cultural and social terms, as well as in economic and political.“ Zohran Kwame Mamdani‘s campaign got the point!

NVIDIA and OpenAi: Concerns that their “increasingly complex and interconnected web of business transactions is artificially propping up the trillion-dollar AI boom.“ @bloomberg.com $NVDA 👀 www.bloomberg.com/news/feature...

Either real-time bidding data or app SDK data is being bought and used by ICE to seize people from the streets. If you didn’t believe that the adtech stack is a human rights concern, I’m not sure what more evidence you need. www.404media.co/ice-to-buy-t...

It would be good if the only political pushback against Nigel Farage's plans to mass deport hundreds of thousands of our neighbours, friends, family and colleagues didn't just come from a handful of backbench MPs and one regional mayor

The headlines — £31bn in investment, new jobs, faster medical treatments — sound almost too good to be true. What are we giving up in return? Me in @theguardian.com on today's US-UK Tech Deal: www.theguardian.com/commentisfre...

Met Police’s policy on when they use oppressive facial recognition technology is made pretty clear here… ✅ Black community celebration / event ❌ White supremacist rally

Very glad to see that someone is doing the important work of aligning AI alignment. alignmentalignment.ai

hey im putting something together for my boss for his birthday. yeah can you write a letter or better yet an insane poem or horrifying drawing about the unspeakable crimes you all have committed together

Radical right accommodation really does not work. New paper out with this exceptionally talented team @katharinalawall.bsky.social @robjohns75.bsky.social @drjennings.bsky.social @sarahobolt.bsky.social @zachdickson.bsky.social @danjdevine.bsky.social & @jack-bailey.co.uk doi.org/10.31235/osf...

Please apply for / pass on this vacancy with @tuc.org.uk! You'll be working with the brilliant @adamc-c.bsky.social and with us @connectedbydata.org to ensure workers' voices are an effective countervailing force that shape the adoption of AI. tuc.current-vacancies.com/Jobs/Advert/...

1. Incitement to violence is a serious offence, carrying a sentence of up to six months in prison. 2. What's it got to do with the health secretary? 3. I'm sure if I posted that Wes Streeting needed a good kicking, he would be very eager to send the cops round. www.theguardian.com/politics/202...

Keir Starmer: cut me and I bleed union j

the level of disorder on show in the UK right now in terms of kneejerk conspirational thinking against every possible intervention or source of authority is quite the concern!!

Spending on AI data centers is now so great that it’s crowding out investment in housing and other critical parts of the real economy, through higher energy costs and by propping up interest rates, argues The Economist economist.com/finance-and-...

Far right demagoguery - thanks to crucial, repeated validation from the political mainstream - has painted a target on the back of every man of colour in this country. Everyone knows what is likely to happen next.

Interesting work from Turing, adding to the growing evidence that significant performance gains — and possibly real-world utility — can still be gleaned from smaller and previous-generation foundation models www.turing.ac.uk/blog/why-we-...

resonates with what we're finding in forthcoming Ada research on transcription tools in social work. these tools are being rolled out without systematic evaluation, often with a singular theory of change -'productivity gains' - which risks obscuring other important impacts on work & care

Last month, Anas al-Sharif sounded the alarm on starvation in Gaza. Israel accused him of terrorism. "We warned this felt like a precursor to justify assassination," said Jodie Ginsberg of the Committee to Protect Journalists. Today he was killed in a targeted strike on a tent housing reporters.

Arresting 400+ people for the "terrorist" offence of protesting against mass murder while large segments of the right are openly encouraging riots against immigrants is a tale of a very unhappy country and government.

'It is less that AI is the cause of degradation in reading and thinking, and more that the creation of a culture that views knowledge primarily in an instrumental manner has made it easier to misuse AI.'

fascinating paper on 'character training' LLMs - models fine-tuned to be 'warmer' are significantly more likely to promote conspiracy theories or validate incorrect facts, particularly when users express sadness 🤯 arxiv.org/abs/2507.21919

Stand by this: www.politico.com/newsletters/...

The Online Safety Act assumes that all tech, social networks, etc, are run by big tech, raking in profit from illegal advertising. Community-run and funded networks/sites — real non-Big Tech alternatives — are left in the dirt. I have never seen OSA proponents consider their needs or their value.

This, and it doesn't work. Now, it is possible to address the harms of mass engagement-driven platforms. But doing so requires pushing back on the core biz models of the platforms, not tinkering on the edges w surveillant 'solutions' that increase surveillance w/o addressing the root cause.

Today (w/ @ox.ac.uk @stanford @MIT @LSE) we’re sharing the results of the largest AI persuasion experiments to date: 76k participants, 19  LLMs, 707 political issues. We examine “levers” of AI persuasion: model scale, post-training, prompting, personalization, & more!  🧵:

'Today's career incentives & social norms prize solutions to scientific problems over human understanding [...] If we use AI to short circuit this process of understanding, that is like using a forklift at the gym. You can lift heavier weights with it, sure, but that's not why you go to the gym.'

just one more supply side economics please bro trust me this time it will fix everything. just some more trickle for my economics please www.theguardian.com/business/202...

It's out! New @adalovelaceinst.bsky.social and @cdt.org report exploring what 'calling in the professionals' might achieve for AI assurance www.adalovelaceinstitute.org/report/going...

Ahead of the expected proscription of Palestine Action - the first time a protest group has been legally redefined as terrorist - I went through the Hansard of the 2000 Act they're being banned under. Turns out MPs were assured the law wouldn't be applied this way. Thread 🧵🔽

first £8 pint, harrowing threshold breached. screaming crying throwing up

New paper out today, accepted at #FAccT2025, led by Jake Stein and me: "Not Even Nice Work If You Can Get It; A Longitudinal Study of Uber's Algorithmic Pay and Pricing" arxiv.org/abs/2506.15278 Over the last 3 years, we've been working with @workerinfox.bsky.social to audit Uber's algorithms. 1/

I too hope to have novel insights next year, at the same small small price of $10billion

"If human goals and preferences become increasingly co-constructed through interaction with AI systems[...] then AI safety requires paying as much attention to the psychology of human-AI relationships as the wider societal factors and technical methods of alignment" arxiv.org/pdf/2502.02528

The Guardian story does not mention that the EHRC's Akua Reindorf has previously advised and represented Sex Matters and been instructed by LGB Alliance. Seems relevant.

"Governments should view companies that drape their products in a democratic banner with great scepticism. When Washington decides to weaponise technology access, partner countries will discover their “sovereign” infrastructure is anything but." www.ft.com/content/4b2a...

not to be histrionic but I kind of feel like the confluence of Trumpism, algorithmic social media, and LLMs have placed language and the fundamental proposition that words have meaning under unprecedented strain and it’s freaking me out more than a bit

'a pluralistic lens on safety creates room for moral disagreement: societies may never converge on extinction probabilities, but they can still agree that finance models should resist manipulation and autonomous vehicles should obey speed limits.' www.techpolicy.press/from-doomsda...

There's a lot of chatter around the proposal being inserted into a budget bill that would put a moratorium on any AI legislation being passed by the states for the next 10 years. I thought I'd say a bit about why this is an absolutely disastrous move. www.404media.co/republicans-... 1/n

an AMERICAN?!!?!

Looking forward to speaking at this event next month with @dianecoyle1859.bsky.social @aboxiwu.bsky.social @alturobov.bsky.social! I'll be talking about my work with @jaivipra.bsky.social on how public compute—state provision of computing resources—can be used to reshape the AI sector. Sign up 👇

enjoying this so far. less excoriating than I was expecting, after the furore around its publication - but very entertaining ‘insanity confirmation’ material to file for Zuckerberg/Sandberg