Besides the obvious, horrible, systemic risk that embedding AI into critical systems demonstrates... ...dude, is organisaing your desktop really something you need AI to do? Really?

Higher house prices not only add to household debt but also increase the need to work for longer. We can’t imagine a future of less work while housing costs keep on rising www.theguardian.com/business/202...

Ivan Illich's radical monopoly: You can buy the product from many different vendors, but you must buy it from someone.

For those I've been discussing this with earlier - a new version of the AI/AGI and the social contract paper. Feedback very much appreciated. papers.ssrn.com/sol3/papers....

Thrilled to share a new paper with @stuartmills.bsky.social & @casssunstein.bsky.social. The paper discusses ManipulationDetect (previously NudgeDetect), a tool I launched a few months ago that uses LLMs to highlight 'manipulative' practices online. ideas.repec.org/p/not/notcdx...

New podcast episode: AI Bubble Economics 101! (W/Stuart Mills) A word that is increasingly being discussed is "bubble". But what is a bubble really, are we in an AI bubble now, and if we are: what’s next? Have a listen, share widely! podcasts.apple.com/us/podcast/a...

2025 is coming to end and some people who claim not to be fascists are still associating themselves with fascist outlets like X. I claim, perhaps charitably, that it's because they don't understand how the internet works. Either way, this needs to change. berjon.com/fascintern-m...

Interesting! We are getting better at doing the wrong things! Drucker would have a few few things to say.... "There is nothing quite so useless, as doing with great efficiency, something that should not be done at all."

I have a new paper in the BMJ Open. We find 71% of GP surgeries do not follow NHS guidance around patient registration. The guidance says patients do not need documents like photo ID, but many surgeries demand them. This will reduce people's access to care. bmjopen.bmj.com/content/15/1...

"One feature of capitalism is its ability to create work for people to do. A world without work is incompatible with capitalism," says me in this piece on AI and job loss - heartening to see a link to my book here www.dazeddigital.com/life-culture...

Something I'm working on: 1) On average, only 9% of US firms were using any form of AI to produce products in September 2) US firms are consistently overestimate the rates of AI adoption (dotted line) 3) Expectations are maybe(?) starting to outpace reality...

This is precisely it. We are living in an era of commoditised conflict, where local conflicts can now have global effects, conflict itself can be more ambiguous, and expensive prestige weapons are increasingly for show.

Nice to be quoted alongside stalwarts like Gary Marcus and Carl Benedikt Frey in this article about the AI bubble. I'm not sure I think AI companies *should* raise prices, only that I think they will *have* to. www.dw.com/en/will-the-...

This reads more like Rogoff advertising his books than him making some profound comment about the state of economics discourse.

Data suggest generative AI is being used for backroom firm activities, rather than on the frontline. Adoption rates imply low productivity returns, too, which is also consistent with backroom rather than frontline applications. Some thoughts from my ongoing research: siu2lh.com/articles/wha...

When people talk about AI boosting productivity, do they ever account for the time lost now that every website on the internet has to have an 'are you human?' check which depends entirely on whether Cloudlfare decided to switch their servers on today?

The US military took 15 years to restructure itself around IT. When they did, they followed Soviet ideas. The Soviets lacked IT, so needed to predict how the US would use IT, to learn how to counter them. Jumping head-first into AI means you're likely to get concussion. siu2lh.com/articles/how...

Right now, a smart government would be massively investing in publicly owned renewables, supported by commitments to domestic manufacturing and nationalised steel production. They'd be talking about driving the price of electricity in the UK to zero, and labelling anti net zero an energy tax.

Our article on the politics of AI prompts is out now @econsocjournal.bsky.social www.tandfonline.com/doi/full/10.... 1/n

Is generative AI profitable? Probably not. Would businesses pay more for it? Also no. Is it politically useful? Yes. Would politicians pay a lot for it? Also yes. New paper with @richardrwhittle.bsky.social papers.ssrn.com/sol3/papers....

This would be an enormous mistake for the UK and an error of judgement which will only compound as climate change bites. Gross negligence on the part of our politicians to bet the house on this stuff. Selling out our future for today's naive delusions.

10,000 have joined us as paid-up members! Some people have experienced issues due to such high traffic! It is a safe and secure platform so please keep trying: www.yourparty-membership.uk

Last week, I spoke to the BBC about contactless payments. I appreciated them reaching out to me, and they quoted me in their subsequent article. It's a reasonably normal thing which happens every now and then. www.bbc.com/news/article...

Last week I wrote to Coventry City Council about their £500,000 contract with Palantir. Palantir is a partner in Israel’s genocide in Gaza, collaborates with ICE in the US & is embedded in our NHS. Coventry (& the NHS) must divest from genocide & protect jobs, services & data.

The rumour is OpenAI and co. are seeking revenue splits from retailers as a way of making their products profitable. This raises big concerns about dark patterns and online manipulation. A whole new frontier for consumer protection. Some of my thoughts below. theconversation.com/openai-looks...

I increasingly struggle to separate the AI productivity debate from the computer productivity debate of the 1990s. Actually, I think it's all still the same debate. Some initial thoughts on what I think will become something more substantial: siu2lh.com/articles/are...

Again, I have some mild concerns about universities paying for unrestricted access to this technology to every single undergraduate on campus!

I think this is a larger pattern: "AI" allows you to roleplay competence.

This received quite a positive response so I wrote the paper. Comments are welcome. papers.ssrn.com/sol3/papers....

AI models downplay women's health issues. This is an important result, and one which should seriously temper calls to introduce AI into public life. www.theguardian.com/technology/2...

Letter of the day (in the Times)

truly abhorrent, and for a labour party to be doing this. it should be unconscionable. the UK needs jeremy corbyn and zarah sultana’s new party more than ever.

If you think A.I. will solve your problems, you don't understand technology and you don't understand your problems.

I laughed so hard I thought I was going to cough up an organ. Every line is gold.

I do firmly believe that refusing to outsource your basic cognitive functions to LLMs right now - as so many are stumbling over themselves to do right now, like hogs trotting to a pile of slop that’s been set up inside of a butcher’s delivery truck - will very much pay off in a few years.

We now have 300,000 people signed up and the establishment is clearly frightened. Westminster is broken, but the real crisis is inequality: just fifty families now own more wealth than half the population. Let’s change that together. Join us: yourparty.uk

I had the privilege of giving some lectures on AI and behavioural science this year, and I was struck by how pretty much all the applications which were proposed were, for lack of a better word, fascist. I'm thinking of developing this idea (image is the abstract). Comments or suggestions welcome.

<Angela Collier video intensifies>

Hey, remember when the Panama Papers were released and they showed how basically every wealthy person on the planet was avoiding taxes by offshoring their money and nothing was done about it except the reporter who broke the story was murdered with a car bomb?

Am I wrong?

1. LLM-generated code tries to run code from online software packages. Which is normal but 2. The packages don’t exist. Which would normally cause an error but 3. Nefarious people have made malware under the package names that LLMs make up most often. So 4. Now the LLM code points to malware.

In 2029, the choice will be stark: Socialism or barbarism.

27 people — including a priest, professor and health workers — were arrested today under terrorism laws. Their crime? Peacefully holding signs opposing genocide. This is how we sleepwalk into fascism.

This suggests that the BBC will now vet all bands and musical acts for any political views the government deems undesirable.

I have a new paper with @futuritas.eu in Technovation, all about whether AI can lead to more representative decision-making (sometimes, but only in some relatively narrow circumstances). www.sciencedirect.com/science/arti...

What in the George Orwell is this shit? Why the fuck do cops need to know if we're neurodivergent or not? If cops treated people like human beings it wouldn't fucking matter.

Website of an initiative we have been working on for the last few years on building ethical behavioural science frameworks within corporate contexts. Has been a fascinating year developing the cohorts and materials. www.forgoodframework.com

Worth a watch: Head of Signal, Meredith Whittaker, on so-called "agentic AI" and the difference between how it's described in the marketing and what access and control it would actually require to work as advertised.

I've posted my seminar on 'The role of the university is to resist AI', which takes as its text Ivan Illich's 'Tools for Conviviality'. danmcquillan.org/cpct_seminar...

On LBC @zoejardiniere.bsky.social lists 3 previous instances where military aircraft have been targeted by campaigners in order to prevent them being used in war crimes, and she describes the govt banning Palestine Action as terrorists as an abuse of state power. Well said.