AI is now so unpopular amongst Democrat core voters that in the LA mayoral race Karen Bass clearly feels “his campaign uses AI” is an effective enough attack line—even in such a short rebuttal as this (from the @nytimes.com)

AI is now so unpopular amongst Democrat core voters that in the LA mayoral race Karen Bass clearly feels “his campaign uses AI” is an effective enough attack line—even in such a short rebuttal as this (from the @nytimes.com)

For a while it looked like a truce, but Republicans’ AI fault lines have reemerged. Last week conservative influencers including Bannon urged Trump to make pre-deployment testing of frontier models mandatory. Days later, Trump shelved an executive order on AI safety after industry lobbying (1/2)

AI companies are writing their own constitutions. Meanwhile, a shared vocabulary for contesting algorithmic power has been traveling across red and blue states, from legislatures to civil society. New from me in @science.org: "A Civic Grammar for AI Rights" www.science.org/doi/10.1126/... 🧵

An interesting dynamic in the AI jobs apocalypse discourse is how everyone thinks that their area of expertise is safe, but that areas they don’t know much about are at risk. Eg writers being able to recognise AI writing as turgid and unoriginal; SWEs seeing vibe coding as buggy and unsafe… (1/3)

Read my last two posts to understand (some of) what’s going on here: open.substack.com/pub/verityha... open.substack.com/pub/verityha...

Anna Wintour taken aback by the backlash to the Bezos Met Gala sponsorship. A commencement speaker in Florida shocked by the students booing her every mention of AI. Old people do not understand what’s happening with AI and the young. Here’s what’s going on, why it’s happening, and what it means:

Anna Wintour taken aback by the backlash to the Bezos Met Gala sponsorship. A commencement speaker in Florida shocked by the students booing her every mention of AI. Old people do not understand what’s happening with AI and the young. Here’s what’s going on, why it’s happening, and what it means:

Among Gen Z AI’s net favourability is now minus 44, driven by fears about jobs, the environment, creativity. But there’s another trend that many have missed: AI just isn’t very cool. My latest post explores the trend, the reasons behind it, and the consequences for the OpenAI/Anthropic rivalry:

Among Gen Z AI’s net favourability is now minus 44, driven by fears about jobs, the environment, creativity. But there’s another trend that many have missed: AI just isn’t very cool. My latest post explores the trend, the reasons behind it, and the consequences for the OpenAI/Anthropic rivalry:

The space race shows is that competition and collaboration needn't be mutually exclusive. I'd be amazed if, in 2026, a longer-term dialogue on safety wasn't be part of the agenda this week. Any hint of recognition that AI isn't just a race to be won would be a huge step in the right direction

In 1967 the UN agreed a treaty which proclaimed space as "the province of all mankind" and prohibited weapons of mass destruction from orbit and celestial bodies. It meant that, in 1969, "We Came in Peace for All Mankind" was not just a slogan but a statement underpinned by international law.

In 1967 the UN agreed a treaty which proclaimed space as "the province of all mankind" and prohibited weapons of mass destruction from orbit and celestial bodies. It meant that, in 1969, "We Came in Peace for All Mankind" was not just a slogan but a statement underpinned by international law.

Two years ago I wrote about what a missed opportunity the Biden-Xi Summit was for progress on AI cooperation. If only some foundations had been laid then, we would be in a lot better position today. All the same lessons from then still apply today:

Two years ago I wrote about what a missed opportunity the Biden-Xi Summit was for progress on AI cooperation. If only some foundations had been laid then, we would be in a lot better position today. All the same lessons from then still apply today:

At the Truth Tellers Summit today hosted by the indomitable @tinabrownlm.bsky.social Much discussion already of the challenges facing journalism in a world of unreality. Read our latest post on how Iran is serving as a “toxic laboratory” for the future of epistemic collapse

In this piece Mahsa talks clearly about the current information ecosystem in Iran. A lot of the lessons are equally applicable to misinformation in times of crisis as we’re seeing with the conspiracy theories about the assassination attempt. Recommend this interview to understand what’s happening.

In this piece Mahsa talks clearly about the current information ecosystem in Iran. A lot of the lessons are equally applicable to misinformation in times of crisis as we’re seeing with the conspiracy theories about the assassination attempt. Recommend this interview to understand what’s happening.

I highly recommend this excellent interview between my colleague Catherine Westall and the world leading expert on deepfake and Iran. This week’s edition of AI Needs You:

I highly recommend this excellent interview between my colleague Catherine Westall and the world leading expert on deepfake and Iran. This week’s edition of AI Needs You:

The headline of a recent NBC poll was that AI is now even less popular than ICE in the US. But even more interesting was the finding on how hugely unpopular it is with Gen Z. AI's young people problem is one of the reasons I wrote this post recently about why AI in recruitment is a huge risk:

Everyone wants to talk about AI replacing jobs, but what about AI replacing the hiring process itself? In my new post I argue that it’s the more immediately important issue — because of its effect on young people, who are already on the sharp end of AI’s impact.

NEW POST! On why we should all care about young people being examined, interviewed and rejected by AI:

NEW POST! On why we should all care about young people being examined, interviewed and rejected by AI:

Fascinating political story from 1984 that feels even more urgently relevant today than ever.

📣 My new post on what Silicon Valley can learn from a 1984 report about IVF—and, what they can learn from politicians…

NEW POST: Three groups within the left in the US are increasingly united against one bogeyman: AI. What does this mean for Democrats in the year of the midterms? It’s a coalition they didn’t ask for, but will have to lead.

A coalition on the left won it for the Greens in Gorton & Denton. Read my latest piece about an emerging coalition of the left in the US—how AI is their uniting force, and what that means for the Democrats as they head into the midterms:

Find out how opposing AI became a uniting focus of the Democrat base, and why this causes the party a headache as they head into the midterms. Someone on LinkedIn compared this piece to Tocqueville, so how can you resist?!

NEW POST: Three groups within the left in the US are increasingly united against one bogeyman: AI. What does this mean for Democrats in the year of the midterms? It’s a coalition they didn’t ask for, but will have to lead.

Why are tech leaders flocking to Delhi next week? My latest post explores India’s rise as an AI power broker.

Why are tech leaders flocking to Delhi next week? My latest post explores India’s rise as an AI power broker.

AI is challenging, but it doesn’t need to be frightening. So why do so many talk about every new aspect of it in such hyperbolic terms? My latest ‘AI Needs You’ post argues for calm.

The Sunday Times today very much proving my point with this headline

Talking about AI in apocalyptic, frantic terms serves only to distract from real work that needs doing and has been done. Moltbook was the latest in a long line of triggers for a breathless panic that helps no one. My latest edition of ‘AI Needs You’ on why we crave hyperbole in AI:

New: we dug into Epstein’s dealmaking in Silicon Valley. He used his connections with tech’s elite to get into hot startup deals, some of which paid off handsomely (coinbase) and others that did not (jawbone). W/ @rmac.bsky.social + matt Goldstein www.nytimes.com/2026/02/05/b...

Talking about AI in apocalyptic, frantic terms serves only to distract from real work that needs doing and has been done. Moltbook was the latest in a long line of triggers for a breathless panic that helps no one. My latest edition of ‘AI Needs You’ on why we crave hyperbole in AI:

Not everything in AI has to be scary.

Not everything in AI has to be scary.

Beneath the calls for all-out American dominance in AI, there is a huge amount of division with the Republican party over what to do. The Silicon Valley critique is that Europe regulates tech too much, but plenty of states and MAGA-supporters want to regulate it too.

The consensus seems to say that the Trump administration is joined up in its support for AI and AI companies—but in fact, AI is one of the few issues causing rupture inside the MAGA movement. Read all about it in the latest edition of AI NEEDS YOU (free to read)

The Trump movement is divided over AI, and it highlights something interesting about the future of AI and the future of US politics. My latest post:

New AI NEEDS YOU post is up: On how AI has become an unlikely fault-line in the MAGA movement. Come for the 2am Senate drama, stay for the evangelical conservative wing who think AI scientists are 'apostates'. Plus, Book of the Week!

New AI NEEDS YOU post is up: On how AI has become an unlikely fault-line in the MAGA movement. Come for the 2am Senate drama, stay for the evangelical conservative wing who think AI scientists are 'apostates'. Plus, Book of the Week!

This is good and, more generally, it's always welcome to have someone talking about AI and politics that has actually worked in both AI and politics.

In case you missed my new Substack on AI and politics — first edition on whether comparisons to an ‘arms race’ are appropriate. With thanks to the substack OG @samfr.bsky.social for this kind recommendation!

My TED talk on why an ‘arms race’ is the wrong way to think about AI is now available online. In it I argue: (1/n)

My TED talk on why an ‘arms race’ is the wrong way to think about AI is now available online. In it I argue: (1/n)

My new Substack launched yesterday with this piece on whether there’s a future for AI and geopolitics beyond the US-China binary. It’s something I’ve been thinking and writing about for years, but which is more urgent to consider now than ever. Please check it out and subscribe for future posts👇