My letter in today’s FT quoting the late, great Charles Kennedy

The rise of "victims" in political discourse

Some good 🇬🇧news! Jade brings intensity and ambition to whatever she touches. She's consistently made great bets at the frontiers of technology. Great appointment.

Fancy something that goes against the narrative? Then try this puckish piece from Tim Leunig on the left case for Reeves.

Jeff McMillan, Chief AI Officer at ⁦@MorganStanley⁩ is interviewed for the *Age of Intelligence* podcast. Very lucid on how corporate execs should approach the technology.

So, if I get this correctly, the answer to 1000+ state level regulatory proposals is to ban all regulation? Because clearly there is no signal that maybe something needs to be regulated?

The criticisms of the India FTA are based on a series of misunderstandings. 🔴"Easier immigration". No. There is nothing in this deal about migration. Nothing. There is an extension of the CATEGORIES under which people can apply for work permits, but the total number can still be

London is the second-best city in the world to start an AI startup (and could be #1, in my opinion): - Abundant early-stage capital - High density of exceptional AI talent - Significantly cheaper and more loyal talent compared to the Bay Area - Pro-growth government and

Everyone has written off the London Stock Exchange. The best companies (Revolut, Monzo etc) will not go public in London. Companies that do (Wise, Deliveroo, Darktrace) are punished on price for doing so. Pension reform and stamp duty has destroyed so much economic value for

Time for King Charles to visit Canada, where he is head of state and has every right to go whenever he wants. I have no doubt the welcome would be ecstatic. And very symbolic. Retweet if you agree.

“Humanity has two choices. A future with safe and ethical AI or one where there is no AI. We are currently pursing the 3rd option. The one where humanity has no future.” Stuart Russell #IASEAI25

This OPM memo is going to be the most impactful news of the day, but I'm not sure it'll get much reporting.

Bad DeepSeek takes flying thick and fast today. Thread of good ones instead: (all subject to Matt's correct meta-take, caveat lector ⬇️ )

As someone who has reported on AI for 7 years and covered China tech as well, I think the biggest lesson to be drawn from DeepSeek is the huge cracks it illustrates with the current dominant paradigm of AI development. A long thread. 1/

1/ In Sifted today, Yann Lechelle is quoted as saying "it will be decades before Europe can build big, big companies like those in the US. So the main question for me is, what do we do in the meantime?". I strongly disagree with this point of view, here's why:

Amusing to hear the @TheNewsAgents talk about this wonderful app that the democrats are using. Singing its praises about how it has all the voter data, ages, names, voting history, lines to say etc. It’s an app the Lib Dems have been using for years. MiniVAN!

“My using technology more effectively”

Sure there will be #AI titbits in #Budget2024 (“£xm for compute! ��ym for skills!”) but key question is what AI-related savings assumption is baked in to projected Gov spending. @InstituteGC have suggested £40bn… Labour’s equivalent of “efficiency savings”. Deus Ex Machina

Many organisations have internalised that AI is learning and improving from its users. The lived reality is the reverse. AI users are learning from, and being trained by, the systems that they use.

But great focus on: - Using AI to humanise Government (example is DWP tool to check if you sound in trouble which speeds up response) - Never taking human out of the loop - Radical Transparency (eg they share code - leading to people downloading and trying to sell back to HMG…)

A slightly rocky start (in Boris Johnson’s time): “If you show them the data and it contradicts what they have always believed then you don’t get invited back to the meeting”