In your briefing today:

📣 John Swinney says Andy Burnham will be the last Prime Minister of the UK and says he wants a second Scottish independence referendum by 2031.

He made the comments after talks with nationalist leaders from Northern Ireland and Wales in Cardiff: the trio signed a joint declaration stating that “Westminster’s time is coming to an end”. (BBC)

📣 The oil and gas industry is pushing for Andy Burnham to approve the Rosebank and Jackdaw North Sea developments - and unlock more than 100 other new oil and gas projects. Offshore Energies UK also wants Chancellor John Healey to make changes in his budget, to be delivered on October 28, that they say would generate an extra £15 billion for the Treasury. (Scotsman)

📣 Former First Minister Nicola Sturgeon has been reported to police for allegedly handling stolen goods.

Pro-independence blogger Stuart Campbell, who runs the Wings Over Scotland site, made the complaint alleging Sturgeon knowingly possesses items that were bought or stolen as a result of fraud or embezzlement - known as “reset”.

Sturgeon has previously said some of the items stolen by Murrell were given to her as gifts. (Times) (Wings Over Scotland)

The 8-hour workday is becoming a 4-hour workday for people who know how to use AI.

Everyone else is still catching up.

This AI work playbook shows you exactly how to cut your work hours in half using AI.

Sign up for Superhuman AI and get:

Claim your AI playbook

📣 Police have confirmed the man killed in a suspected shooting in West Lothian was the victim of a targeted attack. (BBC)

📣 Long-lost oyster beds on the West Coast of Scotland have been rediscovered thanks to researchers searching for references to eisir, eisirein and eisirean in Gaelic placenames. (BBC)

📣 Oasis will begin a mammoth 2027 tour with five nights at Celtic Park next May. (Daily Record)

📣 The public inquiry into Lucy Letby’s crimes will publish its findings today: it has been investigating how concerns about her conduct were handled, and may tell us if babies could have been saved by earlier action. (BBC)

📣 Bosses at some of the world’s leading AI firms are calling for regulation of the sector - but Washington, and in particular President Donald Trump - aren’t especially interested. (AP)

📣 Prince Archie and Princess Lilibet, the children of the Duke and Duchess of Sussex, have moved schools after only two days due to security concerns. The family relocated from the US in the summer. (Hello! has the exclusive)

📣 Rapper Macklemore has been dropped from Ed Sheeran’s US tour after a “free Palestine” speech infuriated venue owners, who threatened to halt the shows. (Guardian)

📣 Matthew Rhys has become the first actor in 78 years to win two Emmy lead actor awards for performances in different shows in the same year. (Guardian)

⚽️ New Scotland boss Sebastien Pocognoli will unveil his first Scotland squad later today, ahead of the Nations League games between September 26 and October 6, and we should expect to see some new faces. Stoke City defender Luke Graham and Sevilla striker Robbie Ure should make the cut. (Daily Record)

⚽️ Leeds thrashed Newcastle 4-1 to rise to third in the Premier League. (Guardian)

🗣️ It’s tempting to dismiss the hue and cry around AI as being just another moral panic, and a commercially useful one at that: a lot of over-excited technologists vastly overestimating the impact of their work while also - as it turns out - seeking to make billions from their sales pitch.

Can it really be that, just as AI companies such as Anthropic and SpaceX are raising vast, vast sums of money from investors, their CEOs are taking to social media to wring their hands about their systems being just so gosh-darn powerful, they might spin out of control and wipe out all humanity?

Later in this briefing, I’ll bring you two sceptical, expert voices who make powerful cases for the AI panic being exactly that.

But first, let’s look at the slightly odd case for the prosecution - odd, because it’s being advanced by people who are in or around the industry that’s said to be going rogue. This isn’t an ill-informed pressure group… there are enough sensible, non-aligned voices demanding that the rest of us pay attention.

First: what are people actually alleging? The alarm was raised by Jacob Coxton, who resigned from AI company Anthropic last week and posted a long thread on X about his concerns. His beef in brief: AI systems are getting too powerful too quickly - evolving more quickly than our understanding, and eons more quickly than any policy response.

That means the human race is naive, or lacks the necessary skills to cope: like someone encountering Covid without a vaccine, we’re more susceptible to being laid low because this is the first time we’ve faced the problem.

Many people have now spoken up in support of Coxton’s concerns.

Second, what could rogue AI actually do, and couldn’t we just switch it off? There are a lot of memes suggesting we could just turn the AI machines off, like the one below. Jack Clark, one of the founders of the AI company Anthropic, has encouraged that idea by suggesting in an interview with the BBC that a “kill switch” might need to be made mandatory.

Alas, one of his own colleagues - Anthropic’s CEO, Dario Amodei - suggests it’s just not that simple.

What keeps Amodei awake at night is the thought of a “botnet” - a swarm of AI agents, spawned by an AI model in pursuit of a particular task, acting independently of each other and of the “parent” AI in a destructive way. He points to the “OpenAI-Hugging Face incident” (OAI-HF), in which a swarm of agents went rogue and tried to hack the software that was “grading” their performance.

“Given the accelerating rate of AI capability development, it’s my worry that in 6–12 months such a swarm could be capable of taking over the entire internet with a persistent botnet (potentially causing hundreds of billions of dollars in damage), and that the scale of damage would continue to increase from there if AI becomes more powerful without the necessary guardrails.”

There isn’t necessarily a “kill switch” that can stop that damage before it happens.

Third, why would AI misbehave? Yoshua Bengio, the world’s most-cited scientist, attempts to explain why AI systems might misbehave.

It’s all to do with the way they’re brought up, he suggests, if not quite in such glib terms.

Specifically, like a baby growing up, newborn AI systems go through three types of trial-and-error training before they’re released to the wild. Because the training materials are all created by humans, the AI inherits the human motivations and foibles behind the content it creates and consumes.

Bengio suggests there are other ways we could train AI systems without using humans, thereby introducing fewer problems into their behavioural models.

Finally: I did promise you those sceptical voices. So, first, former Federal Trade Commission boss Alvaro Bedoya, who states: “AI will not eradicate humanity. Humans survived an ice age, the Black Death, two world wars, and (so far) the advent of nuclear weapons. Anyone who is loudly warning of AI-caused human extinction should not be taken seriously.”

Second: author and activist Cory Doctorow, who’s been calling bullshit on the tech revolution for almost as long as I’ve been on it, calls it here again, and loudly. His writing is accessible and enjoyable: this is worth a read.

His killer paragraph: “‘AI’ – chatbots that wake up, ‘set their own goals,’ and ‘spontaneously’ start hacking servers – is fake. It doesn't have ‘a 10% chance of ending the human race.’ The Hugging Face hack isn't a mysterious, supernatural occurrence. It's a Python loop and a chatbot. The people responsible didn't accidentally create god: they created autonomous malicious software and then failed to closely monitor it, resulting in it doing something both foreseeable and bad.”

In other words: Don’t have nightmares, sleep well?

Sent this by a friend?