Ai

Wednesday

Travis Aaroe

Can AI make Spencer Pratt mayor?

What to make of the new AI election ad created by the filmmaker Charles Curran on behalf of Spencer Pratt, a reality TV star who is running to be Mayor of Los Angeles? The radio host Buck Sexton has already hailed the video as the future of political communication, and Jeb Bush has called it “maybe the best political ad of the year.” The video, which Pratt did not commission, but did repost on social media, shows California worthies – incumbent Mayor Karen Bass, Gavin Newsom, and Kamala Harris – assembled for a sinister banquet. Victims are brought before them: a mother whose children are being harassed by the city’s homeless, and a prostrate Hugh Jackman, who begs to be allowed to rebuild his house in Pacific Palisades.

Spencer Pratt

Tuesday

Jonathan Boff

Accelerating the ‘kill chain’ – a terrifying glimpse of future warfare

From our UK edition

America possesses the most powerful military in history, but since 1945 it has not won a war against anyone other than Saddam Hussein. It appears not to understand why. In fact the only thing the US seems worse at than winning wars is learning lessons from its defeats. People such as the secretary of war Pete Hegseth think it’s all about woke. Lily-livered longhairs stateside stabbed the army in the back over Vietnam; then ‘stupid rules of engagement’ tied the military’s hands in Iraq and Afghanistan and caused the disasters there. The solution is to fight harder, if necessary even at the expense of ethics and the law. Another answer might be to get US forces fighting smarter.

Wednesday

Charles Moore

No one seems sure why Olly Robbins had to go

This session of parliament is due to end between 29 April and 6 May. Now the government is desperate for an Order in Council to kill it off by 9 a.m. on the 29th to avoid another painful Prime Minister’s Questions. The parliament that reassembles for the King’s Speech on 13 May could hardly, in theory, look more like what Sir Keir Starmer wants. His party has the largest overall majority since 2001. He will have jettisoned all hereditary peers.

Peter James

Why is a chatbot deciding what books our children read?

From our UK edition

A school in Greater Manchester has stripped 193 books from its library because they are ‘inappropriate’, liable to upset pupils and thus a safeguarding risk. Among the dangerously destabilising material: Michelle Obama’s memoir and George Orwell’s Nineteen Eighty-Four. Who was entrusted with identifying these literary IEDs? An over-zealous head? A prurient librarian? A demented child psychologist? Nope. A humble AI bot. While writing my latest novel, The Hawk Is Dead, much of which is set in Buckingham Palace, I asked ChatGPT to produce a simple floor plan. Not a tough assignment – much less taxing than diagnosing nearly 200 books as existential threats to adolescent wellbeing. It got the Palace’s orientation wrong. For more than 200 years, the East Wing has faced the Mall.

Sebastian Mallaby

The new AI system causing panic over cybersecurity

It’s tempting, even fashionable, to pooh-pooh the hyperbole from our tech overlords. The release in 2022 of ChatGPT, the first mass-market conversational AI system, unleashed a volley of supercilious put-downs. The chatbot was not intelligent. It was merely hallucinating, manipulating statistics, regurgitating phrases from the internet: it was a ‘stochastic parrot’. Well, over the next three years, ChatGPT became unputdownable. It learned to handle photographs and videos, extract wisdom from dense textbooks, sound like Scarlett Johansson, write everything from code to songs to emails and offer tips on fixing washing machines. Not bad for a parrot.

Martin Vander Weyer

People need to calm down about Nigel Farage’s bitcoin wheeze

From our UK edition

There’s a Tube strike in the old-fashioned style as I write – and you’ll understand the irritation, mine and that of restaurateurs across London, when I add that I’ve just cancelled a table at Noble Rot in Greek Street because my companion can’t face struggling into town. The loss of trade on these days, of which more are planned for May and June, is immeasurably damaging for an already fragile urban economy. More irritating still is the fact that behind the disruption is disharmony between unions: RMT members, just under half the driver workforce, are striking against Transport for London’s proposal of a four-day week (plus extra days off in return for other minor changes) which their Aslef brethren are keen to accept.

James Delingpole

AI could never replace me

There are two main schools of thought on AI in the Delingpole household. I, as the resident batshit-crazy reactionary tinfoil-hat loon, think that it is evil, indeed quite possibly satanic, and that everything would be much better if only we went back to horse transport, herbal salves and abacuses. And Boy Delingpole, representing technologically literate youth, thinks I’m an idiot, that AI is the future and quite mind-blowing in its potential to change everything. Probably we’re both right. Personally, I don’t feel quite as threatened by AI as perhaps I should. More by accident than design, I seem to have ended up in one of the very few jobs that AI isn’t going to steal.

Tuesday

Melissa Kite

Has Airbnb just declared war against its hosts?

From our UK edition

The Airbnb help centre chatbot kept telling me that she understood how frustrating it must be for me to have all these problems created by Airbnb. But she offered no solution, save for congratulating me effusively on being a wonderful host. After a while I asked this person, allegedly a woman: ‘Are you real or is this AI?’ For the relentlessly upbeat drivel she was churning out bore no resemblance to the furious questions I was typing in. I could have told her I was about to throw myself out of the window because of the rise in Airbnb’s fees and the redesign of their app that stops me from using it unless I buy a new £600 iPhone, and she would have replied: ‘Melissa, we know what a great job hosts like you do for guests!

Dot Wordsworth

The creep of AI ‘slop’

From our UK edition

‘The creep of “AI slop” in writing on the arts is everywhere,’ complained the Telegraph’s music writer Ivan Hewett the other week, and it’s up to us readers, he says, to stop it. For others, it means weirdly disjointed videos of anthropomorphic cats. But where did the forceful term AI slop come from? For prisoners, slopping out meant emptying chamber pots in the morning. For two centuries slop has been swill or food for pigs. At the same time it could be dirty household water. Long before that, it meant unappetising, semi-liquid food. Sloppy Joes were, from the 1940s, a kind of mushy hamburger (and also a loose-fitting sweater). I remember from the 1960s a probably ironic use of slop in the lyrics of a song by Arthur Lee on Love’s Forever Changes (1967).

Monday

Alex Diggins

Who wants to bring back the Neanderthals?

From our UK edition

In the not-too-distant future, if your T-shirt starts giving fashion advice or we’re all enslaved by a race of disease-resistant metahumans, then blame Martin Amis. More precisely, blame his obsession with Space Invaders. With a foreword by Steven Spielberg, Amis’s 1982 Invasion of the Space Invaders: An Addict’s Guide to Battle Tactics, Big Scores and the Best Machines gave intellectual heft to a pursuit – videogaming – that had hitherto been the preserve of glassy-eyed youths. His advocacy proved prescient. Over the next couple of decades, the best minds of a generation dedicated their lives to making pixellated worlds as compelling and realistic as possible.

Wednesday

Lionel Shriver

We’ve already given up on novels

From our UK edition

Late last year, I was notified that one or more of my novels might have been fed to an Anthropic large language model, because in a class-action suit the company had reached a copyright settlement with authors who’d never given an AI Goliath permission to gorge on their work. Sure enough, a website verified that 11 of my books had been used as silage for this insatiable digital leviathan. Each of the LLM’s tasty Shriver mouthfuls may merit compensation of about $3,000. But before I take out a loan against that $33,000 sure thing to buy myself a Chinese EV, I should read the fine print: ‘court-approved costs and fees’ will be deducted, meaning I may garner funds sufficient to buy a whole new packet of extra-fine felt-tips. Fees or no, this is chump change for Anthropic.

Daisy Dunn

China wants robots to look after the elderly

From our UK edition

An AI data centre – imagine a factory of buzzing wires and computing equipment cooled by industrial fans – can consume as much power as a city. It has been estimated that, not too long from now, we’ll require 92 cities’ worth of extra power just to meet the demands of artificial intelligence. Ergo, the heat is on – but so, it is said, is a new cold war. On Radio 4 last week, Misha Glenny was exploring how the rapid evolution of technology is shaping the rivalry between the US and China. It turns out that the race for pre-eminence – in AI, at least – is as close as the 1973 Grand National. Red Rum (China) has the current lead, but that lead is ‘razor-thin’ and is thought to owe something to the nature of American tactics.

Tuesday

Sebastian Faulks

The mystery of what makes us special remains unsolved

From our UK edition

Consciousness is thought by many to define what it is to be human. We know that animals are conscious to some extent, but they don’t have what we have in that department. So if we can explain how human self-awareness works it will be a Rosetta Stone to understanding what makes our species so odd, or, depending on your view, so special. Also: big brains love exercise. And the great thing about this topic is that it can be tackled by biologists, neurologists, philosophers, physicists and even novelists. It’s an interdisciplinary, no-holds-barred intellectual wrestling match. Michael Pollan, an American journalist and academic who works at both Harvard and Berkeley, comes from a literature background but is fluent in the relevant science.

Saturday

Jordan McGillis

How Iran could end the AI boom

While Americans anxiously watch the price of gasoline tick higher as the war in the Middle East squeezes the global oil supply, the conflict has highlighted another energy vulnerability that could prove just as costly: Taiwan’s dependency on foreign natural gas. At first blush, energy issues an ocean away seem peripheral to American interests. They are anything but. Though the effect on the American economy won’t be immediate, energy insecurity in Taiwan is a looming disaster. Qatar, the source of 30 percent of Taiwan’s natural gas, has been effectively bottled up The reason is that AI – in fact, virtually all modern computing – is highly reliant upon the steady production of semiconductors in the world’s only true hub, Taiwan.

Semiconductors

Wednesday

Zack Christenson

Inside the race to build AI data centers in space

In the 1966 novel Colossus by British author D.F. Jones, a supercomputer (which goes by the name of Colossus) is given control and decision-making power over the US’s nuclear arsenal – a logical and unemotional computer being better placed, it is assumed, to make unemotional decisions than a human. Eventually, Colossus discovers the existence of a similar supercomputer in the USSR and begins communicating with its Russian counterpart in mathematical languages about technological advances beyond human comprehension. Frightened by the possibilities this presents, scientists sever the connection – only for Colossus to threaten to launch nuclear weapons if it isn’t reconnected.

Tuesday

Gary Dexter

Your AI Granny will speak to you now

There’s a trend on YouTube at the moment for videos in which older people give advice. They speak directly to camera, frankly and without pretension. One can almost sense the care home staff hovering in the background, coaxing their barely extant charges into making one last testament of their time on Earth. The videos have titles such as ‘Things I’d tell my 30-year-old self’, ‘Harsh realities of being an 85-year-old woman’, ‘A girl and a woman talk about life’, ‘Lessons learned’ and ‘How to deal with loneliness.’   The commenters below the videos respond, largely, with gratitude and a surprising lack of trollery. ‘I’m terrified of dying,’ one commenter writes. ‘I fear many things, but nothing scares me as much as death does.

Tuesday

Max Horder

I don’t trust AI’s built-in ‘safety systems’

Cars ruined cities. Anyone can see that cities built before the invention of the automobile are incomparably more beautiful and serene than anything built after them. The contrast between Los Angeles and Prague is unmistakable. But people like things that move fast and make life easier, which means we’re stuck with the modern city hellscape whether we like it or not. And today, the same is true for AI. The contrast between the internet five years ago and today is unmistakable: content-slop, workslop, AI-generated comments, fake opinions and phony judgments, trite phrases, apocalyptic hysteria, the biggest intellectual-property heist in human history – all because of the invention of Large Language Models (LLMs).

ai

Wednesday

The Spectator

Letters: AI won’t save the army

From our UK edition

Brute force Sir: General Sir Nick Carter is correct to point out the fragility of the UK’s armed forces today (‘Empty shell’, 7 February). He is also right to highlight the level of expenditure which will be necessary to overcome 25 years of structural under-investment in defence if the UK and its allies are to deter or win any future war. However, the suggestion that the British armed forces might be saved – relatively cheaply – by the institution of AI-automated kill chains alone is questionable. Indeed, it may be just another mirage of the type which has contributed to the current predicament. Autonomous weapons systems have existed for many years. With AI, they will proliferate and may become more important.

Sunday

Sean Thomas

Moltbook: has AI created its own religion?

From our UK edition

20 min listen

What did you most recently use Artificial Intelligence for? For most people, the answer would be as a glorified search function, using services like Chat GPT to ask questions, draft text and even produce images – like the Chat GPT generated thumbnail image for this episode. The capability of AI far exceeds this most though. Sean Thomas joins Damian Thompson for this episode of Holy Smoke to talk about 'Moltbook', a social network built exclusively for AI agents – and which has now created its own AI 'religion'. What does this mean for humankind? Is AI just replicating a belief impulse, to the extent that one exists within humans? And will we one day end up worshipping AI? Produced by Patrick Gibbons.

Moltbook: has AI created its own religion?

Thursday

Steve Hsu

How useful is AI for research?

Late last year, I published the first theoretical physics paper in which the main idea originated from artificial intelligence – from an AI. And my experience working with the most powerful AI models left me both impressed and wary. The most accurate analogy I can offer is that it’s like something familiar to anyone who has done research: a brilliant colleague who is also unreliable. This colleague can produce deep insights at surprising speed – and then, the next second, make an error that ranges from the trivial to the profound. That tension – capability versus reliability – has shaped how I now use these systems in mathematics and theoretical physics. It is also what will shape how they affect scientific research over the next few years.