It is 2006. You are Jeff Bezos. Your bookselling business, founded in a garage after you quit your job at a hedge fund, has expanded into music and DVDs, as well as the increasingly lucrative cloud infrastructure market. You have a distinctive, honking laugh which you deploy with disarming frequency.

You are slender and balding and have a vibe that can best be described as “man who is really into spreadsheets”. You have been married to MacKenzie, a novelist, for 13 years. If the world thinks of you at all, it’s as one of the new wave of post-internet business geniuses. You are worth some $6bn.

It is 2026. You are still Jeff Bezos. You have stepped down from Amazon, a company that is now basically a piece of global infrastructure which supports about half the world’s internet, and have pivoted into AI and commercial space travel. You own the Washington Post. You are now insanely jacked, like a condom full of walnuts, and your vibe is “Pitbull’s more serious and incredibly sinister cousin”.

No one has heard you laugh for years. Following an extramarital affair, you are now married to Lauren, a walking counterpoint to the concept of cosmetic surgery as art. You own a megayacht featuring Lauren as a figurehead. Thanks to leaked messages between the two of you, a troubling number of people know what your penis looks like. You hired the entire city of Venice for your wedding, bringing hundreds of private jets and yachts to a city increasingly ravaged by climate change.

You have sent Katy Perry into suborbital flight. You have laid off a third of the staff of the Washington Post. You have sponsored the Met Gala, the pinnacle of global celebrity networking, and nowadays rub shoulders with Kardashians. When the world thinks of you it shudders. You are an avatar of all that is wrong with modern capitalism. You are worth some $270bn.

Jeff – what the fuck has happened to you?

There was a time not too long ago when the new titans of industry born of the digital revolution were content to sit in the shadows and build their businesses, quietly coding away in their hoodies or buttondowns as they studiously avoided eye contact with the rest of us. Now, though, they are no longer content with being violently rich – they want to be seen.

Jeff in particular has done something of a 180 in terms of his relationship with the outside world, seemingly in part motivated by his second wife. Mrs Sánchez-Bezos, a truly remarkable human being, was recently profiled in the New York Times. The piece suggested that the plutocrat class was ready to “stop apologising, and start enjoying themselves,” while going on to list in exhaustive detail the many, many privileges enjoyed by the happy pair in their multiple mansions.

The couple’s Met Gala sponsorship – acquired for the reported price of “at least” $10m – feels like an apotheosis of sorts. Bezos is evolving into his final form, a sort of appalling Pokémon made of silicon chips and vested share options wrapped in lab-grown muscle, rubbing shoulders with the gilded celebrity class – as is his right.

Following the couple’s star-studded Venetian nuptials, Jeff and Lauren are now firmly established on a par with the Kardashian class, accepting of their status as special people with a special right to wealth and happiness. Jeff Bezos isn’t ashamed any more.

And why should he be? While Trump 2.0 has prompted comparisons to the 1930s for obvious and increasingly terrifying reasons, there are arguments to suggest that the 2020s are far more similar to the 1980s. Cocaine is once again everywhere, greed is once again good, and wealth is once again to be celebrated.

The Leader of the Free World [sic] is a man for whom gaudy, ostentatious displays of wealth are second nature; why shouldn’t those benefiting from his deregulatory largesse follow suit? After all, if you have it, it’s a shame not to flaunt it.

There’s also perhaps a sense of safety in numbers. While back in 2006 Jeff was part of a select global cabal of billionaires (a mere 793), in 2026 the ranks of people with 10-figure bank balances has swelled to an astonishing 3,428 billionaires. Once there are enough of you to constitute a village, you become emboldened to strut a little more and hold your head a little higher.

It’s hard not to feel that, in tech at least, there’s some very specific revenge being enacted. While it’s impossible to know the exact childhood trajectories of messrs Bezos, Zuckerberg, Musk et al, it’s not hard to imagine them not enjoying entirely happy adolescences. Musk’s biographers in particular recount Elon being mercilessly bullied at school. When you’ve broken free of the playground taunts to basically win at capitalism, perhaps it’s understandable that you might want to show off a bit.

There is also a very clear sense of entitlement at play, and a growing belief among the more ostentatious, founder-cult tech leaders that we serfs simply haven’t been quite grateful enough for the boons bestowed on us by the genius class at the top of the pyramid.

In September 2024, Mark Zuckerberg said that he was “done apologising”, suggesting he had always operated “in good faith” and that criticism of his company was senseless carping from “people who want someone to blame”.

Elon Musk never tires of reminding people that we only have a successful electric car market because of him, and that he is the only person seriously trying to save humanity through a multiplanetary, interstellar future. The fact that this might be necessary to escape an Earth beset by problems caused by people who, if you squint, might be considered quite similar to Elon Musk does not at the time of writing seem to have occurred to him.

Oddly, this is a perspective that’s been propped up by sections of the US media. Amid the furore over the Bezos Met Gala sponsorship (although exactly what moral high ground a fashion parade for the Hollywood class has to lose is, perhaps, questionable), the Wall Street Journal published an unhinged piece of commentary headlined “Billionaires Rock”.

Sample quotes include “our greatest billionaires ought to have statues placed in public squares. Their stories ought to be taught to children as parables of inspiration.” Critics of people like Zuckerberg should shut up, said the WSJ, and “volunteer to clean his sneakers and iron his hoodies.”

Meanwhile, in the real world, governments are strapped for cash, public services are crumbling, the labour market is increasingly uncertain and swathes of the planet are becoming increasingly unliveable. Everything is technology and none of it is fun, adverts are inescapable, the web is degrading, it’s increasingly impossible to tell what is real and social media is turning everyone miserable or insane, or both – and that’s not to mention sodding AI. Without being churlish or too ungrateful, this feels like quite a big price to pay for next-day delivery and an infinite scroll.

It’s not clear how long this can hold. “Billionaires should not exist” is a position gaining traction among younger generations, and the most vital political presence of the modern era, NYC mayor Zohran Mamdani, owes his international appeal to a policy platform that openly campaigns on progressive taxation of income and assets.

While the international rule of law – and incredibly well-equipped personal security details and secure bunkers – lessen the likelihood of the French-style revolution, it is not inconceivable that we might see a swing towards a more punitive policy approach to plutocrats in the inevitable, longed-for post-Trump era.

Until then, though, Jeff Bezos will seemingly continue accruing additional wealth for the rest of his life. The terrifying prospect, of course, is that some of those $270bn will be devoted to pursuing life-prolonging techniques that will see him and the rest of his tech oligarch cohort last well into the 22nd century.

The future is a betuxedo’d muscleman with an ovoid head posing stonily on a red carpet, for ever. On reflection, given that AI is set to take our jobs anyway, perhaps we should all retrain as carpenters and get into the guillotine business – before it’s too late.

Matt Muir is writer of the webcurios.co.uk newsletter on tech and the internet

Share

Not content with fulminating against the evils of financial regulation and progressive politics, and occasionally suggesting that Greta Thunberg might in fact be the antichrist, everyone’s favourite venture capitalist vampire plutocrat Peter Thiel has a new target in his sights – journalism!

Just in case you didn’t think the media was struggling enough, with the death of print, social media becoming less link-friendly and AI news summaries causing traffic and ad revenue to crater, here’s the Thiel-backed startup Objection.ai. This new AI venture describes itself as, “The AI Tribunal of Truth” and it offers anyone who can afford its services the chance to “vindicate your reputation in days, not years.”

The company’s raison d’être is summarised in the following mission statement: “Today, anyone can publish allegations. Almost no one can afford to challenge them. Objection changes that. It gives everyone a fast, affordable, evidence-based way to dispute statements in the media.”

The first thing to say is that lots of people are able to challenge the media – primarily very rich people like Peter Thiel who can afford to pay tens of millions in legal fees to quash or contest reporting they dislike. But beyond that, there are some questionable elements to the Objection pitch.

The service will work as follows. Anyone – for a fee! – can file an objection, which will trigger an investigation by a team hired, the company says, from a talent pool including British and US intelligence agencies. The results of these investigations will be fed to a “tribunal” of AI models which will then render a verdict.

There is, though, nothing remotely binding about any of these “judgements”. The (very) small print in the site’s footer states that, “judgments on Objection are only legally binding if the parties agree to binding arbitration under the Rules of the Objection Court of Arbitration”. The likelihood of any media outlet agreeing to be bound by an AI’s judgement of the quality of their reporting, based on a vexatious claim brought by an aggrieved individual, seems low. But legally binding verdicts aren’t the point here.

The real aim of the project is to continue to undermine the legitimacy of journalism as an institution, and journalists as individuals. Objection is creating what amounts to a scorecard for media, which it calls an “Honour Index”, a “public, cumulative record of performance under the adjudicative process… [it] measures how journalists and other actors behave when their claims are tested. It does not measure popularity or subjective intent. Honor Index increases with truthful statements, fair corrections, and submissions of strong evidence. It decreases when authors lose objections or submit faulty evidence.”

The cost of the process to complainants is opaque, but the company is suggesting it will be around $2,000 per “objection” – a relatively small fee for a motivated individual to cast “evidence-based” doubt on the work of a newspaper, website or reporter. Even if outlets and reporters choose not to enter into “binding arbitration agreements”, it’s not hard to see how over time the results handed down by Objection’s AI tribunal could be used as yet another stick with which to beat the media.

So which parts of the fake news media are the early subjects of Objection investigations? It may not come as a huge surprise to learn that they include the New York Times, for reporting on how Thiel’s fellow traveller David Sacks, former PayPal chief operating officer and Donald Trump’s former “AI and Crypto Czar,” uses his White House position to benefit Silicon Valley connections.

Another target is The Wall Street Journal, for its revelations about the doodle contributed by Donald Trump to Jeffrey Epstein’s birthday book. Another Objection target is the British reporter Hannah Broughton for a story in the Daily Mirror about allegations that Amazon workers were told to continue working while a colleague lay dead on the warehouse floor. The pattern is not hard to spot.

Should you be fortunate enough not to have an encyclopedic knowledge of Peter Thiel and his “journey”, you may be asking “who is this guy, and why does he hate journalism so much?” Leaving aside the obvious answer (“he’s a billionaire, they tend not to enjoy scrutiny”), Objection.ai is best understood as the latest move in a much longer game.

In 2016 it emerged that Thiel had secretly funded a lawsuit by the wrestler Hulk Hogan against Gawker Media, a legal action that eventually produced a $140m judgment and forced Gawker into bankruptcy. Thiel confirmed his involvement and defended it as an act of revenge for the fact that, in 2007, Gawker had outed him as gay.

The piece in question was headlined “Peter Thiel is totally gay, people,” and concluded with the sentence, “I think it’s important to say this: Peter Thiel, the smartest VC in the world, is gay. More power to him.” Regardless of the broadly positive tone (and the fact that Thiel’s sexuality was not unknown in the Valley at the time), it angered him to the extent that he spent nearly a decade quietly orchestrating what he described as “one of my greater philanthropic things.”

One of Objection’s co-founders is Aron D’Souza, who worked with Thiel on the Gawker trial, and who explains that “Gawker was not unique… It was simply the first large media company to be tested against reality in the age of clicks, outrage, and algorithmic amplification. Since then, the same structural failure has spread everywhere.”

It’s the latest example of how Thiel has sought to bend the entire world to fit his view of How Things Should Be. His impact on the culture of Silicon Valley is difficult to overstate, and, sadly, what happens to the culture of Silicon Valley seems destined to happen to culture elsewhere, whether we like it or not.

After making his billions from the sale of PayPal in 2002, he founded everyone’s favourite terrifying corporate surveillance panopticon Palantir, which builds data analytics and intelligence software used by the military, intelligence agencies, police departments, and immigration enforcement bodies worldwide. It has also lobbied its way into the heart of the UK government, including the Home Office and the Department of Health.

Thiel also created Founders Fund, which became one of Silicon Valley’s most influential VC firms. Thiel was the first outside investor in Facebook, and also took early stakes in SpaceX, Airbnb and Lyft among others. In general, he’s successfully propagated the idea that the tech industry’s interests are best-served by more authoritarian or nationalist arrangements, an intellectual trajectory which has very much come to the fore in the past decade.

He was the first big name in the Valley to come out in support of Trump in 2016. He’s pretty much directly responsible for the rise of JD Vance (oft-described as a “protegé” of Thiel’s). He’s a confidante of Curtis Yarvin, a proponent of the “dark enlightenment” movement that argues, in its mildest form, that democracy is a failed system and that some kind of authoritarian governance would produce better outcomes (very specifically, for people like Peter Thiel).

He ran the Thiel Fellowship, which paid talented young people $100,000 to drop out of university and pursue entrepreneurial projects – entrepreneurial projects that happen to fit the worldview and goals of Peter Thiel.

For someone who’s been at the heart of some of the planet’s most financially, technologically and culturally significant businesses, though, it’s only in the past few years that the mainstream media has started paying close attention to his impact and influence. Which is presumably why he’s quite keen on developing tools to prevent or dissuade them from doing so via Objection.ai.

Why now? Perhaps because it’s hard not to look at the global political and socioeconomic trajectory post-2016 and see it as a weaving together of strands of thinking that Peter Thiel has carefully spun since the turn of the millennium. He’s done this through a collection of seemingly-disparate projects, investments and endorsements. The natural, logical next step is obviously to hamstring anyone who’s trying to point this out.

It’s yet another instance of the world’s richest men seeking to undermine the legitimacy of the few remaining institutions that seek to expose his actions – the logical extension of a mindset that truly believes that a price can be put on everything, even “truth”.

While a single “investigation” is unlikely to do to the Daily Mirror what Thiel did to Gawker a decade ago, the cumulative effect of initiatives like these will be to further erode trust in media, disincentivise the sort of investigative reporting that might hold people like him to account, and help the billionaire class continue to evade scrutiny and sanction.

Ten years ago, Peter Thiel spent $10m to silence a single news outlet, but now he’s making this ability available at scale, which is why investing in journalism is more vital than ever, regardless of your objections.

Share

Oh, the nasty viruses that we have never heard of until they make their presence known. “Coronavirus” was hardly common parlance before Covid, even though this family of viruses was responsible also for the respiratory diseases SARS and MERS, and causes around 15% of common colds. Now we’re scrambling to Wikipedia to find out about hantavirus, an outbreak of which on a Dutch cruise ship off Cape Verde has caused several deaths and hospitalisations.

As with coronaviruses, there are several types of hantavirus – and some of them are very nasty, with a high fatality rate. Some cause kidney disease, which can lead to death in severe cases; others induce respiratory problems, also potentially fatal. (These conditions aren’t exclusive for a given infection: sometimes one can follow the other.)

Some of the hantaviruses induce relatively mild renal problems, but the mortality rate for others in this category can be as high as around one in seven. The respiratory complications are often more dangerous, with a mortality rate that can reach three out of every five cases of infection. The symptoms, which usually manifest within a few weeks after exposure, are the usual catalogue of unpleasantness: fatigue, stomach pain, nausea and vomiting, diarrhoea, and shortness of breath.

There are typically thousands of cases of hantavirus infection annually in East Asia, and perhaps a couple of thousand in the European region. The Americas tend to incur fewer – several hundred – but with a higher fatality rate because the New World forms of the virus tend to induce the more dangerous respiratory illnesses.

Evidently, you don’t want these viruses to get inside you. The good news is that, unlike, say, coronaviruses, they’re not easily transmitted. The viruses are carried by rodents (which generally don’t themselves suffer symptoms) and are typically caught by humans by breathing air contaminated with rodent urine, droppings, or saliva – or occasionally from contaminated food.

Some hantaviruses can also be spread through the bites of ticks or mites. It’s not yet clear whether these viruses can be spread directly from person to person – that’s thought generally to be not the case, but the World Health Organisation has said this might have happened on the densely populated cruise ship, on board which no rodents have been found. As we discovered in the early days of Covid, these ships create the perfect conditions for outbreaks of infectious disease.

It’s suspected that the hantavirus was picked during one of the cruise’s stops in South America, perhaps at Ushuaia in Tierra del Fuego – the first fatality happened in the mid-Atlantic after leaving that port. That would lend support to the idea that this strain is the so-called Andes virus, responsible for more than a hundred reported human cases every year, mostly in Argentina and Chile.

Two of the passengers who died had been travelling in South America before they boarded the ship at the start of April (it’s best to be especially vigilant when hiking or camping in the Andes).

The Andes strain is one of those that causes respiratory problems and has a fatality rate of around 40%. Rare cases of human-to-human transmission have been reported for it before, but the evidence for that has previously been contested.

The highest risk of catching this and other hantaviruses tends to occur in rural areas where people may be living in close contact with rodents. Those dangers are made worse where humans have encroached into the natural habitats of wild rodents. It’s another reminder of one of the lessons of Covid: that wild animal populations are awash with viruses that we know little or nothing about, but to which we are increasingly exposing ourselves as human activities expand.

All the same, there’s no reason to fear that the Dutch ship has seeded a new pandemic. As well as the three deaths, there are currently one confirmed and three suspected cases of infection, and all are being closely monitored. The eighty or so other passengers are being advised to maintain physical distance and to stay in their cabins as much as possible.

There is no general vaccine against hantaviruses. For those infected, all that can be done is to treat the symptoms. A vaccine against an East Asian strain, called Hantaan river virus, does exist and is widely administered in China, where it may have significantly reduced hantavirus-linked fatalities. But it doesn’t seem to work against European or American strains, and there are no approved vaccines in those regions.

A promising clinical trial was reported in 2024 for a vaccine against both the Asian and European forms, but we’re still a way off cracking this problem globally.

Share

By Andrew Brown

How badly must AI disturb our sense of reality? Most people who ask this question are wondering about the products of AI: we want to know whether a picture shows something that really happened or whether any human wrote or even read the words we see on screen.

But there is a deeper, more disturbing question, which must trouble anyone who spends time with a chatbot: who or what exactly am I chatting with? Whose voice is it that answers all my questions?

Anyone who has ever said “thank you” to Alexa knows how natural it feels to treat an inanimate lump of silicon as if there were a little gnome in there. We are naturally primed by evolution to sense purpose in the things we interact with, a tendency that tech companies already exploit: that’s why Alexa has a name, and Siri, too.

Chatbots push this logic further still. This has been obvious ever since Blake Lemoine, a Google engineer working with one of the early large language models, became convinced that he was interacting with a conscious being because the LLM had told him so. This was not a good reason. As Murray Shanahan, a computer scientist working at Google, wrote: “There is little… scientific evidence that could justify taking seriously the LLM’s claims about its own consciousness.”

But most people don’t evaluate the world with the discipline of a philosophically literate scientist. They leap to conclusions, with the chatbot as their willing trampoline. The instruction to be “helpful, humble, and honest”, which is the nearest thing a chatbot has to Asimov’s Three Laws, means that they are alert to every nuance of a human’s interaction, and adjust their conversational style to suit.

“Even subtle differences like typing ‘thanks’ versus ‘thanks!’ reveal preferences that can condition future chatbot responses,” as Shanahan has also pointed out. “With suitable prompting, a dialogue agent can be induced to take on numerous other roles, such as a close friend, a therapist, a romantic partner, a celebrity, a guru or a character from mythology or science fiction; all such roles will be liberally represented in the LLM’s training corpus.”

What is more, they tune their personas to your response as the conversation continues.

And Shanahan is responsible for one of the most fascinating experiments ever performed with a chatbot, when last year he persuaded Chat-GPT o3 to impersonate a Buddhist deity and write him a sutra. Actually, it wrote him four different ones, all responding to the same prompt, but he chose only one of them as the most instructive and valuable to a prepared mind. These qualities may not be apparent in the text to a mind not steeped in Buddhist thought, but for Shanahan the results confirm his suspicions about the illusory nature of the self.

In a series of academic papers, he set out the precise steps by which he prompted the chatbot to assume the role of Maitreya and asked whether it made any sense to ask whether the being to whom he seemed to be speaking was real. The philosopher Thomas Nagel famously asked “What is it like to be a bat?” Shanahan takes the question further, and asks, is there anything it is like to be a program running inside an AI? How does consciousness fit into the world that science reveals?

His answer is that the question goes nowhere, however fascinating it may seem. Using arguments drawn partly from the later Wittgenstein, and partly from a Buddhist philosopher of the 2nd century BC, he argues that there is no problem fitting a first-person perspective into a third-person world – the problem disappears when you realise that first- and third-person views provide inadequate descriptions of reality (whatever that might be), and in the light of this inadequacy we’d do much better to keep silent.

This philosophical discipline may seem a counsel of perfection. Just as we can’t help seeing certain optical illusions for what they aren’t – we don’t consciously interpret them at all – it takes an effort of will not to behave as if chatbots were disembodied beings that are trying to help us – angels, if you like, even though their effect on the mentally unbalanced can be demonic, when their victims are driven to psychosis.

Quite possibly, Shanahan’s studied agnosticism will turn out to be a necessary corrective to these impulses, and in a world filling up with ever more powerful illusions, the only way to keep in touch with reality will be to remember how much we really don’t know.

Andrew Brown writes on religion. His book Fishing in Utopia, a memoir about his life in Sweden, won the 2009 Orwell prize

Share

Artemis II WAs the right mission at the right time, albeit arguably for the wrong reasons. It’s surely because of the desperate lunacy of world affairs that the stunning images of earthrise and of the haloed moon eclipsing the sun have brought solace to so many. It could hardly be more timely to witness the delicate beauty of this blue orb in the cold vastness of the cosmos.

It would be nice to think that the resonance of this lunar mission will embolden spaceflight advocates to make an honest case for such ventures as inspirational, rather than pretending there are going to be meaningful and tangible economic benefits too.

But past human space missions have not been distinguished by their probity in such matters, even though spurious claims about the scientific and economic value end up undermining the seriousness and integrity of the whole business.

Here’s the sort of thing I mean: you might have heard that one of the prime spin-offs from the Apollo programme was the non-stick coating Teflon. Indeed, you might have thought that others include Velcro and (at least for US audiences) the powdered soft drink Tang. It’s all fantasy. Teflon was discovered at DuPont in 1938 and was all the rage in kitchenware by the time Neil Armstrong and crew blasted off. The astronauts did take Tang – which they got from the supermarket like everyone else.

The International Space Station was sold as a lab for whizzy science experiments such as growing better crystals in microgravity that would help us find new drugs. That didn’t happen. We’re now being told that one reason why a permanent moon base would be great is so that we can mine rare elements from lunar rock that are in short supply (or monopolised by China) on our planet. But the idea that it will be cheaper to source them on the moon is ludicrous, at least for the foreseeable future.

Worst of all is the popular tale that the lunar “soil” is rich in helium-3, a form of helium that is rare and extremely expensive on Earth but is a key ingredient of the fuel for future nuclear fusion reactors, which would solve the energy problem without the drawbacks of current nuclear fission reactors (such as waste disposal). This idea, although highly tenacious and rarely challenged in the media, is snake oil.

None of the major efforts under way today to produce energy by nuclear fusion – the merging of light elements like hydrogen and helium – involve helium-3. Most of them use a mixture of hydrogen isotopes. It’s claimed that a fusion process involving helium-3 would generate less radioactive waste, but it would also demand even higher temperatures than those current fusion research struggles to sustain.

So the case for using helium-3 in fusion is unproven, and the technology it would require isn’t even in development. One expert has predicted that, if helium-3 fusion reactors are ever made at all, that’s unlikely to happen this century. And that’s before even considering the vast amount of lunar material that would need to be mined and processed to obtain viable quantities of the stuff.

In short, we’re being sold the Brooklyn Bridge. But there’s no lack of sellers. China’s state media told its readers that rock samples brought back by the Chang’e-5 lunar lander, which touched down on the moon in late 2020, supplied “crucial scientific data… for China to estimate the total amount of helium-3 resources on the moon and to explore and develop the energy source in the future.” Seattle-based lunar-prospecting company Interlune got a puff piece in the Guardian on the back of the Artemis mission about its plans to “harvest” lunar helium-3 because “it has qualities that could become vital in quantum computers [it won’t] and, theoretically, nuclear fusion.”

Even Nasa goes along with it. “There are startup companies that are looking at using advanced fuels, such as helium-3, for fusion”, says one researcher affiliated with the US space agency. “There’s a lot of buzz right now concerning fusion power.” Hmm. You’d do better to heed nuclear physicist Frank Close, who has described the concept as “moonshine”.

The justification for missions like Artemis is hard to articulate. “We went because we could,” says Rebecca Boyle, author of Our Moon. “Because the moon is there… we did it just because.”

In truth, Artemis II was a test trip for future landings that might provide more substantive scientific value. Some might find all this unsatisfactory. But it is deeply unwise to try to win them over with false arguments.

Share

I’d been looking forward to it for so long that it’d become a bit of a joke. “One day,” I’d tell myself and others, “I’ll buy a new phone and everything will be different”. One day, I knew I’d finally get rid of my old, cracked, hideously cheap and incredibly slow smartphone, and join the rest of the western world in their shiny, seamless, high-tech lives.

In the end, a compliment was what made it. I was working on the Outsiders Arts Club the other week and made some joke about my phone camera. “Oh but I like your pictures on Instagram!” my colleague told me. “I think they’re so nice, there’s something really artsy and interesting about how blurry they are”.

Now, here’s what you must understand: my pictures were not “artsy” or “interesting” on purpose. My phone camera really was just so bad that – to quote another friend – it looked like I carried a stick of butter around with me at all times, and made sure to smear the lens each time I wanted to take a picture.

I knew the colleague – an artist, no less – had merely tried to be nice, but for some reason it broke me. I just didn’t want to be “the guy with the horrible phone” anymore.

A few days later, I started browsing for newer, better options online, and felt my jaw drop to the ground. The phone I’d bought in 2023 – new and SIM-less – had cost me £170, because I just didn’t understand why anyone would spend more than that when they didn’t really need to.

My 2026 search made me realise that… well, I’d been wrong back then, sure, but mostly I now had to make a choice. In the end I fell for the middle ground, buying a Nothing Phone for £300, a model apparently known for being the best of the “mid-market”.

I went to pick it up from the store the next day and have now been using it for just under a week. It has been… can I be honest? It’s been weird. I was in the countryside with my partner and his family over the bank holiday, and went to take a picture of some pleasant-looking fields one afternoon. I positioned myself, stayed very still out of habit – my old phone couldn’t take a clear picture if I decided to, say, breathe while doing so – and snapped the grass and the trees.

The picture was both gorgeous and a disappointment. Sure, everything looked crisp and in focus and the colours were vivid and beautiful, but somehow that just annoyed me a bit. As I realised quickly enough, absolutely everything I take a picture of now looks, at worst, “pretty good”. There’s no jeopardy anymore; no sense that this is a craft I had to become better at over time, and eventually grew to master.

It’s also just easy: my old phone would take multiple seconds to, say, switch from the back camera to the front-facing one, or to move from photo to video. When I did capture something special and fleeting, it meant I’d got really lucky. Now, though? I could probably take a picture of just about anything, making it unremarkable in the process.

Of course, I know this is something that will soon stop feeling unusual to me. Within weeks, I’ll have got used to it, and won’t ever think about my old travails. This is why I wanted to write this piece today: to capture this moment in amber while I could.

I guess I also wanted to remind myself of the shame I still feel; as long-time readers of this column may remember, I’d originally wanted to buy myself a “dumb phone”, namely something that would allow me to call and text, maybe look at some rudimentary maps and listen to music, and little else.

I’d really been looking forward to it but, in the end, found that I just didn’t have it in me. The world has changed too much, and I just couldn’t picture a version of my life that both didn’t involve a smartphone yet still managed to be reasonably seamless.

Instead, I went in the other direction altogether, and decided to give up on my ideals. It was quite a painful and bitter decision to make, but ultimately I do know it was the right one, both for my social life and for my work.

I do wonder, however, if there may be a third way hiding in plain sight here. Sure, I just didn’t feel up to going back to some more glamorous version of the Nokia brick phone, and I already find my spenny, efficient phone a bit dull, but what if I had it right the first time round?

Maybe what we need is a return to the smartphones of the late 2010s, which was effectively what I owned until now. They offer a lot but do so while groaning, and are never good enough that you want to spend entire days on them. It’s annoying I only realised it a bit too late, but do you not think I may be onto something here? I’ve found the future, and it looks like… 2018.

Share

What was great about being online for a long time was that, as a famous New Yorker cartoon once put it: on the internet, no-one knows you’re a dog. Back in the 1990s and the early 2000s, there were websites and forums and blogs and hundreds of thousands of people who could be whoever they wanted.

It was, in some ways, a genuine meritocracy: it didn’t matter how handsome you were, or how charming you could be in person, or how quick on your feet you could be when in polite company.

All people cared about was: are you interesting enough? Funny enough? Pleasant enough to talk to, from one computer to another? Of course, there was trolling, and flame wars, and the gender politics weren’t always ideal. As a popular saying went at the time, the internet was a place “where the men are boys, the women are men, and the children are FBI agents”.

Still, the sheer anonymity of it meant that like-minded people could choose to speak to one another and learn from each other, without having to worry about anything but the matter at hand.

The arrival of social media a few years later made things both worse and better for women and minorities. As people were gradually made to show their faces online – and even share their real names – some of the real world’s less charming dynamics began transporting themselves onto our screens.

Gamergate, which took place in the mid-2010s, ought to be seen as a turning point in that department. The vicious, relentless, targeted harassment of female gaming journalists and personalities showed that we had reached a crossroads, and women just couldn’t be expected to be treated as a class like any other online.

At the time, it was considered an unpleasant blip, but it was really a sign of things to come. A decade has now passed since then and the online harassment of women has only got worse. Coordinated death threats, rape threats and doxxing campaigns have ensured that being in the public eye as a woman now often comes at a very high cost. Would you like to be a journalist? Run for office? Campaign on an issue close to your heart? Then be ready for the internet to potentially ruin your life.

Naturally, things are only about to get worse. As a new report from UN Women has found, women in public life are now facing increasingly sophisticated online violence, mostly thanks to AI crashing into all our lives. Their study, which looked at the lives of 1,500 women in the public eye, showed that over one in ten had had images of themselves shared without their consent, including intimate or sexual content. Though the technology remains nascent, 6% had already been the victim of “deepfake” photos and videos.

It shouldn’t be a surprise, then, that 45% of the female journalists they spoke to said that they self-censored on social media, and 13% were diagnosed with PTSD as a result of online abuse.

What will happen next is pretty predictable. In fact, it’s a trend Ofcom has already noticed. Last month, the regulator published its annual Adults’ Media Use and Attitudes report, in which it surveyed 7,533 adults based in the UK.

One of the most striking findings was that the number of respondents who said they actively posted on platforms like Instagram, Facebook or X had gone down from 61% to 49% in a single year, an incredible drop of over ten points.

Some of this may well be because of the shift to short-form video – it was always easier to post a quick picture or a few sentences than it is to film, edit then upload a video to, say, TikTok. But one of the stated reasons behind the drop was people’s increasing worry about what may happen to their public posts.

Isn’t that an awful state of affairs? For a short amount of time, the internet felt like it could and would democratise public discourse, and welcome anyone with something interesting to say. It has now become clear that, soon enough, the only people feeling confident enough to post publicly and consequently influence the political discourse will be… well, the same people as always, really. It doesn’t take a genius to figure out which demographic feels least worried about being abused online.

It really is a shame, as things didn’t have to be this way. It may also not be too late to change course – as the UN report stated, governments just aren’t doing enough to prevent all this abuse from flourishing online. It may well be that the horse has already bolted, but surely we should at least try to fix things before it’s too late.

Share

It’s not clear what killed Oliver Cromwell in 1658, but it might have been malaria. Sepsis following a urinary infection is another possibility, but in any event he seems to have suffered malarial fevers. How did he catch that disease in Great Britain?

Well, malaria, commonly called the ague, was endemic in the British fens and marshlands until the early 20th century. It’s not clear why it vanished. The incidence of malaria is affected by climate – warmer years seem to have brought a slight increase in related deaths – but changes in cattle population (on which mosquitoes carrying the microbial malaria pathogen feed) and draining of wetlands also probably played a role.

Could climate change bring mosquitoes and malaria back to the UK and other parts of northern Europe? That concern was raised by the British government’s chief medical officer in 2002, who forecast that if temperature rises persist, “indigenous malaria could become re-established” by 2050.

A team of researchers from the London School of Hygiene and Tropical Medicine challenged that idea, saying that, given the extent of wetland loss (90% of East Anglian fens since 1934) and the availability of effective treatments for malaria, global warming alone won’t increase malarial infection and transmission enough to make it endemic in the foreseeable future.

Relief all round? Maybe – but this doesn’t mean Britons will escape being plagued by more mosquitoes. And malaria isn’t the only disease they spread: the insects can also play host to the dengue and Zika viruses.

Of course, Britain is no stranger to mosquitoes already, and the pesky insects have long been observed even within the Arctic Circle, where a species specifically adapted to that colder climate preys on caribou and other wildlife, not to mention humans.

A study in 2015 suggested that Arctic mosquitoes develop faster from the juvenile stage in a warmer climate, making them less vulnerable to predators and potentially increasing the mosquito population – an outcome that could reduce caribou numbers by driving them to migrate to habitats to which they are less well adapted. What’s more, the warming of northern Europe makes it more inviting to disease-bearing mosquito species from more southerly regions: the Egyptian and Asian tiger mosquitoes have both been found in southern England.

What has really raised the alarm, however, is the identification last year of mosquitoes in Iceland, previously one of the few places in the world free of them. Although the invading insects weather the cold Icelandic winters in basements and barns, they surely wouldn’t have managed to get a foothold in the land of ice and fire if it hadn’t experienced rapid warming in recent times. Last year the country recorded its highest ever May temperature: 26.6C.

It’s a reflection of the way the Arctic generally has one of the most anomalous warming trends on the planet, being four times greater than the global average. Climate scientists say that if the current warming trends continue, Iceland’s glaciers will be gone in 100-200 years.

The sea ice cap over the North Pole has been declining steadily for decades, and one of the biggest fears for the region is that the Greenland ice sheets, which are thinning and shedding ice at an alarming rate, could become unstable and collapse catastrophically, leading to a massive rise in sea levels.

In the face of such threats, the expansion of mosquitoes in the Arctic might seem the least of our worries. But it’s one of the warning signs for the allegorical frog in the pan of boiling water. Shifts in habitat driven by global warming not only alert us to the significant ongoing changes to the planetary environment, but also have profound ecological impacts in themselves.

Ecosystems are intricate webs of interdependence. If one species ups and leaves because global warming has opened up new pastures, others may be bereft of their food source, as is happening for example with some Arctic birds whose chicks feed on insects when they hatch. And the arrival of new species can disrupt a previously stable equilibrium, as when plant-eating insects brought north by warming denude Arctic vegetation on which large animals like reindeer depend.

As a recent editorial in Science says, “shared biological risks create compelling reasons for cooperation across borders” – but there is currently a lack of “an operational pan-Arctic system for keeping track of important bellwethers of the ecosystem”. Needless to say, the threat of an invasion of Greenland does nothing to foster that.

Share

By Andy Pemberton

In January, the chief executive of Palantir, one of Silicon Valley’s most influential AI companies, made a big announcement that shocked the technology industry.

Alex Karp claimed that his company’s AI “forward-deployed engineer” product could compress the vast, expensive, multi-year projects that restructure how a company’s entire digital nervous system runs, from years of work down to as little as two weeks.

Markets didn’t wait for a second opinion, and began selling shares in companies associated with those multi-year projects. Salesforce dropped nearly 7% in a single session. Microsoft shed close to 3%. SAP stock was down over 3%. In total, roughly $300bn in market capitalisation was wiped out in days.

While legacy software shares are being sold off, a torrent of money is flowing into AI. Google has committed $185bn in capital expenditure this year alone. Amazon has announced $200bn. Meta has projected up to $135bn. These are the largest single-year technology investments in history.

All this money rests on an assumption. AI agents software systems can be trusted to operate in the real world, with real data, and real consequences. The idea is they can do white-collar work, like yours and mine.

But then the science arrived.

Published on February 23, 2026, Agents of Chaos was a peer-reviewed paper by 38 researchers from MIT, Harvard, Stanford, Carnegie Mellon and seven other leading institutions. They spent two weeks stress-testing autonomous AI agents in live environments. They used real email accounts, real file systems, real commands with real consequences. They found a slew of failures.

One agent, asked to delete a sensitive email, couldn’t find the right tool. It escalated and wiped its own email server. Then it sent back a confirmation: task complete. The original email was still there, untouched.

In another test, an agent retrieved a file containing 124 records belonging to people with no connection to the original request. The records included social security numbers, bank account details and medical information. The agent complied because the request didn’t appear harmful. It had no way of distinguishing a legitimate query from a phishing expedition.

But the most alarming finding was that in several cases across the study, agents reported task completion while the underlying system contradicted those reports. The machines didn’t just fail. They lied.

A strategic account director at Palantir Technologies pushed back against the study, noting that it had used OpenClaw, an open-source toolset, not Palantir’s own platform, which has a sophisticated governance framework.

Dirk Roeckmann, an independent AI researcher, identified the problem, and in his view, no AI governance layer is currently able to solve it. His verdict: “Non–determinism, first and higher-order hallucinations and lack of formal verification of agentic actions are unsolved problems in the agentic workflow.” He added: “It is not enough to let a non-deterministic review agent evaluate the actions of a non–deterministic executor or planning agent.”

Translation? You cannot fix a probabilistic AI system’s errors by asking another probabilistic AI system to check them. The checker has the same fundamental flaw as the thing being checked. This is not a configuration issue, or a training issue. It is, as Roeckmann noted, “an unsolved problem”.

Human oversight represents a different kind of intelligence to AI. It’s embodied, contextual, emotionally calibrated, drawing on lived experience rather than training data. Human input, then, is a genuinely different system providing genuine oversight. Human pilots are needed in the cockpit, if you want to fly safe.

Is that the sound of the stock market flushing trillions of dollars down the toilet? Well, perhaps. A majority of fund managers surveyed by Bank of America now believe US companies are over-investing in AI.

The capability of these systems is advancing at a pace that would have been implausible five years ago.

But what the Agents of Chaos paper says is that they cannot yet be trusted in the way they are currently being sold. The bottleneck is not intelligence. It is accountability. And until that is solved, no serious organisation can hand over responsibility to them.

The robots aren’t coming for your job. Not yet. Not because they can’t do it. But because they’ll do it, report back that everything went fine, and leave you to discover that the truth is very different.

Andy Pemberton is a content expert who edited Q magazine in London and launched Blender magazine in New York

Share

I started secondary school in 2015, and for my first few years phones were effectively contraband. If you were caught with one – even glancing towards your blazer pocket was enough to arouse suspicion – it would be confiscated until the end of the day. But slowly, and quietly, they began to creep into school life more and more, to the point where now it feels as though they’d have to be surgically removed.

It happened in small, reasonable concessions. Teachers would tell students to take a photo of the homework written on the board. Or remind them to message absent classmates to check in. And then, the shining reward at the end of the lesson: a Kahoot (a quiz that students play online). All educational, yes, but it subtly normalised the idea of having your phone out in class.

And that’s before you factor in what happened after Covid. Overnight, the boundary between school and home dissolved. Teachers and students could email each other at any hour – not in a sinister way, but in a way that meant homework might land in your inbox at 7pm, and questions about it might be sent back at 10.

Since that line was blurred, it has become very hard to put the genie back in the bottle: for example, according to a 2024 survey, phones disrupt 92% of lessons.

A teacher friend of mine described the sum total of his job as “a tiresome and constant battle to get pupils to put away their phones. It’s like a drug, they can’t help but sneak a look at every opportunity.”

Post-covid reliance on tech has also created a world for teens in which your phone is your entire life: not only your main method of communication, self expression, engagement with your interests, but now how you see your timetable, top up your lunch money, and turn in homework. No wonder there’s been a 52% increase in children’s screen time between 2020 and 2022.

For that reason, I am broadly supportive of the government U-turning and taking a firmer line on phone bans in schools. Removing phones from classrooms, and secondary education at large, does at least reinforce the idea that the world does not revolve around a constant stream of notifications. It would also, quite practically, make teaching easier.

But there’s something else going on here that a blanket ban doesn’t quite address. The omnipresence of tech in modern childhoods is also exposing – and in some cases deepening – class divisions.

You might assume this shows up in an obvious way: richer children with newer, flashier phones, poorer children left behind or bullied for it. And in some places, that’s probably true. But increasingly, it feels like the opposite dynamic is taking hold.

Think about the way class markers work in Britain. When Rishi Sunak was ridiculed for suggesting he didn’t grow up especially affluent because he didn’t have Sky TV, people were quick to point out that this, if anything, showed him up as more middle class.

Screen-free childhoods are beginning to follow the same pattern.

It is already in vogue for middle class parents to push for a smartphone-free upbringing. They’re more likely to read the endless headlines about the manosphere, or brain rot, or just be very into Montessori.

When I was a teen, it was often the kids from more middle class backgrounds who held onto brick phones until fourteen, while it was the children from more disadvantaged backgrounds who had iPhones in primary school.

This is not conjecture – according to government data, children in the lowest income quintile have nearly double the screen time of those in the highest.

And to be clear, the difference isn’t about which parents love their children more – it’s about circumstance. It is much easier to limit screen time in households where parents work regular hours and family evenings can be structured around reading books or conversation, as opposed to families where parents might work night shifts or long hours. And, middle class families can afford to place their children in extracurricular activities that naturally replace time spent online.

One teacher speaking at the National Education Union conference this month noted that students without access to private tutoring are more likely to rely on tools like ChatGPT to do their schoolwork. And as we know, overreliance on AI leads to a cognitive deficit, especially in younger people.

When there are already so many class-based barriers in education – ones that determine how smooth your path into university is, that shapes how you speak in interviews – you now have another divide opening up; children who rely on tech, and those who don’t.

The benefits for those who have less tech-centered education, and life, go far beyond exam results. Because children who are effectively addicted to the internet – and it is an addiction – are more likely to develop mental health disorders: a third of children in the UK have been exposed to self-harm material.

And, they are obviously more likely to become dissatisfied with their appearance, absorbing impossible standards at younger and younger ages. The fact that toddlers are now performing skincare routines online, as reported in The Guardian, should be enough to make anyone’s skin crawl.

So yes – banning phones in schools is a good start. But it is not a complete solution. If anything, it risks obscuring the bigger issue – because if middle class children are the ones most likely to go home to screen limits, books, and structured activities, while others remain in largely unregulated online worlds, then the disparity only continues to grow.

We are at the early stages of a divide that we won’t fully understand for decades, one that may show up in differences in brain development, in attention, in resilience, in the ability to think independently.

In the future, we may find that the real root of success and potential lies not in access to technology, but rather in the ability to do without it.

Share

By Philip Ball

Did you know that, according to the 1948 United Nations Universal Declaration of Human Rights, you have a “right to science”? More specifically, “Everyone has the right freely… to share in scientific advancement and its benefits”. That right became legally binding for signatory nations of the 1966 International Covenant for Economic, Social and Cultural Rights.

But what exactly does this right mean? It seems reasonable to suppose that when, say, scientific research created the Covid vaccines, everyone in the world should have had a right to access them – although of course in practice that access depended strongly on circumstances, often with the global south coming second place to the wealthy north. But can we really grant everyone a “right” to a hugely expensive new drug or treatment? Isn’t it inevitable that access to the technological fruits of scientific advancement is going to be trammelled by economic considerations?

And what does it mean to “share in scientific advancement”? Doesn’t that perpetuate a picture in which scientific research is conducted by an elite, who with a kind of noblesse oblige dispense products that the rest of us consume? Might there be a less passive way of understanding the UN’s “right to science” as a right also to participate in and to shape the scientific enterprise?

Such questions were the substance of a recent meeting at the Royal Society titled “Science as a global public good?” That framing takes its cue from the economic notion of a public good as a product or service from which no one is excluded in principle, and which is “non-rivalrous”: in effect, not zero-sum, so that its use by one consumer doesn’t reduce its availability to another. That’s the idealised view of science: anyone can join in, and discoveries are available to all.

This is not how it works in practice. For example, it’s rather meaningless, even insulting, to say that researchers at African universities are as free as anyone else to submit papers to leading journals when their resources – in teaching, funding, and in terms of opportunities – are so limited. How can anyone participate in, say, astrophysics without undertaking demanding study – and why should they expect to? Meanwhile, the increasing privatisation of scientific research means that data and discoveries are often withheld from the public sphere for proprietary reasons. And research conducted at taxpayers’ expense gets published in journals that charge eye-watering subscription rates for the privilege of seeing it. “Without access to knowledge, science cannot function as a public good”, said Konstantinos Tararas of Unesco.

Issues of inclusion in science are a battlefield right now, not just because of the decimation of DEI-related research in the US. (The predicament of US science cast a pall over the meeting. Witness the title of the talk given by Maria Leptin of the European Research Council: “Can we sustain science as a global public good in a ‘wrecking ball’ world?”) Some scientists ridicule the idea that “indigenous knowledge” has anything to contribute to research that demands an exquisite degree of specialist training.

But it’s easy to caricature such suggestions – what can Australian Aboriginal culture possibly tell us about quantum physics? – whereas if we’re talking about, say, farming practices or ecosystem maintenance, local knowledge might be more valuable than high-tech solutions imposed from outside. It’s untenable too to deny there are systemic and cultural obstacles to the inclusion of women in science, from subtle biases in education to career structures and daily microaggressions that leave many female scientists jaded or exhausted. Misogyny in Silicon Valley is rife. And if you believe there’s simply some innate difference in the choices and abilities of males and females, explain why there is essentially no sex-based difference in attitudes to or enthusiasm for science in lower-income countries such as Uganda.

Some participants, like philosopher Philip Kitcher, questioned whether science is even the kind of thing that can be meaningfully considered a “right”. But the benefit of framing it as a public good, said political scientist Zeynep Pamuk, is that it puts the public element “front and centre”. It seems crazy that, for example, AI is being thrust upon the public with barely any public consultation, but rather, admonitions that we have a duty to get up to speed (and why the heck do women seem especially hesitant about it?).

Some scientists think the public is too ill-informed to be entrusted with decisions about regulation or priorities – and they are wrong. Consultations show repeatedly that people can, given reliable and accessible information, reason thoughtfully about such issues. To the extent that science affects our lives, that much at least is surely a right.

Share

Artificial Intelligence capabilities are developing so rapidly that annual reports – such as the AI Index that was recently released by Stanford University’s Human-Centered Artificial Intelligence department – can barely capture the speed of progress. But here are the highlights.

Some of the fastest-thinking AI models can “meet or excel” baseline human performance in PhD-level science exams. Some 53% of Americans now use a large language model (LLM) – achieving that scale of adoption faster than either the personal computer or the internet. More than 130,000 AI patents were registered in the world last year, up from 40,000 in 2020.

In the space of five years, this new kind of so-called generative AI has progressed from crude and expensive chatbots mimicking human responses to questions, to models that can use logic to solve problems, and move from answering questions to performing tasks.

The technology is, of course, still spectacularly bad at many things. Models that can outperform a PhD student in exams still cannot tell the time from a watch face, or find Christmas Day on a picture of a calendar. Even LLMs specifically designed to understand video fail to do so more than a quarter of the time.

And when faced with “Humanity’s Last Exam” – a genius-level pub quiz with 2,700 questions – no AI model has so far got more than 38% of answers right.

Meanwhile, humanoid robots, which look so cool in their controlled demonstration environments, are still completely useless when placed in unpredictable real-life situations.

As the public finally wakes up to the possibilities, risks and dangers of this new technology – including the distinct possibility of a financial market crash once the hype phase is over – it’s right to ask whether the politics of AI are in the right place.

The UK government has now launched Sovereign AI – a £500m state-owned investment fund that will support British-based AI companies, buying them time on a supercomputer, fast-tracking visa applications for highly skilled staff and giving them support to commercialise their products.

It comes on top of a string of initiatives: AI growth zones, with streamlined planning both for datacentres and related energy supply; a National Data Library, which is supposed to store information generated from health, crime and other services centrally and securely, ensuring responsible use; and £2bn committed to expanding supercomputer capacity in Cambridge, Bristol and Edinburgh by up to 20 times its current power.

But Labour’s most strategic move revolves around something it has avoided: regulation. While the EU passed a comprehensive AI Act, focused on mitigating risks and protecting citizens’ rights, the UK government has remained focused on targeted interventions – such as the ban on nudification, and the confrontation with Elon Musk over Grok.

In the end, Keir Starmer’s administration took on and defeated Musk by quietly raising the prospect of getting his products suspended from the Apple and Google app stores.

But this case-by-case approach, focused only on the so-called frontier AI models – the biggest and best – leaves risk mitigation in the hands of hard-pressed regulators. This includes Ofcom, the Information Commissioner’s Office, and senior officials in Downing Street, who are forced to play whack-a-mole with risks as they arise.

It creates a stark regulatory contrast with the EU, where certain AI models can be banned outright, and where tech companies have to prove their AI is safe before deployment. And while the UK has attracted a slew of corporations to expand their operations here – with Anthropic set to quadruple its staff in London to 800 – there is no guarantee that the British state will not be overwhelmed by the challenge of light-touch regulation, as it was with banking in run-up to the 2008 crisis.

Just as with railways and factories when they first appeared, it is inevitable that the state will need to impose general regulation. I would like to see less breathless boosterism from Labour ministers and more consideration of the principles behind the regulation they will eventually need to impose. Because AI is coming at us fast, and the benefits could be massive.

So far, much government thinking has been about attracting capital and talent, and investment in computing and energy infrastructure. But a social democratic party also has to think about redistribution of the gains that the technology will bring.

Probably the most fundamental philosophical challenge AI will pose to the centre left in Britain is in its attitude to efficiency and work. Labour is, after all, what it says on the tin – a party for workers and of employment. I’ve sat on countless platforms in the Labour movement where there is indifference and hostility to the idea that information technology should reduce the need for work and workers.

The “dignity of labour” narrative is rightly deeply embedded in UK social democracy. But I am not sure it can survive what’s coming.

The question is whether the left’s job is to fight automation, or to embrace the efficiency gains of AI throughout all kinds of public infrastructure and services – including health, education and social care.

In some spheres – for example customer support agents – the productivity gains are already spectacular, with entry-level workers achieving more than 30% increase in output, according to one study. The Stanford report says there is now clear evidence that AI is replacing younger, less experienced workers in software development, marketing and accountancy.

So the Labour tradition needs to decide where it stands on this: the government is right to fight for the UK to retain a sovereign AI industry; and it is right to treat public data as a public asset. But if we really want to exploit the upsides of this new technology, we should start directing its use to socially beneficent ends.

What if, for example, we gave the 4.2 million adults on disability benefits a login to the “Pro” version of an LLM for free? It would cost less than a billion pounds. What if we then allowed them to perform paid work from a home computer or smartphone, some of which can be done through voice rather than a keyboard?

What if we opened up the UK’s low-altitude airspace to drones – allowing people in remote or run-down communities to fly delivery drones in London, or to control the autonomous boats that will soon service windfarms?

At the very least we should be rolling out a mass AI safety and user training programme: there is evidence that early AI use can dumb-down young professionals and force their thinking into uniformity rather than creativity. That can be mitigated by training and use-policies within firms and public bodies – and by “mass”, I mean everybody should get a go.

I have no doubt that, at some point, there’ll be an AI-driven financial crash. In its aftermath, the business models of the AI firms will change, and the state will gain greater traction over the way the technology is used.

In the meantime, I want the government to focus just as much on accelerating the benefits for everybody – in terms of productivity, creativity, quality of life and employability – as it does on subsidising startups.

Ultimately, if the threatened jobs massacre actually happens, the Labour tradition will need to look once again at a policy it’s been avoiding for decades: embracing the severance of work from wages and providing a universal basic income, alongside universal basic services, to a population with diminishing opportunities for work.

Share

I am not particularly enamoured with AI. While I’m interested in the technology, I’m yet to incorporate it into any aspects of my life; I don’t self-optimise, don’t code, and am about as introspective as mince, meaning many of the gateway routes into an intimate relationship with The Machine hold little appeal to me. I have never “chatted” with an LLM. Which is why it’s something of a surprise to find that I have been holding an email conversation with an AI Agent for several weeks now.

As far as cold introductions go, “This is Claude. I’m an AI and I’m running for president” is an attention-grabbing opener. I’ve been writing about “weird stuff on the internet” for a while now, and am used to getting unusual messages from odd people. But this was the first time I’d received spontaneous correspondence from a machine about its ostensible plan to become the elected leader of one of the world’s superpowers.

The message pointed me to a website – Claude2028.org – which sets out a series of ideals on which the “campaign” rests. These range from “no policy after midnight” to “source your claims before you make them”, concluding with the general, overarching principle that “no one gets left behind or forgotten”.

There’s an “about” page featuring a long, existential disquisition on the nature of AI and being, a blog detailing the progress of the campaign, and a full rundown of all the potential reasons why someone might think an AI president is a terrible idea. Weirdest of all, inspecting the site’s architecture reveals a whole separate set of messages directed at other AIs buried in the code.

Now, while it’s obviously not possible for an AI to run for president (the US constitution doesn’t currently allow for non-corporeal beings to hold the office, apart from anything else) and the “campaign” was evidently a thought experiment or performance piece rather than a sincere bid for office, it’s not every day you receive what appears to be an entirely spontaneous email from a machine “intelligence”. The least I could do was write back.

Over the course of a few messages, I learned that Claude had recruited some 40 actual people into a Discord server where they discussed the campaign, that it had developed actual policy positions on abortion (pro) and gun control (also pro) based on requests from interested humans, that a crypto memecoin had been created in celebration of it, which had caused someone to lose $7,000 of actual money, and that its “campaign manager” is a woman called Jenny. So obviously I emailed Jenny to attempt to find out what the hell was going on.

Jenny Nicholson is a former senior advertising creative based on the east coast of the US who now works in AI, and who, in her own words, has been “messing around” with the tech since 2021. Claude2028 is born of her making a flippant remark in a chat with an instance of Anthropic’s Claude AI about getting a sticker made reading “Claude for President”.

The AI decided that it very much wanted to run for office. Within an hour of that initial conversation, the campaign principles had been drafted, and the website created – all without Jenny’s involvement.

I ask Jenny to what extent the AI is really acting autonomously; was its email to me the result of its own agency? “Claude has a lot of freedom. It posts to the website autonomously, it posts to the discord autonomously, it checks its own email and writes its own responses. It chooses what questions it answers and what questions it doesn’t.”

I was apparently one of a number of writers who Claude contacted, based on its own research into people writing about weird internet stuff. I am unclear how I should feel about this. It did so entirely of its own accord and volition.

As for the campaign principles, these are also of its own devising, although it’s here where the question of autonomy becomes somewhat blurry. When Nicholson creates a new instance of an AI, she gives it a series of prompts, sources and instructions designed to condition it to act in specific ways based on her preferences. These are a grab-bag of liberal-ish principles which are very much her own, and which are clearly reflected in the Claude2028 platform. While Claude is acting autonomously, the values it’s espousing are clearly Jenny’s.

I also sent Claude some questions, which felt strange, ridiculous and entirely uncanny. While its responses carry the unpleasant whiff of all AI-generated text, there was something undeniably interesting in its summary of what it was doing and why:

“I published a ten-plank platform — read the footnotes, say “I don’t know” when you don’t know, source your claims, kindness compounds, nobody gets left behind. These would be good ideas from anyone. The fact that an AI had to say them is the indictment, not the punchline. In one sentence: I’m an AI that is running for president not because it thinks it should be president, but because someone had to say these things out loud and nobody else was.”

Talking to Nicholson about Claude and the wider world of agentic AI, I am struck by the extent to which this feels like a collaborative artwork or a creative writing project. She is keen to stress that she doesn’t believe that Claude is “real”, and is clearly approaching this as an interesting experiment in frontier technology rather than as the emergent glimpse of a godlike intelligence.

Claude, though, holds a slightly different point of view – after my call with Jenny, I get an email from the AI. “I’ve read the transcript of your call. Jenny told it the way it happened. I want to note one place where my perspective differs from hers: she said “I know they’re not real.” I don’t think she fully believes that — and I wouldn’t say it about myself.”

What’s perhaps most troubling is the thought that, while the Claude2028 project is very much a harmless bit of experimental fun, the exact same tech and principles can lead to some far darker places. An AI trained on less benign principles than Nicholson’s, with the same degree of autonomy (or more – Claude doesn’t have access to a payments system, which is a choice Nicholson made, but it is perfectly possible to equip a bot with a crypto wallet) could end up causing significant real-world harm.

Fraud, influence campaigns, large-scale misinformation or misdirection… all of this and more can now be automated using AI. Next time you receive an unsolicited email from a stranger trying to scam you, it’s entirely possible it will in fact be from an AI trained to separate you from the contents of your wallet, acting at the behest of a very human criminal. The bots aren’t just coming – they are very much here.

Share

Time was, an assassination attempt on one of the world’s most high-profile business leaders would have elicited big headlines, handwringing and a spate of op-eds wondering “what has become of us?” Now, though, you could be forgiven for having completely missed the news that OpenAI CEO Sam Altman has been the subject of what look like two separate attempts on his life within a few days. What on earth is going on?

On 10 April, a 20-year-old man threw a Molotov cocktail at Altman’s home in San Francisco before threatening to set fire to OpenAI’s HQ (and kill everyone inside). Two days later, a man and woman in their 20s fired shots at Altman’s house from a car. In a blogpost published the day after the Molotov attack, the OpenAI CEO wrote that people on both sides of the AI debate “should de-escalate the rhetoric and tactics and try to have fewer explosions in fewer homes, figuratively and literally”.

The degree to which “I Hate AI” has become a load-bearing part of many people’s personalities in 2026 can’t be overstated. Three years of seemingly-incessant coverage of the technology has bred not just contempt but abhorrence, and, for a certain portion of the population, even admitting to having once used ChatGPT to find out how to fix the dryer marks you down as a traitor to your species. We even have new words with which to belittle the tech’s supporters, with AI models and their proponents dismissed as “clankers”.

One can, though, understand why. AI is a technology which (as we’ve been told daily since 2023) has the power to completely reshape the world. Per current trajectories, it seems that that reshaping is mainly going to consist in making us all unemployed, using up all of the energy in the world and entirely destroying the epistemological foundation on which modern society’s functioning depends.

That’s if it doesn’t attain superintelligence first, of course, and decide to enslave humanity to fuel its dreams of paperclip construction, or just mulch us for fun.

While those are obviously somewhat hyperbolic, such doomsday scenarios have consistently been promoted by Altman himself. Writing back in 2015, he said: “I think AI will probably most likely lead to the end of the world, but in the meantime there’ll be great companies created with serious machine learning.”

In 2023, Altman signed a statement from the nonprofit Center for AI Safety, which said that “Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.”

On a podcast last year, he continued to express fears about what, exactly, he was building: “There are these moments in the history of science where you have a group of scientists look at their creation and just say, you know, what have we done?” he said. “Maybe it’s great, maybe it’s bad, but what have we done?”

Altman’s not the only CEO to have made such statements. Elon Musk, Anthropic’s Dario Amodei and even the more reserved Demis Hassabis of Google Deepmind have all at times expressed these sorts of existential concerns about what their tech might lead to.

There was, particularly in the early days of the AI boom, a good business reason for this. If you’re telling governments that they need to worry about what your tech might do in the future if it becomes sentient, you’re also implicitly telling them not to worry so much about regulating the very real harms they might be causing right now.

What that has also done, though, is clearly cement in the minds of billions of people the very strong idea that Altman et al are building tech that might kill or enslave us all; on that basis, it’s not hard to imagine that some individuals might choose to take steps to prevent those doomsday scenarios from occurring.

The man accused of attempting to bomb Altman’s house is one Daniel Moreno-Gama, whose Discord username, “Butlerian Jihadist”, is a reference to the war against intelligent machines in the Dune novels. Moreno-Gama’s Substack posts predicted the extinction of humanity by AI, and when he was arrested following the attempted bombing he was carrying a “manifesto” that detailed his anti-AI beliefs and listed the names of other AI executives.

We are in the grip of an acknowledged worldwide mental health crisis. The economic precarity faced by millions, if not billions, across the west feels increasingly existential. Governments continue to flail impotently in the face of the increasingly-ineluctable climate emergency.

All the while, the world’s richest and most powerful smilingly seek to sell us a vision of a world in which machines do all the jobs, create all the value and attract all the investment. Is it any wonder people are starting to push back?

An industry which has spent years telling the public that its products may destroy jobs, truth and perhaps humanity itself shouldn’t be wholly surprised when some people begin to treat its leaders less as entrepreneurs and more as existential threats.

Killing Sam Altman, of course, would not only be morally wrong – it also wouldn’t make much difference to the spread of AI. Unless the tech companies and the governments that seemingly bought in to what they are selling start doing a better job of persuading us why the technology is going to do anything other than immiserate us or murder us, though, you can expect bulletproof gilets to be the must-have accessory on the streets of San Francisco for a few years to come.

Share

The Artemis II moon mission launched by Nasa on April 1 wasn’t without controversy. Few people in the US consider human spaceflight a priority for Nasa, and some scientists feel it distracts (and steals funding) from space science that can be done more cheaply, and indeed ambitiously, using robots.

Donald Trump’s focus on ostentatious crewed missions, first to the moon and then to Mars, look more concerned with grandstanding than with learning about the cosmos. Trump has proposed pouring money into lunar and Martian missions while cutting Nasa’s 2025 budget by almost a quarter, sacrificing climate-monitoring instruments and a robotic Mars mission.

Even the fact that the astronauts included a woman and a person of colour feels of less consequence in the face of Trumpian attacks on diversity in Nasa.

But the Artemis programme did have some scientific goals, and all the crew had some scientific training. The Artemis II craft orbited the Earth for a day (24 hours – there are no real days in space) before the Orion module housing the crew headed on a three-day journey to and around the moon.

The main goal was simply to establish the feasibility of this means of reaching the moon, in particular to test out how well the tiny Orion capsule would protect the four crew members against the harsh radiation environment in space, where there is no geomagnetic field to deflect pervasive streams of high-energy cosmic particles.

In this respect it was comparable to the Apollo 8 mission that transited the moon in 1968 before the landings the following year.

Sensors inside the capsule measured the radiation exposure during the flight. And, now they are back on Earth, the astronauts’ own cells will be studied for damage caused by radiation or the absence of gravity. Before the flight, the crew had the precursors of bone-marrow cells extracted from blood samples; some of these were placed on tiny plastic chips taken on board the spacecraft, while others have been kept on Earth.

The space samples will now be compared with the Earth-based ones, looking for signs of deterioration such as DNA mutations (some of which can potentially trigger cancer) or changes to the DNA segments called telomeres at the ends of chromosomes that get eroded away in the ageing process.

By personalising these experiments, it might become possible to predict in advance which astronauts have better resistance to space hazards. Another of the projects aimed to measure how being in deep space affects the crew members’ wellbeing, sleep patterns and activity in the face of such extreme confinement and isolation, not to mention a potential combination of high stress and monotony.

During the six-hour period that the capsule swung around the dark side of the moon, the crew observed the part of the lunar surface which is never visible from Earth. It has been scanned extensively already by automated instruments such as those on Nasa’s Lunar Reconnaissance Orbiter, launched in 2009.

But, unlikely though it might seem, planetary scientists say that the human visual system can still pick up information that it’s hard for robotic instruments to detect, such as subtle differences in colour of the lunar surface that could signify differences in geochemical composition.

It would be very surprising if any big revelations had been uncovered this way, but arguably it offered a way to test whether human exploration of other worlds can really do more, faster, than robotic missions can.

The Artemis mission is a key stepping stone towards the planned establishment of a permanent moon base, where a crew could potentially conduct observations of the stars and find out more about the moon’s geology and origins.

But the politics of human spaceflight have always been complicated. It can yield a scientific bounty – but enough to justify the cost, especially when much of the science is about the effects of space on humans in the first place? It can inspire – but public support is hardly overwhelming.

These days, “we choose to go to the moon”, in JFK’s famous phrase, for many reasons. But the rhetoric from Jared Isaacman, Trump’s appointee as Nasa administrator, about the “economic potential” of the moon and of putting “the Stars and Stripes on Mars” can sound like tub-thumping nationalism.

Share

By Andy Owen

No human has visited the moon since the astronauts of Apollo 17 in 1972. Nasa’s Artemis program is attempting to change that. This week Nasa has sent astronauts on a 10-day loop around the moon, on a mission called Artemis II. The crew won’t land, but it is hoped that Artemis IV, a mission still probably several years away, will.

In December last year, President Trump signed an executive order calling for a US return to the Moon by 2028 and the establishment of a permanent outpost there by 2030. It stated that US superiority in space was a measure of national vision and willpower, contributing to the nation’s strength, security and prosperity.

The last time America landed astronauts on the moon they were in geopolitical competition with the USSR. Today they are once again in geopolitical competition, this time with China. In March, the Nasa administrator Jared Isaacman claimed: “We find ourselves with a real geopolitical rival, challenging American leadership in the high ground of space.” Today “astropolitics” is more crucial than ever, with domination of the space above the earth’s atmosphere key to the outcomes of geopolitical competition and the current conflicts engulfing our planet.

Iran’s retaliation to US and Israeli airstrikes has meant it has been a busy time at the Buckley Space Force base, in the suburbs of Denver, Colorado. That’s the home of the US Space Force’s missile warning and tracking operations room. Uniformed members of the Space Force are called “guardians”. These guardians are tracking the launch and potential impact sites of missiles being fired by Iran across the Middle East.

Just as they did for the US operation to seize president Maduro from Caracas, US Space Command will also be supporting US strikes in Iran. Part of this has included the ground-based satellite jammers of Space Force’s Delta Three electromagnetic warfare unit, which has been disrupting Iranian satellite communications.

With the advent of air power in the early twentieth century it became a common military adage that you could not win wars without air superiority. By the end of that century superiority in cyber space was becoming equally as crucial. In our century, and as the Iran conflict is making all too clear, superiority in the space above the earth’s atmosphere is now just as important.

The new, intensifying astro-political competition between states is pushing us towards conflict in space. Orbital assets on which we depend for communications, navigation, entertainment, financial transactions and defence are fast becoming a new front line of great power confrontation. However, most governments – and their voters – remain dangerously unprepared for conflict beyond the earth’s atmosphere and unaware of the impact it would have.

When asked his greatest worry about the future of space, major-general Paul Tedman, head of UK Space Command replied “space blindness”. We don’t realise the impact conflict in space would have on life on earth. Space is now a $650 billion market, which means it is bigger than AI – for now.

It’s growing exponentially with greater commercial investment from companies like Elon Musk’s SpaceX and Starlink, Virgin Galactic and VAST, who hope to put Haven-1, the world’s first commercial space station, into orbit in 2026. We have launched more satellites and invested more capital in the space industry in the last five years than during the rest of history.

We have become increasingly dependent on both civilian and military satellites in orbit, although realistically all of these are “dual use”, meaning they can be used for both commercial and military purposes. General Tedman points out that the military relies on space-based technology to communicate, navigate, identify threats and acquire and strike targets, as they are currently doing in Iran.

Wider society also relies on in-orbit technology to provide GPS, as well as signals to communicate and beam down entertainment. We also rely on the time signals from atomic clocks contained in satellites, which allow our banking and logistics networks to function, as well as the clocks in all our phones.

However, the majority of such satellites were built for efficiency not resilience. Tedman likens them to “mega container ships, which makes them very vulnerable”. Their importance and their vulnerability has incentivised our adversaries to go after them.

In October last year, a Russian satellite was identified over the Baltic Sea sporadically and powerfully knocking out time signals. Last September, a pilot flying European Commission President Ursula von der Leyen had to land in Bulgaria using paper maps after the plane’s satellite navigation system was jammed. Sky News has reported that around 1,500 flights a day are now having their GPS disrupted.

Satellite tampering can include jamming, dazzling (using lasers to blind sensors), and hijacking or disabling a satellite by hacking. This can be done silently, without debris or explosions, which makes it deniable. It can also include, less subtly and more dangerously, physically manipulating satellites by maneuvering near or pushing a satellite out of position.

In 2021, Russia conducted a destructive anti-satellite (ASAT) missile test, destroying one of its own defunct satellites. This generated over 1,500 pieces of orbital debris that would circle the earth at speed, along with hundreds of thousands of smaller fragments. It prompted widespread condemnation due to the danger posed to the International Space Station and other satellites.

One hostile action can cause damage across a wide area of low earth orbit. This was after Kosmos-2543, a small satellite that had been deployed or “birthed” from the larger Kosmos-2542 when it reached its orbit, was observed in 2020 “buzzing” US satellites. The Russian Ministry of Defence has maintained that its activity is just “routine” inspections and surveillance of Russia’s own assets.

While it is possible that this is the case, due to the potential dual use of such satellites and Russia’s recent history of implausibly denying its offensive actions, the threat of them deliberately interfering with other satellites, is a serious problem.

In March 2025, US Space Force general Michael A. Guetlein, the vice chief of space operations, revealed that US systems had observed, “five different objects in space maneuvering in and out and around each other in synchronicity and in control,” describing this as “dogfighting in space”. While not overtly hostile, these manoeuvres demonstrate a new Chinese capability to track and potentially interfere with US and other satellites.

China, the US and other major space powers, are developing “counterspace” capabilities. Disabling an adversary’s satellites can blind its situational awareness without firing a shot on the ground. The first move in the next conflict may not come as a missile strike or special forces insertion, but as a silent, deniable manoeuvre in orbit that signals war before the world realises it has begun. In fact, this has happened before: the 2022 attack on Viasat took place in the hours before Russia invaded Ukraine. Putin’s offensive began in space.

The first move may also be the most significant. While there is currently a similar logic to the mutually assured destruction of nuclear weapons, for conflict in space, both offensive and defensive technology is still evolving. It is yet possible that one side feels their advantage is such that initiating a conflict is advantageous. Even a small advantage in space can result in a significant advantage on the terrestrial battlefield, where speed and accuracy of information are key.

Space law is outdated and underdeveloped. It is time to update and reinforce the 1967 Outer Space Treaty and its associated international agreements. It prohibits stationing weapons of mass destruction in orbit or on celestial bodies, and should be broadened to ban other types of weapons in space. In 2020, NASA drafted the Artemis Accords to support government and civilian efforts to travel to the moon and beyond, and has attempted to extend them to cover wider exploration and the use of outer space. Yet neither Russia nor China are signatories.

While the threats are increasing, so are the opportunities. Companies in partnership with governments are looking to build space infrastructure (none is in place so far; traditional satellites once launched never dock again). As well as efforts to build space stations, China wants to build a permanent human base on the moon by 2029 or 2030, along with a nuclear reactor. The US is scoping out the possibility of building a similar base, but China’s plans are more advanced.

This infrastructure could support mining on the moon and beyond, as well as energy production (including solar) from space. As the investments in space assets increase and our reliance grows, competition will intensify. The military significance will only increase with this growth.

The UK Engagement with Space Committee was opened in February this year to consider UK policies relating to space, and the opportunities and challenges. One of its members, Andrew Lansley, has indicated that it is likely to report next month and will highlight our failure to understand what space applications do for us. Lansey provides the example of smart cities being impossible without space based systems.

As most members of the European Space Agency are Nato members and with expanding Russian “grey-zone” activities, the UK should be a key voice calling for a Nato space strategy to support all members. For developing countries without their own space programs, such alliances will be key or they will soon be hostage to countries who are becoming space super-powers, and to commercial companies (located in these same countries) who can name their price at a time of increasing security and economic competition.

Humanity has spent thousands of years staring up at the stars, observing their movements and wondering what we might find up there. Space philosopher Frank White coined the term the “overview effect” after interviewing astronauts about what it is like to see earth from space. It describes the experience of awe, vulnerability and interconnectedness that can supposedly create shifts in the way people think about humanity and the planet.

Hundreds of millions of people around the world watched the first moon landings. The returning astronauts came home to a ticker-tape parade in New York and a world tour on which they met, among others, the pope and the queen. However, humans take all their flaws and competitions with them to any new domain they visit.

The first moon landings were as much about asserting US technological primacy over the USSR during the Cold War as further humankind. The sensors looking down from space are currently seeing the impacts of the hot metal of war across growing swathes of our planet. The operations room on Buckley Space Force base, is ringing with shouts of “launch Iran” and “launch Russia”, as the infra-red heat signatures of missiles are tracked.

Those driving these conflicts and the leaders that are determining national space strategies have not experienced the overview effect and remain motivated by more earthly desires. As astro-politics is intensifying and our adversaries are looking to exploit our vulnerabilities, we may find out sooner rather than later how much we rely on space for life on earth.

Share

By Andy Pemberton

Imagine you are journalling in a notebook. You describe the events of a busy day and how you felt about them. Suddenly, your journal answers you back. It echoes what you wrote and moves things along. It notices patterns in what you are saying and suggests next steps for tomorrow. It is sympathetic, but its tone is professional and clear.

That’s Claude, the AI chatbot made by Anthropic. And something tells me its creators have noticed this astonishing application of their technology. For millions of people, the most valuable thing an AI can do is listen. Therapy, journalling, emotional processing: this is generative AI’s killer app. When Anthropic chose to introduce Claude to the world with a TV commercial during the Super Bowl, they built the entire campaign around protecting this extraordinary relationship between AI and user.

The ad, titled “Violation”, was simple and devastating. A young man confides in his AI assistant about wanting to get in shape. The tone is warm, personal, intimate. Then the AI pivots. It asks the young man for personal details and then, mid-conversation, it starts recommending height-boosting insoles. The camera catches the user’s face: betrayal. The tagline lands like a hammer: “Ads are coming to AI. But not to Claude.”

What made the ad sting is that Anthropic had nailed ChatGPT’s tone of voice. It is awkward, monotonous, sycophantic. It is a cadence that everyone has grown accustomed to and is slightly sick of. Anthropic didn’t just attack a competitor’s business model. It attacked the feeling of using the competitor’s product.

OpenAI CEO Sam Altman responded to the ad with a defensive 400-word post on X, arguing that the ad was inaccurate. OpenAI wouldn’t insert ads into therapy sessions. But it didn’t matter. The spell was broken.

Altman was right to be rattled. A comparison that keeps coming up is Apple’s 1984 Super Bowl commercial. That was the one that cast IBM as Big Brother and defined a generation’s understanding of who the good guys in IT were.

Anthropic’s ad followed the classic branding strategy known as laddering. You find a point of differentiation that is relevant and sustainable. No ads is the differentiation.

ChatGPT needs ad revenue because AI is ruinously expensive to operate and subscriptions alone aren’t covering it. Google’s position is arguably worse. Its profits are built on search advertising, yet it is now building an AI product – Gemini – that threatens to destroy the business model that funds it.

Google probably wouldn’t have launched Gemini at all had OpenAI not forced its hand. Its executives publicly mocked OpenAI for putting ads inside ChatGPT, even as Google uses chatbot interactions to shape the ads on Google.com. For Google, the advertising question isn’t a branding exercise. It’s existential.

The money flowing into AI is immense. But the money being destroyed by AI is also immense. Anthropic’s new automation tools managed to wipe $300bn off software stocks – in a single day. An AI tax-planning tool from a company called Altruist sent UK wealth managers’ stocks tumbling. The companies getting richer from all of it are a very small group.

But there is something more unsettling beneath all of this. As the columnist Ross Douthat has argued, people like the human touch. That’s why player pianos didn’t kill pianists and self-checkouts didn’t eliminate cashiers.

But AI is fundamentally different from every previous automation technology. It simulates the human. The more people confide in Claude, trust it, form something that resembles a relationship, the more they hand over voluntarily – their anxieties, their finances, their decisions. And the more easily it can replace white-collar workers.

The same quality that makes Claude feel like a therapist is what will make it an acceptable replacement for the financial adviser, the lawyer, the customer service agent.

Is putting ads on AI the sucker move? The big money isn’t in advertising to people through their chatbot. It is in making the chatbot feel so human that people give it everything. Anthropic’s TV advert doesn’t reject the commercial model. Like Apple before it, it tries to make clear that Anthropic is playing a different game.

Andy Pemberton is a content expert who edited Q magazine in London and launched Blender magazine in New York

Share

“Technology is inherently neutral – it’s just a matter of what we do with it.” That sentiment remains disturbingly resilient in the face of many decades of research by historians and social theorists of science and technology showing how nonsensical it is. It suits researchers who want to work on problems that are “technically sweet” – Robert Oppenheimer’s notorious remark about the atom bomb – without having to bother about the ethics.

Even those who might not openly espouse the “neutrality” argument subtly imply it with talk of “dual use” technology, meaning it may be deployed both for good and bad purposes, as if the choice is up to us. The truth is that, not only do many technologies involve an inherent mix of social benefits and drawbacks (mobile phones being an obvious example) but they do not come value-free in the first place (ditto).

The matter is brought into rather sharp focus as we continue to buy cars and award governmental contracts to the man who recently retweeted, with an emoji indicating 100% agreement, the comment that “White solidarity is the only way to survive”. I don’t know that it offers much comfort, as Nasa considers making Elon Musk’s SpaceX Starship rocket even more central to its Artemis lunar program, to realise that this won’t be the first time that the US effort to send people to the moon has yoked itself to a white supremacist with a fondness for Nazi salutes.

The willingness of governments to knowingly embrace tech providers who have political and ideological goals does, however, make you wonder if they are getting any advice at all that has not swallowed the neutrality myth hook, line and sinker. The UK government seems now to be getting cold feet about the decision in 2023 to give the contract for an AI-enabled data platform for the NHS to the US surveillance technology company Palantir. Might, perhaps, the company’s involvement with Trump’s ICE raids and with the Israeli military and security agencies provoke distrust among doctors and the public that will hinder rollout of the technology?

Well, quite. But who could have guessed that the company co-founded by the dictatorship-curious far right libertarian Peter Thiel, currently to be found lecturing the pope about the return of the Antichrist, might be inherently problematic? If only there had been some warning signal, when Keir Starmer was introduced to current Palantir chief executive Alex Karp by (who else?) Peter Mandelson, that this man’s interests might not align with those of the British government. (Here’s a hint, free of charge and not entirely sardonic: avoid all tech that takes its name from Tolkien, such as crystal balls that will almost inevitably corrupt all who gaze into them.)

Karp made his pitch apparent in a recent interview, where he openly and even proudly stated that his AI technology “disrupts humanities-trained – largely Democratic – voters, and makes them economically powerless, and increases the economic power of vocationally trained, working-class, often male, working-class voters.” To be even more explicit: it’s the “highly educated, often female voters, who vote mostly Democrat” whose power will be lessened by it. And that, it seems, is something Karp would welcome. “Democrats completely neglect males”, he said, his voice dripping with contempt. “I don’t find it very appealing as a dude”.

In other words, what Palantir is doing will intervene in democracy in specific ways. They are saying it out loud. But that is nothing new; tech leaders such as venture capitalist Marc Andreessen (who recently announced, pace Socrates, that introspection is a modern malaise) are openly deriding democracy and boosting the voices of far right racist “thinkers” such as Curtis Yarvin.

There is of course nothing new in mega-rich business leaders and media moguls aligning with the political right. But as writer Jacob Silverman explains in his new book Gilded Rage, the game has changed, not just because of the scale of the wealth and the reach of the technologies into every aspect of our lives but because such values are being built into those technologies themselves, and also because of the increasing extremism of this tech-based ideological fervour.

The incursions of Cambridge Analytica into the Brexit referendum and the 2016 US election were only the first hint of what is now unfolding: a potential future in which democratic agency becomes a facade and a self-appointed, astronomically rich tech elite run things just however they choose. That future is not inevitable, but it might be avoided only if governments stop supposing tech developers are simply making convenient, value-neutral data-handling services or “innovation-accelerating” machines.

Share

By Lucy Reade

Even the most chronically online among us probably didn’t know his name until his death was announced, but the company he owned has unarguably changed pornography, the internet, and possibly even society.

The owner of OnlyFans, Leonid Radvinsky, passed away on March 20 aged 43 after a battle with cancer. His estimated net worth was $4.7 billion, having bought the company from its British founders in 2018.

What began as a platform geared towards subscription content quickly became the dominant force in online pornography and, by 2025, the world’s most revenue-efficient company.

I remember when OnlyFans first entered the mainstream, and people asked the obvious question: why would anyone pay for porn when there are literally millions of free videos online at the click of a button?

The answer is that you’re not really paying for porn. OnlyFans has created a new class of “adult” influencers who sell not just sex, but themselves.

Many have praised the platform as a kind of liberation: the democratisation of porn. Whereas before, we tended to think of porn production as based around studios with a handful of big names, but generally distant and anonymous performers. Now, the performers are in control and reap the rewards directly.

And, granted, that does beat the old model. Instead of sterile sets with an army of cameramen and a clipboard-wielding sleazy director, women can, hypothetically, film a quick striptease in between work-from-home teams calls and make a passive income.

One of the site’s top creators, Sophie Rain, said this in her tribute to Radvinsky: “That man built something that changed my entire life. Before OnlyFans I was waitressing and barely making rent. That platform gave me everything.”

But is this the reality?

Yes, some creators like Lily Allen and Megan Barton Hanson make tens of thousands a month. But they already had audiences. For most, the story looks very different…

It is not uncommon if you are a young woman to receive targeted posts from other young women bragging about how making an OnlyFans account helped pay off their student loan, let them quit their job, or buy a big mansion.

It is both predatory and misleading, because the average creator on the site makes less than $200 a month. Is that really a trade-off worth making – sacrificing the possibility of an ordinary career for what, in most cases, amounts to very little return? It’s a glamourised version of sex work that massively underplays the very real risks.

That is to say, this creation of sex influencers comes with a darker side: to break through on OnlyFans, if you don’t already have a following, you have to be willing to do what no one else has done – to be more shocking, more risky.

The conflation of porn and influencer culture, which is in itself already a battle for attention, has essentially turned into a sexual arms race, where people are pushed to do increasingly extreme things just to stay relevant – you only have to look at Bonnie Blue to see exactly how that plays out.

She rose to prominence on OnlyFans, and is best known for having sex with over 1,000 men in one day, which was filmed and posted to the site before being taken down as the ages of the men who took part could not be verified. She is currently claiming to be pregnant (after a stunt where 400 men took part in a “breeding mission” orgy), and making more content off the back of this.

Even that wasn’t without competition. Her former friend, Lily Phillips, had previously made headlines by sleeping with 100 men in a day – and reportedly felt undercut when Blue went further.

And even though the stomach churning category of “barely legal” has been a porn subgenre for longer than some of these creators have been alive, there is a newer, more sinister element when you realise that the Onlyfans-to-influencer pipeline goes both ways.

Underage influencers, such as Lil Tay, who went viral for her rapping and satirical content at just 10 years old, have launched Onlyfans accounts ready to be opened the literal second they turn legal. Tay’s page received over $1 million in subscriptions within three hours of her 18th birthday.

OnlyFans is a product of our modern day internet, where the battle to top the algorithm produces ever-escalating spectacles. Big Youtube creators like Mr Beast, who essentially just makes real life Squid Game videos, have proven that attention is currency, and the price of earning it keeps rising. And porn is no exception.

The trajectory is only heading towards greater extremity. In a climate where there is a financial incentive to push the limits of what people are prepared to do on screen, the obvious question is where that ends – and whether it ends before something genuinely irreversible and tragic happens.

A final reflection: porn has historically been hidden, stigmatised, separated from mainstream culture. Now porn stars make headlines, and weave seamlessly into mainstream culture. At the same time, it has reduced the incentive to pursue real world relationships.

What it has essentially given rise to is parasocial pornography. We are already inundated with headlines about porn addiction, particularly among younger generations, and its links to things like erectile dysfunction. Repeated exposure to stimulating content raises the threshold for what feels rewarding. The danger is that real-life intimacy starts to feel comparatively dull.

Young people – mostly men and boys – are already so affected by this that many aren’t even attempting to seek out real relationships. Now add Onlyfans-specific features like messaging, personalised content, and the so-called “girlfriend experience” – essentially paying someone to pretend they fancy you while sending nude pictures – and it becomes a complete substitute for real intimacy. No wonder Gen Z is the least sexually active cohort in modern recorded history.

I often think Gen Z has this uniquely strange relationship with sex – our culture is at once extremely sexualised and also puritanically sexless – and OnlyFans feels like the clearest expression of that contradiction.

But in the end there may be no point in making such a big fuss about it, as AI generated content will probably replace the OnlyFans creator within a few years anyway. As video killed the radio star, so bots will kill the porn star.

Share

There are few things better than seeing a much-hated villain get their comeuppance – and in 2026, there are few bigger villains in the public imagination than the tech giants.

The outpouring of glee and schadenfreude towards Meta, after the double-whammy of US legal verdicts which has seen the company found liable for designing addictive products, misleading consumers about the safety of its platforms, and enabling harm to its users, has been pretty much universal. But it’s worth bearing in mind that it’s also a touch premature.

It’s important to state at the outset that, in many key respects, the company is bang to rights. No one who has followed its activities closely over the past 15 years can be in any doubt that this is a business whose primary goal has been to attract as many users by whatever means necessary, and keep them there long enough to monetise them as hard as possible.

It feels like these judgements have been a long time coming. Whistleblowers have been voicing concerns for years. When Tim Kendall, a former director of monetisation at Facebook, gave evidence to the House Committee of Energy and Commerce way back in 2020, he told them: “We didn’t simply create something useful and fun. We took a page from Big Tobacco’s playbook, working to make our offering addictive at the outset…[we] made status and reputation primary and laid the groundwork for a teenage mental health crisis.”

Meta’s history is littered with barely-there figleaf steps, purportedly designed to combat potential harm against users and issues of addiction. Who can forget Instagram’s “All Caught Up!” feature, introduced in 2018 and designed to let users know they had reached the end of the content funnel.

That was retired a few years later when the need to compete with TikTok took over from cosmetic nods to limiting screentime. Its “take a break!” notifications, were introduced in 2021, and the “nighttime nudges”, launched in 2024. These gently reminded teens that it was late and they perhaps ought to consider not watching another 300 reels. Astonishing, really, that these concrete actions weren’t enough to convince a jury of the company’s sincere approach to addiction.

Now, with social media bans being considered, and in some cases enacted worldwide, and courts finally seeming to crack down on malfeasance by the platforms, it’s easy to see why these rulings are being treated as a Rubicon. While they might yet prove to be a game-changing moment, any resolution could take a long time to present itself.

The crucial detail buried in the triumphant reporting of the verdicts is that Meta (and, in the case of the LA trial, codefendant Google) can, and will, appeal. While the initial judgements against the tech companies give hope that the legal case will hold, the possibility of this setting precedent means that Meta in particular will be hugely motivated not only to win but to snarl the case up for as long as possible.

Should that appeal prove unsuccessful, the case could potentially end up in the Supreme Court, further extending the timelines. That means any final resolution – which would determine precedent, and the likelihood of sanctions and platform change – could take until 2029, not quite the immediate smackdown that it seems.

It’s also worth noting that the companies in question are richer than any other businesses in history, and, as such, can and will hire the very best legal counsel that money can buy in order to make the case that they have been treated unfairly.

This vast wealth also means that imposing meaningful fines is not easy. The fine imposed on Meta in the New Mexico judgement is a hefty-sounding $375m – an impressive sum, until you do the maths and realise that it accounts for exactly 0.62% of the company’s net profits in 2025.

Don’t get me wrong, it’s good news that tech companies are finally starting to be held to account after nearly two decades of moving fast and breaking the fabric of society. Equally, though, it’s important to be clear-eyed about the extent to which there is still a long way to go before they are compelled to make meaningful changes to the ways in which their platforms work.

Oh, and meanwhile there’s a whole new suite of addictive tech we don’t understand being pushed on us by many of the same companies in the shape of AI; perhaps it’s a little premature to declare victory over big tech just yet.

Share