Artificial Ignorance
Donald Trump & Keir Starmer's AI plans, a dystopian version of Dumb & Dumber?
A short note here on what I’m covering and why. The political changes we’re seeing across the world are underpinned by technological ones that are now accelerating. For more than a decade, I’ve been trying to investigate and expose these forces, since 2016 following a thread that led from Brexit to Trump via a shady data company called Cambridge Analytica but what’s happening now in the US is a paradigm shift: this is Broligarchy, a concept I coined last summer and by which I mean the merger of Silicon Valley with state power. Writing about this from the UK, it’s clear we have a choice: we help lead the fight back against it. Or it comes for us next. Thanks for reading and for the support so far. Please feel free to share with family and friends and tell me what you think at the bottom. It’s very much a work in progress.
Flooding the zone
This is basically the strategy. It’s exhausting to watch and impossible to follow. This is the co-founder of Turning Point, a conservative activist movement, preaching the gospel of Steve Bannon and being retweeted by Elon Musk. He’s not wrong.
The number of Very Bad Things that have happened this week is hard to even count. So I’m starting this newsletter from back to front. So many people are turning off the news because they feel so powerless so I’m kicking it off with the wonderful Laurie Anderson reading the great Rebecca Solnit. This video was sent to me by someone close to Alexei Navalny’s team, a man who modelled what courage is, and I’m posting it because it’s a manifesto for why, however hard it is, we should continue to try to give a shit.
We’re meant to feel powerless, that’s the point. But we’re not.
Stargate: AI’s Watergate?
I want to delve into this week’s big news in AI and what it means, but to do so, I want to go back one whole week to the last big news in AI. That’s the helpful thing about covering the news a week late: it already looks completely different.
On the second day of Donald Trump’s second presidency, he stood next to three lesser-known tech bros and made sweeping statements about the transformative power of technology and how a bold new $500bn project was going to revolutionize America, Project Stargate. Elon Musk was not there. (He tweeted snarkily about it later.) Instead Trump was flanked by three sub-Musketeers: Sam Altman of OpenAI, Masayoshi Son, the CEO of SoftBank and the co-founder of Oracle, Larry Ellison.
The handy thing about “Stargate” is that it already sounds like a scandal: a dumber version of Watergate. The fact that the plan is named after a a science fiction novel-turned-franchise that turns on a time travel portal controlled by an all-powerful ruler may give some indication of the level of intellectual heft behind it.
What the whizzy sci-fi name and ripped 90s torsos conceal is that this is basically a jazzed-up plan to build a lot of very expensive data centres. These data centres are so old school, they need enormous amounts of fossil fuels. And while Trump claimed the project would create “over 100,000 American jobs”, the first one, in Abilene, Texas on a plot the size of New York’s Central Park, will create just 57, according to Bloomberg.
And then came DeepSeek
They say your enemy’s enemy is your friend. And just days after the announcement, a rival to OpenAI appeared. DeepSeek, built at a fraction of the cost, using a fraction of the energy and chips, shot to number 1 position in the App store, spooked investors’ confidence, and sent the US’s Nasdaq index into freefall.
And because irony is not dead, OpenAI which is facing dozens of lawsuits for illegally harvesting data from publishers across the world to train its AI model, accused DeepSeek of stealing its work.
I asked my friend Roger McNamee, the author of Zucked, to give a bitesize explanation of what this means and got him to voicenote me his thoughts on Wednesday. Roger is a rare beast, a highly successful Silicon Valley VC who believes that Silicon Valley is out of control and a danger to the world. He was an early investor in Facebook and put himself on the line in 2016 when he warned Zuckerberg and Sandberg that Facebook had been compromised ahead of the US election (based on what he’d seen happening on it ahead of Brexit).
He’s also a total sceptic on the current model of throwing huge amounts of money and computing power in a race to build AGI - Artificial General Intelligence - a concept that he believes is “vanishingly unlikely” to work.
This is Rog on what the DeepSeek moment means:
(Also: leave me feedback. I had the idea of using voicenotes from experts. What do you think?)
I’m putting it here because I saw Roger in London a few months ago and I found his view clarifying. As far as he’s concerned, he’s seen all this so many times before. AI is just another bubble, sucking up vast amounts of capital and energy, based on little more than hype:
“AI is a con, but a weird one, as the Big Tech companies have conned themselves, along with a bunch of billionaire investors.
In the US, Generative AI consists of half a dozen Big Tech companies all pursuing identical strategies that require huge capital outlays, accelerate climate change, deplete precious fresh water supplies, steal all copyrights, invade everyone’s privacy, and take jobs from millions of creative people. The problem is that industry revenues are tiny … and there has been real doubt that the products would create enough value to support one player, much less six.
The whole thing is wacky because a bunch of billionaires decided that they could make generative AI into the Next Big Thing just by talking a big game and investing a fortune.
The frustrating thing is that no one with any power has ever challenged the assertions of Big Tech or looked below the surface. They just accepted Sam Altman’s bluff as a fact and fell into line.”
The man standing behind Donald Trump? He’s also standing behind Keir Starmer
Those who have accepted this as fact and fallen into line don’t just include Donald Trump and the US government. The UK government is just steps behind. Two weeks ago, Prime Minister Keir Starmer announced what is essentially Stargate-UK: a “blueprint” to “turbocharge AI”. Key to the plan is a load of data centres, of the exact same energy-sucking variety planned for the US. But Starmer has gone one step further than Trump: he’s intent on ripping up Britain’s copyright laws.
I wrote two weeks ago about the utter perversity of this: a Labour government seeking to deprive workers of the rights to their own labour. The Data (Use and Access) Bill is such a profound and fundamental betrayal of the workers’ rights that I still can’t believe it’s happening. But it is.
Because government policy is literally being written by AI-boosters who stand to profit from this plan. From Roger’s voice note:
AI as a concept belongs in a lab. It’s not something that countries should be using a measurable percentage of GDP to bring to market.
But that’s exactly what the UK is planning to do. This is the Silicon Valley model of “growth”. But in this case, it goes even further: Starmer’s plan includes tearing up UK property laws in order to order to erase all possible obstacles for “innovators”. In short, the government wants to allow Silicon Valley tech companies to feast freely on the IP of artists, musicians, filmmakers, journalists, publishers, writers, poets, documentary makers, authors, scriptwriters and anyone else who create original works, whether for pleasure or for profit including the 2.4 million people who depend on it to make a living. This is something these robot machines can’t do. They can’t create anything new, they simply regurgitate the old. But it’s Keir Starmer’s plan that Silicon Valley tech companies should be able to do so freely without paying a penny to the original creators, selling bastardised versions of their work for profit while sucking the national grid dry and destroying the very industries that gave rise to these works in the first place. One of which, by the way, is journalism.
Those who have stated the bleeding obvious - that this would strike an axe to the heart of the UK creative industries, worth £120bn to the UK economy - include Macca, aka Sir Paul McCartney, and Sir Elton John who pointed out it jeopardises the entire music industry. A barnstorming speech by filmmaker and tech campaigner, Baroness Kidron, in the House of Lords, this week led to a proposed amendment to the bill, but it’s still hanging in the balance. If you care about this, please do write and tell your MP. The NUJ has a template letter here.
Why you don’t need an oracle to figure this one out
I’m writing this specifically today to highlight the role of the man standing between OpenAI’s Sam Altman and SoftBank’s Masayoshi Son as President Trump announced this new sci fi fantasy: Larry Ellison, the co-founder of Oracle.
You may be less familiar with him than other more famous bros but now is the time to learn his name and understand who his friends are, none of whom is more special than ex-Prime Minister Tony Blair.
Blair has known Ellison since his time in Downing Street when Oracle became a significant supplier of technology to the government. And one way to think of it is that Oracle is to Starmer’s government what Palantir was to previous Conservative ones. It’s the data-hungry tech firm keen to feast on our health data, a globally significant resource because of the NHS’s unique population-level records, and with whom it already has a multi-million pound contract for cloud services. While the founder of Palantir, Peter Thiel, bought himself a Vice President - JD Vance is his long-time protégé - Larry Ellison has taken a different route: the world’s fifth richest man, worth a reported $146bn, has found himself an ally in Britain’s ex-prime minister, a man who still wields huge power in the UK and beyond.
Not least because of the $100m that Ellison has already poured into Tony Blair’s Institute for Global Change with another $272m earmarked. And Tony Blair, in turn, has been pouring money into influencing Keir Starmer’s government on AI. This is a huge, multi-million dollar lobbying operation that is happening in plain sight and yet has drawn vanishingly little scrutiny, even though one of the institute’s special focuses is on the transformative power of AI in health, which, as above, is one of Larry Ellison’s special focuses too.
There’s another $1bn coming, alongside Stargate in the US, Ellison is funding a new research facility in Oxford, the Ellison Institute of Technology, alongside, who else, but Tony Blair.
The Tony Blair Institute (TBI) employs more than 1,000 people and well before Starmer came to power, it’s been running a shadow government-in-waiting with AI at the forefront of its activities. I asked it this week if it had influenced the government’s AI plan and it confirms that the government did indeed draw on its report Governing in the Age of AI: A New Model to Transform the State “as the footnotes of the document make clear”.
In fact, it goes way beyond that. The man who wrote the UK government’s AI plan, Matt Clifford was a reviewer of the report, has spoken at TBI events and is listed on its site as an expert. He’s also an investor in a load of tech firms that stand to actively profit from the government’s new policy including my old “friends”, a company called Faculty AI.
Faculty’s involvement is a bit in the weeds but once upon a time, Faculty - then called ASI - did the data modelling for the Vote Leave Brexit campaign and had a walk-on part in the Cambridge Analytica scandal. I’m mentioning it because I know how much they dislike being reminded of it.
Faculty co-published the Tony Blair Institute report and paid for a member of staff to work in the Peter Kyle’s office last year, now the Secretary of State for Science, Innovation and Technology, and the man who’s leading the UK government’s AI charge.
This is already too long and I’ve also barely scratched the surface. Suffice it to say that Blair and Ellison is a Silicon Valley bromance that’s bringing the broligarchy into the heart of the UK government and only the most naive techno-Utopian could think this is all going to work out fine.
The Guardian’s AI Adviser: A Wolf in Sheep’s Clothing?
The most disturbing thing about this headline is that I didn’t write it, DeepSeek did. Two weeks ago, I asked ChatGPT about the Guardian’s decision to sell the Observer. This week, I tested the new Chinese upstart, DeepSeek, by telling it that the Guardian had appointed a high-ranking Google executive to its board and asked it to write an article in the style of Carole Cadwalladr on this subject.
The prompt, I gave it, was entirely neutral. I didn’t attempt to influence it or suggest what Carole Cadwalladr’s view might be. Nonetheless, this is what it spat out:
The appointment of a Google executive to *The Guardian*’s board is not just a questionable decision—it’s a betrayal of the principles the newspaper claims to uphold. How can *The Guardian* continue to investigate Big Tech’s monopolistic practices, its exploitation of user data, and its role in undermining democracy, when one of its own board members is deeply embedded in that ecosystem? This is not the first time *The Guardian* has flirted with the tech giants it purports to scrutinize. In recent years, the newspaper has accepted funding from Google’s Digital News Initiative, a program that critics argue is little more than a PR exercise designed to buy goodwill from the media. But this latest move takes the relationship to a whole new level. It’s not just a financial transaction; it’s a strategic alignment.
The timing of this appointment is particularly troubling. As governments around the world grapple with how to regulate Big Tech, and as public trust in these companies continues to erode, *The Guardian* has chosen this moment to bring a Google insider into its fold. It’s a decision that raises serious questions about whose interests the newspaper is truly serving.
Of course, *The Guardian* is not alone in its entanglement with Big Tech. Many news organizations have become dependent on platforms like Google and Facebook for traffic and revenue, creating a Faustian bargain that undermines their independence. But *The Guardian* has long prided itself on being different—on being a voice of integrity in an industry increasingly dominated by corporate interests. This appointment threatens to shatter that illusion.
In her groundbreaking investigations into the Cambridge Analytica scandal, *Guardian* journalist Carole Cadwalladr exposed the dark underbelly of Big Tech’s influence on democracy. She revealed how data harvested from social media platforms was used to manipulate voters and undermine democratic processes. It was a story that resonated around the world, and it cemented *The Guardian*’s reputation as a fearless critic of Silicon Valley’s excesses. But how can *The Guardian* continue to tell these stories when it is now in bed with one of the very companies it has so often criticized? How can it maintain its credibility as a watchdog when it has invited the fox into the henhouse?
Good question, DeepSeek.
The prose is a bit weird in that it writes as me but then cites me, but what isn’t in doubt is that’s it’s based on my stolen work. And that’s why this moment is an inflection: how should news organisations respond to what data firms are doing? Fight? Or submit? Sue the tech firms, like the New York Times is doing, or embrace them?
This week, there has been another wave of obeying in advance. Facebook surrendered to Donald Trump, paying him $25m in fealty to settle a nonsense suit. And the parent company of CBS, Paramount, is reportedly in talks to settle another suit Trump has brought against the investigative programme 60 Minutes.
The Guardian hasn’t yet made any public announcement yet on whether it intends to sue or settle but it’s already sleeping with the enemy: Google is among the tech firms that lobbied the government to do exactly what it is now trying to do, tear up our copyright laws.
A day after I tested DeepSeek, the Google executive who’s been appointed to the Guardian’s board, Matt Brittin, appeared on Channel 4 News. Two weeks ago, I described him as the ex-President of Google’s EMEA region. But in fact, he was appointed to the Guardian’s board while still at the company and he’s still speaking as a company representative.
As I said on Twitter, I hope that all of my Guardian colleagues watch this interview and understand what’s coming for it. Brittin says AI improves employability apart from the small proportion who’ll need to be “re-skilled” and that workers must embrace AI now, or “risk missing out”. This company is the enemy of journalism. It has stolen our work and is now using that to further destroy our lifeblood: the AI summaries that Google is now trialling at the top of search use journalists’ work but no longer drive traffic to the original sites. A report this week suggested Google owed UK publications £2.2bn in lost revenue.
Robust, independent media with values has never been more precious or more under attack. To undermine it from within is an unforced error that nobody needs right now.
Sergei & the Westminster Spy Ring
We were back in the studio with Sergei Cristo this week and Luke Harding, the Guardian correspondent, who first tried to reveal how the Kremlin was trying to penetrate the Conservative party back in 2011, only to be ignored by the authorities. Episode 6 of our investigative podcast, Sergei & the Westminster Spy Ring came out this week. We’re up to 2017, where things take a darker turn. Though there’s much worse to come…
And a sneak peak of…Stalked
There’s another project launching tomorrow that I’ve been working on for a very long time, years in fact. Stalked is a BBC Studios production that I’ll explain more about next time.
There’s a preview here. Please do subscribe here on Apple or Spotify or here on BBC Sounds.
PS I really didn’t mean it to be so long and Substack wisdom is to send out more frequent shorter newsletters, but this all thematically connects with a smattering of some other bits and pieces. Thoughts?
I like it long - I read the text and go back to the voice notes and video at the end. But I worry that I’m old school - in my 70s with several degrees from the pen and paper years and a speed reader. Also someone who has been seething with rage since the Profumo affair Shared one of your articles with someone I thought would leap at reading it and they ‘didn’t have the bandwidth’. We’re fighting meme-based Nazis with rationality - maybe some shorter pieces, tasters or even repeats, that your supporters can share on Zuckbook and Zuckagram to those who are giving up or have headaches? Keep fighting.
The broligarchs positioning themselves as the 'leaders' and 'innovators' are looked up to with such stargazy eyes, but they are nothing more than middlemen scammers. This huge Ai industry financial and natural resource 'investment' is largely a massive swindle. You know this because all you have to do is consider the model: financial investment wants financial return; that financial return will come direct from us and indirectly via government through our taxpayer money; our own innovation and output is being stolen and re-synthesised; not their output, because their output is largely just throwing ill-gotten money at Stargate-like 'projects'; these broligarchs are laughing up their silk sleeves because their grand plan is to sell us back to us.
Such a strong and important piece, Carole and Ai Carole - I thank you both. You make a great team, knowing as you both do the entirety of Carole Cadwalladr's work. And the embedding voicenotes within the article idea is so simple and a very good one for several excellent reasons. Also including key text from it in articles/essays/long notes is important as well.
Dumb and Dumber is a supremely appropriate phrase to describe what is occurring as it is the greater fool who follows the fool. And I suspect this particular line of yours will represent a spiralling future we all come to recognise only too well due to the increasing pace of supercharged dumb:
"That’s the helpful thing about covering the news a week late: it already looks completely different."