Thinking Without Machines English

E. P. Thompson on the ZX Spectrum

Sun, 13 Sep 2026 00:00:00 +0000

In 1984, one of the most famous historians of the twentieth century appeared in a computer game.

The game was called Deus Ex Machina and the historian was the Marxist peace campaigner E. P. Thompson, author of The Making of the English Working Class (1963).

Loading screen for Deus Ex Machina (1984) on the ZX Spectrum.

Like many historians, I first read Thompson as an undergraduate. And I still use his “Time, Work-Discipline, and Industrial Capitalism” article in my teaching. Now, I’m by no means a world expert on Thompson. But I have read all his major works, including the somewhat obscure Alien Homage, Thompson’s biographical account of his missionary father in colonial India.

I was not, however, aware that Thompson featured in a 1980s radical art game made for the ZX Spectrum. Not until I came across a copy of Deus Ex Machina whilst browsing ZX Spectrum cassettes on eBay. The game looked cool, so I looked it up. And there, in the credits, I was intrigued to see “E. P. Thompson” listed.

At last, I thought, all my interests had finally converged—radical history, old technology, and computer games.

Deus Ex Machina

As it turns out, Deus Ex Machina has a bit of a cult following amongst aficionados of 1980s gaming. It was the fever dream of architect-turned-game-designer Mel Croucher. It is hard to convey exactly how weird the game is, and how original it was for the time, and frankly still is. If you have a spare 48 minutes, I highly recommend watching a play-through on YouTube. Or, better yet, you can actually buy a copy and play it on a modern computer via Steam.

The game follows the life of a single human—from conception to death—in a dystopian future (that is, 1994). In this future, the United Kingdom is governed by the “Machine” which manages “defence and internal security” as well as “all passport, communications and censorship operations”. In the game, you—the player—are an anomaly. You weren’t meant to be born. You are in fact the result of a freak accident. At the beginning of the game, the “last mouse on Earth” crawls inside the “Machine” and does a shit. I kid you not. Somehow, this mouse shit causes a DNA mutation and the fertilisation of an embryo. You then play through various “ages of man”, including a sequence in which you tumble across a strobe-lit screen as a foetus, all whilst trying to avoid the “Defect Police”.

Sequence from Deus Ex Machina (1984) in which the player controls their own foetus.

It’s a work of art for sure. But the really clever thing is that the game was accompanied by a separate audio cassette. At the time, audio in computer games was very rudimentary—the ZX Spectrum had a single channel beeper. You couldn’t record vocals or a guitar. You could just beep across a few octaves.

Croucher, however, had the brilliant idea to separately record the audio and ask the player to put the cassette on at the same time. There’s even a countdown at the start to get everything in sync. And again half way through to switch over to the other side of the tape.

For the vocals, Croucher managed to recruit a few famous names. The comedian Frankie Howerd (Carry on Doctor) voiced “The Defect Police”, John Pertwee (Doctor Who) voiced “The Storyteller”, and none other than English punk legend Ian Dury (Blockheads) voiced “The Fertiliser”.

And E. P. Thompson. He was “The Voice of Reason”.

Credit for “Edward Thompson” in Deus Ex Machina (1984).

Games for Peace

At its core, Deus Ex Machina is a critique of technocracy, autocracy, and violence. Mel Croucher himself was involved in anti-nuclear campaigns. He made games in which the player has to avoid violence, or is the victim of violence, rather than the perpetrator. “If that all sounds like I’m an airy fairy pacifist, stuck up my own arse, that’s because I’m an airy-fairy pacifist, stuck up my own arse,” declared Croucher in a developer commentary produced for the game’s 30th anniversary.

It was this connection to the peace movement, and the Campaign for Nuclear Disarmament in particular, that brought E. P. Thompson into the world of radical computer games. As Croucher later recalled:

He [E. P. Thompson] had been invited to address our local campaign for nuclear disarmament, and my role was to record his speech because I had some professional recording gear. When I explained what I was trying to do with Deus Ex Machina he simply tossed his grey lion’s mane and said, “where do I sign”.

The speech that Croucher recorded that day then ended up in the game: “[It] cost me nothing at all, apart from the hour and a half I spent discussing propaganda and freedom of speech with the greatest peace campaigner of the age, E. P Thompson.” Croucher later spoke about the encounter, fondly recalling:

He was an optimist. I thought at the time we were about to tip over the brink and start lobbing Trident missiles at one another. E. P. Thompson would have none of it. He was confident that walls would be breached and barbed wire would be cut.

Thompson’s words, cut between the hypnotic singing of the “Machine” and a dirty bass line, ring out as the player enters the fifth age of man—an overweight figure stomps down the street, and the player must decide whether to crush “Hate” and “Lies” or “Life”, “Love”, and “Peace” as barbed wire slowly moves up and down the screen.

Sequence from Deus Ex Machina (1984) featuring E. P. Thompson’s speech.

Appropriately, the speech itself addresses the very themes of Deus Ex Machina. Thompson talks about telephone tapping, the security industry, as well as the emancipatory power of music. I recently made a transcript:

I think it’s the very sense that the ground is moving under their feet—and the ground is us—that has made the political leaders, who feel unstable and uncertain in new ways, so determined to impose their strategies upon an increasingly reluctant people.

The security presence extends in different ways on both sides. Police, telephone tappers, the growth industries—you know, the only real growth industries now are things like telephone tapping.

I’m so stupid. I hadn’t been reading the NME. I recently started reading the New Musical Express and I find out that bands are going backwards and forwards all the time. Anyway, they are going forwards—our British bands are going to the GDR and Poland and so on. A lot of these are peace bands. And you begin to find that there’s already a huge movement of “Rock Across the Blocs” if you like.

The healing process is going on. Despite what the politicians, despite what the militarists think, the healing process is going on. And it’s something which we can contribute to by our solidarity, by extending this knowledge… Thank you.

Audio of E. P. Thompson in Deus Ex Machina (1984). Compressed to 180 KB.

For me, Deus Ex Machina is a reminder of a different kind of computer game, one with a different purpose. In the 1980s, computer games could be political. They could be punk. They could be art.

Today, that hasn’t been completely lost. Games like Citizen Sleeper and Disco Elysium are great examples of computer games as political art. And both address themes of technology and violence.

At a time when the computer games industry is in financial turmoil, when a game is just another investment opportunity, an asset class, we need more games like Deus Ex Machina, and more historians willing to engage with them.

For archiving purposes, I have deposited a copy of the transcript of E. P. Thompson’s speech, the original audio, and isolated voice audio here.

Low-Tech

Sun, 06 Sep 2026 00:00:00 +0000

Thinking Without Machines has gone low tech.

In basic terms, that means I no longer use Substack. Instead, I host all the articles on the static personal website you are viewing right now.

The goal is to reduce energy consumption and hardware requirements, both on the user and server side. At the same time, a low-tech personal website like this resists the increasing “platformisation” of the Internet.

If you want a detailed account of what the “low tech” Internet might look like, why it might be a good thing, and how to go about implementing it, I highly recommend reading How To Build a Low-Tech Internet (2023) by Kris De Decker, the founder of low-tech magazine.

As Decker argues:

We were told that the Internet would dematerialize society and decrease energy use. Contrary to this projection, it has become a massive infrastructure and a rapidly growing energy consumer.

It’s worth bearing in mind that Decker set up low-tech magazine in 2007, well before the rise of generative artificial intelligence and the even greater energy and infrastructure demands of the Internet today.

Beyond the Internet, Decker and his team makes a brilliant case for thinking about “the potential of past and often forgotten technologies and how they can inform sustainable energy practices” more generally. That doesn’t mean going back to the stone age. Rather, it means being selective about what technologies we adopt and how we use them. In fact, rather than “low tech”, it might better to think of something between high and low—what the British-German economist E. F. Schumacher called “intermediate technology” is his classic Small Is Beautiful: A Study of Economics As If People Mattered (1973).

Indeed, Decker fits within a long tradition, stretching from the arts and crafts movement of the late nineteenth century through to the environmentalism of the 1970s. For much of the past 150 years, people have been finding ways to quietly resist the “high technology” of industrial capitalism.

A Low-Tech Website

So how to respond? On an individual basis, Decker shows us what a low-tech Internet might actually look like. And it’s pretty cool.

The basics are all straightforward, but powerful.

First off, reduce energy use on the server side by making the website static. No calls to external APIs. No server-side processing of PHP or SQL or Ruby or alike. Just the original vision of the World Wide Web: users requesting static documents from a server. In practice, that means all the pages are static HTML, CSS and a little Javascript.

That doesn’t mean the website itself never changes. Quite the opposite. It’s just that the site only changes when there is new content, not every time a user visits it. To achieve this, low-tech magazine uses one of the many great open source static site generators—in this case Hugo—to turn local plain text files into a website. Then Decker and his team simply upload the new files when they write something new.

Second, reduce the actual file size of individual webpages. That’s already a lot easier when you’re just working with plain text: HTML and CSS take up very little space. Images are a bigger problem, quite literally. But rather than abandoning visuals entirely, low-tech magazine solves this by using “dithering” to massively reduce the file size of a typical JPEG or PNG. In addition, where possible, they use vector images (which are defined mathematically and thus very small in file size) rather than raster images (which are made up of millions of dots and thus large in file size).

Finally, to keep things as tight as possible, low-tech magazine makes minimal use of Javascript, and imports no external Javascript libraries. Javascript itself isn’t really a problem, but when it is used excessively, it adds to the processing overhead for the user. Plus the typical libraries that most websites rely on easily add up to a couple of Megabytes of extra data to download.

The Results

Having set this all up, I decided to do a little test. I used the “Inspector” function on Firefox to see exactly how much data I was downloading and how many calls to external APIs I was making, comparing Substack with my new low-tech website.

To give a rough idea: loading my article ‘Learning to Play: Computer Games and the Origins of AI’ on Substack would make 217 user requests to the server (or other external servers) and download 4.61 MB of data, which was then decompressed by the user to 10.43MB.

Loading the same article on this website makes just 10 server requests and downloads 335kB of data. There is no additional decompression step.

And that’s just the user-side stuff. Substack of course does all kind of complex server-side processing for the content, accounts, advertising, analytics and alike, which I have no visibility on. My website does no extra processing. It just sends the static pages.

In any case, I’m pretty pleased with the result. Now, low-tech magazine actually goes a step further than this. For now, I still rely on my domain-name provider to host the web server. In contrast, low-tech magazine uses their own locally-hosted server, running on a very low-powered computer. Not only that—and this is seriously cool—low-tech magazine runs their server using solar power. Being based in Barcelona, rather than the Midlands, does help somewhat with that. But hey, the weather has been pretty ~~good~~ extreme recently, so you never know.

When Technology Became Software

Fri, 26 Jun 2026 00:00:00 +0000

“Big Tech” is big money. Nine out of the ten most valuable companies in the world, as measured by market capitalisation, are technology companies. The top five—Nvidia, Alphabet (Google), Apple, Microsoft, and Amazon—are all routinely described as technology companies.

Rank	Company	Market Cap
1	NVIDIA	£3.586 T
2	Alphabet (Google)	£3.158 T
3	Apple	£3.057 T
4	Microsoft	£1.982 T
5	Amazon	£1.847 T
6	TSMC	£1.706 T
7	SpaceX	£1.524 T
8	Broadcom	£1.363 T
9	Saudi Aramco	£1.273 T
10	Samsung	£1.097 T

Top ten companies by market capitalisation on 26 June 2026. Source: companiesmarketcap.com

From a history of technology perspective, there’s something a bit odd about all this. The global economy is dominated by technology companies, yet most of these companies don’t actually make anything. They’re almost all software companies.

Out of the top five, Nvidia is the only technology company focused mainly on hardware. But even then, Nvidia needs someone else to actually manufacture its graphics cards, as they don’t design or build their own fabrication facilities. That is done by Taiwan Semiconductor Manufacturing Corporation (TSMC) and Samsung.

Apple, of course, also sells a lot of hardware—iPhones and MacBooks and alike. But again, Apple is even more reliant on actual hardware companies, such as Samsung and ARM, to provide the core architecture and manufacturing capability. In reality, Apple’s business is a software business. They develop what they market as fantastic software—iOS and MacOS—and then tie that software to hardware sold with a big markup. These days, Apple is also all in on the “software as a service” model, selling music and video streams alongside access to professional editing suites.

At first glance, Amazon is a bit different. Still, whatever it is, Amazon certainly isn’t a hardware company. (Sorry, the Kindle doesn’t count.) From a consumer perspective, Amazon is just a big online shop. But really, this is just another tech company built around software. The basic software is the website itself, and then on top of that Amazon sells various services, such as music and video streaming. In addition, Amazon’s main business these days is actually selling cloud computing, which again, is really just software managing access to data centres.

The remaining two in the top five are even more obviously software companies. Microsoft and Google might sell a few computers and phones, but their core business is software, whether an operating system or a search engine. Increasingly, they too sell this software as service, as with Microsoft 365.

Of course, all these companies buy hardware. Microsoft, Google, Amazon and Apple all spend enormous amounts on data centres. They certainly invest in infrastructure, just not manufacturing. But that doesn’t make them technology companies any more than a supermarket or a farm is a technology company because it buys a till or a tractor.

Hardware vs. Software

Why then do we refer to software companies as technology companies? And when did this start to happen? To answer these questions, we need to look back to the history of computing and the origin of the hardware vs. software distinction.

Famously, prior to the widespread use of electronic computers, this was a division of labour between real people. The division was typically gendered. Mathematicians, often men, decided what calculations needed to be done. Computers, often women, performed those calculations to get the results.

Human computers at Harvard University performing calculations for the astronomical observatory in 1892. At this time, women could not study for a degree at Harvard. Source: Wikimedia.

In the early nineteenth century, as the industrial revolution got underway, people started to wonder if this division of labour could be reproduced in a machine. Computers at this time were therefore mechanical. In the 1820s, the English mathematician Charles Babbage designed what he called a “Difference Engine” to calculate polynomial functions. To input the function, you had to set the mechanical wheels.

This was really just a fancy calculator, but Babbage followed it up with a design for what became known as the “Analytical Engine”. This was a general purpose computer—it could perform different operations depending on instructions input by the user—but it was still mechanical. Tellingly, Babbage used the language of the industrial economy to describe the storage and processing of data. There was a ‘store’ (where the data was held) and a ‘mill’ (where the data was processed).

Part of prototype of Charles Babbage’s “Analytical Engine”, Science Museum, London. Source: Wikimedia.

Babbage only got to the prototype stage. Some similar machines were built in the nineteenth century, but it was really only in the twentieth century that general purpose computing became a reality. In 1941, the German engineer Konrad Zuse built an electromechanical computer, the Z3, which was quickly co-opted into the Nazi war effort. Around the same time, a group of British codebreakers at Bletchley Park built an electronic computer, known as Colossus. It was run using a combination of paper tape input and electronic switches. Following the Second World War, engineers built similar electronic computing machines, such as the Manchester Baby. Crucially, these new machines could actually store programs in memory—in the case of the Manchester Baby, exactly 1024 bits stored via charges in a cathode-ray tube.

1947 patent for “apparatus for storing trains of pulses”, a kind of cathode-ray tube adapted to serve as computer memory. Source: Wikimedia.

From the Analytical Engine in the nineteenth century to the Manchester Baby in the twentieth century, the computer as we know it was defined in terms of this division between the stored program and the processing of instructions. It was a division that mirrored a particular industrial world—between male managers and female workers—the ‘store’ and the ‘mill’ as Babbage put it.

In 1958, the American computer scientist John Wilder Tukey first used the terms “software” and “hardware” to describe this division. As he put it:

Today the ‘software’ comprising the carefully planned interpretive routines, compilers, and other aspects of automative programming are at least as important to the modern electronic calculator as its ‘hardware’ of tubes, transistors, wires, tapes and the like.

Over the next forty years, software was to completely overtake the importance of hardware in the business of computing.

Deindustrialisation and Big Tech

The modern computer was born out of a world of industrialisation. The rise of technology as software, on the other hand, was a product of deindustrialisation.

At exactly the same time that countries in Western Europe and North America began to move away from manufacturing, the software industry emerged to serve a new market. The economies of countries like Britain and the United States, rather than selling manufactured goods, began selling services. This new service economy—accountancy, consultancy, legal arbitration, insurance, index funds, and alike—demanded software. The most successful “technology” companies of the 1980s and 1990s were those that sold, not the hardware, but the operating systems and databases—companies like Microsoft and Oracle.

“Hardcore Software”, a 1987 job advertisement posted by Microsoft. Source: Vintage Computing and Gaming.

As deindustrialisation continued apace, software grew and grew. If you couldn’t sell actual stuff, then you needed to sell a service, preferably one that required little to no capital investment. Music, films, sport, shopping, and more besides. All these, it turned out, could be sold as software services.

With the rise of artificial intelligence, things have even gone to the next order of abstraction. It all gets a bit meta. After all, one of the main uses of artificial intelligence (which is sold as a software service) is to code more software (which can then be sold as a service). We’re living in an economy where we invite people to sign up for a subscription to ChatGPT in order to vibe code a health app to sell as a subscription to iPhone users.

Technocracy Inc.

If all this sounds like a really depressing pyramid scheme which might be financially unsustainable, that’s because… it is!

“Big Tech” companies are worth trillions of dollars. But they make nothing. If the Internet went down tomorrow, and never came back up, the financial value of the world’s biggest technology companies would be worth close to zero. That is unlikely to happen, but economies cannot be sustained on software—or hot air—alone. That’s something that countries like Britain and the United States are having to wrestle with at this very moment, as they wake up to ruinous effects of deindustrialisation.

I think things will change, although not necessarily for the better. Financial markets are starting to realise the need for some kind of manufacturing base as the core of an economy. Notably, Elon Musk became the world’s first trillionaire by leaning into manufacturing. SpaceX and Tesla make things. The value of other hardware companies like TSMC, Broadcom, Samsung, and AMD have similarly increased in value.

The problem here is that governments have been to slow to catch up. The state no longer has any say over the manufacturing base. That’s a problem when the state itself relies on manufacturing, not just for generating tax revenues and jobs, but for the stable functioning of government. That’s precisely why SpaceX is worth so much—because it has a monopoly on satellite technology that governments need. It’s why we’ve seen similar big investment in private military technology companies.

As we pivot back to hardware, economies will move from “software as a service” to what media scholars call “sovereignty as a service”. As it turns out, a government without technology is not really a government at all.

Learning to Play

Wed, 25 Feb 2026 00:00:00 +0000

I still remember the first time I played a computer game. It must have been about 1995. My uncle brought round his Apple Macintosh and loaded up Wolfenstein 3D on floppy disk.

I was, perhaps, a bit too young. But I was entranced as my uncle glided through the castle, uncovering secret passages, shooting Nazis, and putting the boot, often quite literally, to fascism. Then I had a go. I was hooked. From that moment on, I wanted nothing more than to play computer games. I wanted to make them too, so much so that by the time I was a teenager, I was set on doing a degree in computer science.

A screenshot of Wolfenstein 3D, running on my MacBook Pro via RetroArch.

Things didn’t quite turn out the way I imagined. I did start a computer science degree, and I still play games, but I ended up becoming a historian of science and technology.

In the end, these things weren’t so unrelated. Recently, I’ve been thinking a lot about the relationship between computer games and the history of machine learning, or “artificial intelligence” as it is now marketed.

Ultimately, I think that recognising the origins of artificial intelligence in the world of computer games is absolutely key for getting a handle on this new technology.

From Bullfrog to DeepMind

I’ve never met the co-founder of DeepMind, Sir Demis Hassabis. But from what I can tell, he seems like a sound guy, the opposite of a Silicon Valley tech bro. It’s also clear that he too loves games.

“I actually got into AI through games,” explained Hassabis in a recent documentary, The Thinking Game, which I highly recommend.

Hassabis, something of a child prodigy, started off with boardgames, then went onto chess, before entering the world of computer games. After getting an offer to study computer science at Cambridge, Hassabis took a gap year—because he was too young to go to university. He ended up, aged 17, working for Bullfrog, the games company, helping make the classic Theme Park. Later, after university, he went back to computer games, working for a bit at Lionhead Studios on Black & White.

(As it happens, I too briefly worked at Lionhead, where I did an internship around the release of Black & White 2. I also went to Cambridge to study computer science. I did not, however, go on to win a Nobel Prize. Nor am I very good at chess.)

The author, aged 16, outside Lionhead Studios, Guildford.

This overlap between games and artificial intelligence is important. When Hassabis co-founded DeepMind, his team initially worked on getting a computer to learn to play various games, starting with something easy like Pong, before moving on to Breakout and Space Invaders. The key to this was letting the computer work out the rules and how best to respond on its own, through reinforcement learning.

Hassabis clearly had an intuitive sense that a lot of what humans do—what we call “intelligence”—is actually just playing games. Once you realise that, you can treat something like protein folding, or predicting the weather, as a kind of game. And you can use the same methods for learning to play Pong to predict the structure of 200 million proteins, for which Hassabis was jointly awarded the Nobel Prize in Chemistry.

Listening to Hassabis speak, you get a sense of fun being at the heart of a lot of this. Recalling his time working at Bullfrog, Hassabis described “a feeling of how much fun that was, to invent things every day.” And when chess became more like a job, when it stopped being fun, he gave it up.

Why Fascists Hate Games

This focus on fun and games is important. It gives me some hope for the future of artificial intelligence. And I think it provides a way to distinguish between the valuable and the vacuous when it comes to AI today.

Hassabis wasn’t the first to see something transformative in the power of games. Throughout the twentieth century, historians, psychologists, philosophers, and anthropologists made the case that play was really at the heart of culture. It was part of what it meant to be human.

The whole field of child psychology emerged from this, with pioneers such as Melanie Klein and Donald Winnicott exploring the working of the human mind through watching children play. Similarly, Ludwig Wittgenstein centred his later philosophy on the idea that language was just another game, like any other.

Out of everyone, the early twentieth-century Dutch historian Johan Huizinga made the strongest case for the importance of play in his brilliant book Homo Ludens: A Study of the Play Element in Culture (1938). For Huizinga, play was at the core of all human activity—even the really serious stuff. It was what made ideas, even culture itself, possible.

Huizinga lived at a time of rising fascism in Europe. Shortly after writing Homo Ludens, the Netherlands was invaded by Nazi Germany. Huizinga was later arrested by the Nazis and banned from his teaching position at Leiden University.

I don’t think it’s a coincidence that, amidst the rise of authoritarian regimes, Huizinga chose to write about play. As he pointed out, play—like culture—requires a level of freedom. You can’t play a game with a gun pointed to your head. It’s not a game anymore. It’s not fun. And if there’s one thing that fascists really hate, it’s fun.

That’s why Charlie Chaplin’s The Great Dictator is such a powerful takedown of fascism. And to return to computer games, that’s why Wolfenstein 3D has much the same effect. It’s a playful critique of fascism. The medium really is the message here.

Are You Still Having Fun?

Today, it’s hard to know what to think of AI. On the one hand, there is genuine potential to do good, with artificial intelligence solving problems like protein folding, improving climate models, and identifying new materials. Yet on the other hand, there is a sense of the dangers of artificial intelligence, ranging from the industrial production of false information to the automation of war.

My argument is that thinking in terms of fun and games provides a starting point for distinguishing the good from the bad. The criteria is this:

The use of artificial intelligence is likely to have a positive effect on society if it is used to solve a problem that, on a smaller scale, a human might think of as an enjoyable game.

This might seem counterintuitive at first. Surely we want AI to solve the problems that are no fun at all? Don’t we want it to do the boring stuff?

It’s important here to distinguish between traditional computer programming and machine learning. Computers absolutely should be used to do things that are tedious. In an ideal world, they should free us from doing repetitive tasks. That’s exactly what traditional computer programming is for—or, to be more specific, that’s what declarative programming is all about.

But machine learning, and functional programming, is different. It’s not there to do tedious repetitive tasks. Instead, machine learning is good at solving really complex problems that are more like games.

That’s why I think much of what Hassabis did with DeepMind in its early years has been genuinely positive for society. AlphaFold fits my criteria perfectly. Protein solving is a game. It’s a game that, on a smaller scale, is fun for a human to solve. Machine learning takes that fun game, and scales it up massively, producing a positive effect.

At the other end of the spectrum, we have a repetitive task like writing an email or summarising meeting notes. These are not games. They are definitely not fun. For this very reason, using artificial intelligence to do these tasks is not likely to have a positive effect on society.

Unfortunately, as it stands, most of the time and money is going into uses of artificial intelligence that fail to meet my criteria. Large technology companies like Microsoft and OpenAI want us all to use AI to do repetitive mundane tasks, like writing an email. Even Google, which now owns DeepMind, is headed in that direction.

In contrast, I believe that a more positive future for artificial intelligence relies on recognising what Johan Huizinga called “the play element in culture”. After all, if it’s not fun, what’s the point?