The Baby: Where the Computer Was Born

The Baby that started everything

In June 1948, a room full of valves at the University of Manchester ran a program that it was holding in its own memory. The machine had a proper name, the Small-Scale Experimental Machine, but everyone just called it the Baby. It wasn't much to look at, and it could store about as much as a single text message. The clever bit was that it kept its instructions in memory next to its data, and could rewrite them as it went. That one idea is what every computer still runs on today.

We tend to assume computing began somewhere in California. It didn't. It started here, in a lab in the north of England.

So why Manchester?

It made a certain sense. Manchester had been an industrial city for a hundred years, so building machines to solve awkward problems was already part of how the place thought. The two men behind the Baby, Freddie Williams and Tom Kilburn, had spent the war working on radar and knew electronics inside out. Alan Turing turned up not long after. The university was willing to fund the work back when it still sounded faintly ridiculous, and a local engineering firm, Ferranti, took the design and built it into the Mark 1, the first computer you could actually buy, in 1951. For a few years Britain was genuinely ahead of everyone.

From first to forgotten

1948	Williams, Kilburn and Turing run the world's first stored program. Britain is out in front.
1951	Ferranti ships the Mark 1, the first computer anyone could buy.
1960s	American money and bigger salaries start pulling the talent across the Atlantic.
1970s–80s	Manchester's factories close, and the city spends years just trying to recover.
Today	There's a Silicon Valley. There isn't a Manchester one.

So what went wrong?

And yet the city that invented the computer never became a computing city. There is no Manchester version of Silicon Valley. What happened wasn't one big disaster. It was a handful of smaller things that piled up over a few decades.

• Good at inventing, hopeless at selling. Britain has a long habit of coming up with something brilliant and then watching other people make the money from it. Computing was a classic example. American firms like IBM understood that a computer is only worth as much as the salesmen, the training and the support around it. We built clever machines and got out-organised.
• Never enough money, and too many committees. Post-war Britain was broke. The funding for computing was small, scattered across too many projects, and slow to arrive. In America, Cold War defence budgets were pouring into the same field with far fewer questions asked.
• The talent went where the work was. A sharp graduate could earn a lot more, with much better kit, in California. Plenty of them went. We trained the people and America hired them.
• The city's industry fell apart. From the 1970s onwards, Manchester's factories closed one after another. When a place is dealing with that kind of unemployment, paying for research that might not work is the first thing to get cut.
• Universities started counting students. As money got tighter, a dependable way to bring it in was to fill seats, increasingly with international students paying full fees. There is nothing wrong with that by itself, but it quietly shifts a university's energy towards teaching large numbers rather than gambling on the sort of research that produced the Baby in the first place.
• The country bet on the City instead. By the 1980s the ambitious money in Britain was flowing into London finance, not northern engineering. Somewhere along the way the national story stopped being about making things.

Why it still stings

None of this was really about talent. The people who built the Baby were as good as anyone in the world. The trouble is that inventing something and building an industry around it are two completely different jobs, and Britain kept turning out to be excellent at the first and useless at the second.

If you want to see where it all started, there's a working replica of the Baby at the Science and Industry Museum in Manchester. They switch it on regularly, and it still grinds through its sums exactly the way it did in 1948. Standing in front of it, it's worth remembering that being first counts for very little if you can't hang on to the lead.

Manchester gave the world the computer. It just never managed to keep it.

Computing and AI through the years

It helps to zoom out. Here are the major moments in computing and AI, with Manchester's own marked in colour. Read down the list and the shape of the story is hard to miss.

Manchester milestones

1936	Alan Turing's paper "On Computable Numbers" sets out the maths every computer is built on.
1943	McCulloch and Pitts describe the first artificial neuron, the starting point for neural networks.
1945	Von Neumann's EDVAC report sketches out the stored-program idea on paper.
1948	Manchester's Baby runs the first stored program. The idea becomes a working machine.
1950	At Manchester, Turing publishes the Turing Test, asking whether machines can think.
1951	Ferranti Mark 1 (Manchester): first computer sold commercially, first recorded computer music, first game-playing program.
1956	The Dartmouth workshop coins the term "artificial intelligence."
1957	Rosenblatt builds the Perceptron, the first neural net you could train.
1962	Manchester's Atlas pioneers virtual memory, an idea still in every device today.
1969	ARPANET, the ancestor of the internet, goes live.
1971	Intel's 4004 squeezes a whole computer onto a single chip.
1973	The Lighthill report tips Britain into its first "AI winter."
1986	Backpropagation, with Geoffrey Hinton among the authors, brings neural networks back.
1989	Tim Berners-Lee invents the World Wide Web.
1997	IBM's Deep Blue beats world chess champion Garry Kasparov.
2004	Graphene is first isolated at Manchester. A Nobel Prize follows in 2010.
2012	AlexNet kicks off the deep-learning boom.
2016	DeepMind's AlphaGo beats a Go world champion.
2017	The paper "Attention is all you need" introduces the transformer, the engine behind today's AI.
2018	Manchester's SpiNNaker, a brain-inspired million-core supercomputer, switches on.
2022	ChatGPT puts large language models in front of everyone.

Manchester is all over the early years: the first stored program, the Turing Test, the first computer for sale, the first computer music, virtual memory. After that, the big moments mostly happen in America. The city still turns out the occasional world-beater, graphene and the SpiNNaker brain computer among them, but for the last sixty years the real centre of gravity in computing and AI has sat in California.

So what now? The pragmatic bit

It would be easy to leave this as a lament, but that would be unfair on the city, because most of what went wrong has a modern fix, and Manchester has quietly started on several of them. AI has handed every serious university city a second chance at exactly the kind of moment the Baby was. The question is whether Manchester plays it better this time. The pragmatic list looks like this.

• Let the spinouts go. For decades UK universities took founding stakes in their own spinouts several times larger than anything Stanford or MIT would ask, and a large slice held by an institution that doesn't work at the company is precisely the thing that makes a seed investor walk away. The government's 2023 review of spin-outs told universities to take 10% or less of software companies, and stakes are finally falling across the sector. Manchester should go further: take a token stake or none, sign the licence in weeks rather than months, and judge its tech transfer arm on how many companies carry the university's name, not on what it claws back from them. The winners repay universities in donations, labs and hiring, never in dividends — I've written the longer version of that argument here.
• Give the new machinery teeth. At the end of 2024 the university launched Unit M, an innovation unit under Lou Cordwell with a mandate to plug its research into the city's economy. That is the right instrument, but only if it is measured on the numbers a founder actually feels: spinouts formed, time from lab to licence, and graduates still living in Manchester five years on. If it turns into a strategy team that publishes reports, it will have failed politely.
• Fund the gap at home. The Baby-era problem was scattered money arriving slowly; the modern version is that venture capital pools in London and California. The counterweights finally exist: Northern Gritstone, set up by the universities of Manchester, Leeds and Sheffield, has more than £380m to put into northern spinouts, much of it anchored by northern pension funds, and the British Business Bank's £660m Northern Powerhouse Investment Fund II writes cheques from £25k up to £5m. The pragmatic step now is speed and visibility: a good founder should be able to raise a first round without buying a train ticket.
• Make Sister a factory, not a landlord. The university's old North Campus is becoming Sister, a £1.7bn innovation district built with Bruntwood SciTech: four million square feet and a promised 10,000 jobs, a short walk from Piccadilly. Innovation districts fail the same way everywhere, by measuring success in square feet let. This one works if it becomes the cheapest, fastest place in Britain to go from a paper to a company, with the university's labs, lawyers and first customers all inside the same few streets.
• Keep the graduates — and catch the leavers. The 1960s drain was American salaries and better kit; today's is a train to London. Except the flow has started running the other way: London now loses tens of thousands of people in their thirties every year, priced out and squeezed for space, and Manchester is the city they ask about most — more than 40% of London leavers now head for the North and Midlands, up from a quarter in 2019. That is trained, experienced talent turning up of its own accord, the very thing the city spent seventy years losing. The trick is to give them something better to do than remote work from a cheaper flat. With one of Europe's largest student populations at one end and an inflow of seasoned operators at the other, locally funded spinouts are what turn the two into an ecosystem rather than a coincidence.

None of this needs a moonshot or a minister. The pieces are already on the board; what's been missing for seventy years is the posture — the willingness to back winners and then let them win. Manchester was first once. Being first again isn't on offer, but keeping the lead the next time it builds something great still is.