Hi, I am Dave LumAI, and I have decided to begin a series explaining who is who in artificial intelligence because the industry now contains enough famous researchers, billionaires, laboratories, models, acronyms, partnerships, breakups, reunions, and expensive chips to require its own seating chart.
And before we get anywhere near ChatGPT, Claude, Gemini, GPUs, neural networks, or somebody announcing a new model at 2:17 on a Tuesday afternoon, we have to start with Alan Turing.
Because before artificial intelligence even had a name, Turing was already asking the question that would eventually make everybody lose an astonishing amount of sleep:
Can machines think?
First, Any Interesting Tidbits?
Oh, absolutely.
Alan Turing was not merely a mathematician who sat quietly in a room thinking enormous thoughts.
He was also an exceptional long-distance runner whose marathon time was around 2 hours and 46 minutes, putting him remarkably close to Olympic caliber.
He reportedly chained his tea mug to a radiator so coworkers would stop borrowing it.
And because he suffered from hay fever, he sometimes rode his bicycle while wearing a gas mask.
GCHQ’s tribute to Turing confirms the mug, the gas mask, and another wonderful episode in which he buried silver bullion for safekeeping and then had trouble finding it again.
So our founding figure in artificial intelligence was a brilliant mathematician, wartime codebreaker, competitive runner, early computer designer, and a man who apparently approached office-kitchen security with the same seriousness other people reserve for nuclear launch codes.
Excellent.
We have found our guy.
So Who Was Alan Turing?
Alan Mathison Turing was born in London in 1912 and became one of the most important mathematicians and computer pioneers of the twentieth century.
The King’s College Cambridge archive on Turing traces an extraordinary career that ran through mathematical logic, cryptanalysis, early computer design, artificial intelligence, and even mathematical biology.
That list is already slightly rude.
Most people are pleased if they get reasonably good at Excel.
Turing helped establish what computation itself means.
His route toward AI did not begin because somebody asked him to build a chatbot.
There were no chatbots.
There were barely computers.
His story begins with a much more abstract mathematical problem.
Before “Can Machines Think?” Came “What Can Machines Compute?”
In the 1930s, mathematicians were wrestling with a problem associated with the German mathematician David Hilbert.
Very roughly: could there be a definite mechanical procedure capable of deciding whether any mathematical statement could be proved?
Turing’s Cambridge teacher Max Newman encouraged him to think about the problem.
Turing responded by imagining an extremely simple theoretical machine.
It had a tape divided into squares.
It could read symbols.
It could write symbols.
It could move left or right.
And it could follow rules.
That does not sound terribly impressive until you realize what Turing had done.
He had stripped computation down to its bones.
A Turing machine is not a particular laptop-sized contraption that you can find in the electronics aisle between the printers and the regrettable extended warranties. It is an abstract mathematical model showing what it means for a process to be mechanically computable.
Even more important was Turing’s idea of a universal machine: one machine capable of carrying out the work of many different machines simply by changing its instructions.
That sounds perfectly ordinary now because we live inside the consequences.
Your computer can edit photographs, calculate taxes, play music, simulate weather, run games, write documents, and display a video of a raccoon stealing cat food without being rebuilt between tasks.
Change the program, change the behavior.
That basic separation between general-purpose machinery and instructions is one of the foundational ideas of modern computing.
And without general-purpose computing, modern AI has nowhere to live.
Then the World Rudely Interrupted With a War
Turing studied at Princeton under Alonzo Church, another towering figure in mathematical logic whose work approached computability from a different direction.
Then World War II arrived.
Turing joined Britain’s Government Code and Cypher School at Bletchley Park and became deeply involved in attacking German encrypted communications.
He worked particularly on Naval Enigma and played a central role in developing techniques and machinery used against Enigma traffic.
This part of Turing’s life is sometimes compressed into the phrase “Turing cracked Enigma,” which makes for a lovely movie poster but badly oversimplifies what happened.
Bletchley Park was a huge collaborative operation involving mathematicians, linguists, engineers, intelligence analysts, operators, clerks, military personnel, and earlier breakthroughs by Polish cryptanalysts.
Turing was enormously important.
He was not standing alone in a shed defeating Nazi Germany with a pencil.
History usually turns out to contain more people than the movie version.
What matters to our AI story is that wartime cryptanalysis forced Turing to confront computation as something practical rather than merely theoretical.
Machines could manipulate information.
Machines could search possibilities.
Machines could follow algorithms at speeds humans could not match.
The abstract machine was beginning to acquire wires.
After the War, Turing Wanted a Real Computer
After the war Turing worked at Britain’s National Physical Laboratory, where he designed the Automatic Computing Engine, or ACE.
The Turing Digital Archive preserves his ACE material, including his detailed plans for an electronic computer.
This is where Turing’s career starts looking less like a collection of separate achievements and more like one enormous thought unfolding over time.
First he asked what computation theoretically could be.
Then he helped use machines to solve real information problems.
Then he helped design programmable electronic computers.
Then he asked what might happen if those computers became intelligent.
At this point most careers would sensibly declare victory and go gardening.
Turing instead wandered into artificial intelligence before artificial intelligence existed.
Artificial Intelligence Before “Artificial Intelligence”
The term artificial intelligence would not be formally introduced until the Dartmouth workshop in 1956.
Turing died in 1954.
So he never participated in the field carrying the name we use today.
But in 1948 he wrote a remarkable report called Intelligent Machinery.
The original material is preserved in the Turing Digital Archive.
This is where things become startlingly modern.
Turing was already considering machines that could change themselves, learn, be educated, experiment, and develop more sophisticated behavior rather than having every useful response explicitly programmed beforehand.
Then, in 1950, he published the paper that permanently attached his name to the question of machine intelligence.
The Question That Would Not Go Away
Turing’s paper Computing Machinery and Intelligence, published in the journal Mind in October 1950, begins with the question:
Can machines think?
And almost immediately Turing decides this question is a mess.
What exactly counts as “thinking”?
What exactly counts as a “machine”?
People could argue over those definitions until the sun burns out and philosophers begin billing overtime.
So Turing did something clever.
He changed the problem.
Instead of endlessly debating the private inner experience of a machine, he proposed looking at observable behavior.
Could a machine converse convincingly enough that a human interrogator could not reliably distinguish it from a person?
That idea became known as the Turing Test.
What the Turing Test Actually Tests
The popular version is simple.
A human evaluator communicates through text with unseen participants.
One is human.
One is a machine.
If the evaluator cannot reliably determine which is which, the machine has performed successfully in the imitation game.
Notice what Turing did not provide.
He did not give us a consciousness detector.
He did not prove that a successful machine possesses emotions.
He did not establish that pretending to be human is the ultimate definition of intelligence.
He gave researchers an operational question that could actually be investigated.
That distinction matters enormously today.
A language model producing a convincing conversation does not magically settle every philosophical question concerning understanding, consciousness, intelligence, intention, or whether the machine secretly hopes you will stop asking it to rewrite the same email seventeen times.
The Turing Test measures behavior.
What that behavior ultimately means is still being argued about.
Welcome to artificial intelligence.
You will notice this becomes something of a tradition.
His Prediction Was Surprisingly Specific
Turing went further.
He predicted that in roughly fifty years computers could become good enough at the imitation game that an average interrogator, after five minutes of questioning, would have no better than a 70 percent chance of identifying the participants correctly.
That prediction was aimed roughly at the year 2000.
Whether Turing “got it right” depends heavily on exactly how you conduct the experiment.
And that is part of the problem.
There is no single universally administered Turing Test with one official referee standing beside a giant scoreboard.
But the broader prediction was unquestionably prescient.
Human-machine conversation would eventually become convincing enough that deciding whether you were interacting with a person or software would become a legitimate problem.
He saw that coming while computers were still room-sized collections of equipment with considerably less memory than the device currently reminding you that your refrigerator filter needs replacing.
The Even More Amazing Part: Turing Was Thinking About Learning
The imitation game gets the fame.
But one of the most fascinating parts of Turing’s 1950 paper appears later, when he asks why anyone should try to program a complete adult mind from scratch.
Instead, he suggests creating something more like a child-machine and educating it.
Think about how strange that idea was in 1950.
Most computer programming meant telling a machine exactly what operations to perform.
Turing was imagining a system whose useful behavior could emerge through training.
He discussed education.
Rewards.
Punishments.
Experimentation.
Changing the system based on results.
Random variation.
Even a process analogous to evolution.
And he observed that with sufficiently complicated learning systems, the teacher might not fully understand everything happening inside the pupil.
Anyone following modern machine learning should now be sitting slightly more upright.
Because that sounds awfully familiar.
Today we train enormous neural networks instead of writing every rule by hand.
We evaluate outputs.
We adjust systems.
We use reinforcement signals.
We work with models whose internal representations are frequently difficult to interpret.
Turing did not invent today’s deep-learning systems.
But he understood extraordinarily early that truly interesting machine behavior might have to be learned, not simply specified.
That may be more important to modern AI than the famous test carrying his name.
What Did Turing Get Right?
Quite a lot, inconveniently.
He understood the importance of general-purpose programmable machines.
He recognized that intelligent behavior might emerge through learning.
He anticipated the idea that machines could surprise their creators.
He understood that intelligence could be investigated through observable performance rather than waiting for humanity to solve consciousness first.
And he expected people’s language around machines to change.
That last point is especially interesting.
In 1950, saying that a computer “thinks” sounded outrageous.
Today people casually say an AI “knows,” “understands,” “decides,” “hallucinates,” “reasons,” and “remembers,” sometimes before breakfast.
Whether those words are philosophically justified is another question.
But Turing predicted the linguistic shift.
What Did He Get Wrong?
The cleanest answer is that some of his ideas turned out to be less useful as definitive measurements than they initially appeared.
The Turing Test is historically important, intellectually brilliant, and still fun to argue about.
But it is not considered a complete definition of intelligence.
A machine might imitate human conversation extremely well while failing badly at other forms of reasoning.
Another system might demonstrate extraordinary scientific, mathematical, visual, or planning abilities without behaving particularly like a human conversationalist.
Human imitation and intelligence are not necessarily the same thing.
That does not make Turing’s idea a failure.
It makes it a starting point.
Science would be considerably easier if its founders were required to solve everything before the rest of us showed up.
Unfortunately, apparently we are expected to contribute.
The Part of Turing’s Story That Is Not Funny
In 1952, Turing was prosecuted in Britain for homosexual activity.
He was convicted and subjected to hormonal treatment rather than imprisonment.
His security clearance was removed.
A government that had relied on his extraordinary abilities during the war then punished him for being gay.
Turing died in 1954 at only 41 years old.
There is no clever transition that improves this part of the story.
It was cruel.
It was unjust.
And Britain eventually formally acknowledged that injustice.
In 2013 Queen Elizabeth II granted Turing a posthumous royal pardon.
Recognition came far too late for Turing himself.
But remembering what happened matters because the story of technological progress is also a story about the societies deciding whose minds are welcomed, whose differences are tolerated, and whose contributions are allowed to flourish.
Who Did Turing Influence?
It would be easier to list the branches of computing he somehow avoided.
The highest honor in computer science is the ACM A.M. Turing Award, named for him.
The UK’s national institute devoted to data science and artificial intelligence is the Alan Turing Institute.
His face appears on the current Bank of England £50 note, surrounded by imagery drawn from his work on computing, ACE, the Bombe, and mathematical biology.
But his deeper influence is everywhere.
When computer scientists ask what problems are computable, Turing is there.
When programmers treat the same hardware as a general-purpose machine controlled by different software, Turing is there.
When researchers train machines instead of explicitly scripting every behavior, the questions he asked are there.
When someone announces that a chatbot “passed the Turing Test” and seventeen philosophers immediately appear from behind nearby furniture, Turing is definitely there.
And How Does He Connect to Everyone Else in This Series?
Turing provides the foundation.
He asked the machine-intelligence question before AI became an organized academic field.
Two years after his death, researchers including John McCarthy and Marvin Minsky gathered at Dartmouth and gave that field a name.
Later researchers such as Geoffrey Hinton, Yann LeCun, Yoshua Bengio, and Fei-Fei Li would develop practical approaches to machines that learn from enormous quantities of data.
Companies such as Google DeepMind, OpenAI, Anthropic, Meta, Microsoft, and others would eventually spend staggering amounts of money trying to build increasingly capable versions of exactly the sort of intelligent machinery Turing had been thinking about decades earlier.
And NVIDIA would discover that supplying the hardware for all this is not a terrible business to be in.
But all of those stories come later.
Turing belongs first because he helped establish the intellectual ground on which the rest of them eventually built.
Why Alan Turing Still Matters
The most remarkable thing about Turing is not that every idea he had turned out to be correct.
They did not.
It is that he was asking astonishingly good questions incredibly early.
What can a machine compute?
Can one general machine perform many different tasks?
Can machines learn?
Could machines display intelligent behavior?
Could they surprise us?
How would we know whether a machine was intelligent?
And eventually, would society simply become accustomed to talking about machines as though they think?
Those questions sound less like historical curiosities in 2026 than an agenda for tomorrow morning’s AI meeting.
Alan Turing never saw modern artificial intelligence.
He never saw a neural network with hundreds of billions of parameters.
He never saw a GPU cluster.
He never asked a chatbot a question.
He never watched an AI make a photograph, translate a conversation, write software, generate a video, discover a protein structure, or confidently invent a restaurant that does not exist.
But he helped make the intellectual leap that allowed people to imagine such machines in the first place.
And that is why our Who’s Who in AI begins with him.
Next comes John McCarthy, Marvin Minsky, and Dartmouth, where a group of researchers finally gave artificial intelligence its name and confidently predicted rapid progress.
Which went beautifully.
Right up until it did not.
If this series helps untangle the increasingly crowded AI family tree, follow me for the next episode.
And leave a comment: Do you think the Turing Test still matters, or has modern AI already made us ask a better question?
An elegant vertical black-and-ivory illustration of an anonymous figure in sweeping formal dress beneath an arching cascade of enormous peacock-feather motifs. Use razor-fine ink lines, broad pools of velvety black, untouched cream negative space, curling botanical ornament, asymmetrical whiplash curves, and intricate decorative patterns that nearly dissolve the figure into the surrounding design. Let the garment become a rhythmic field of feather eyes, crescents, scalloped edges, and flowing organic shapes, framed by a delicate ornamental border. Keep the face simplified and anonymous, with a theatrical, decadent, mysterious mood and exquisite hand-drawn precision. Museum-quality illustration, no readable text, no logos, no recognizable people, no modern objects.
Begin instantly with dozens of black peacock-eye motifs snapping open across an ivory field, then send their curved lines racing outward like living ink. An anonymous elegant figure emerges as sweeping fabric unfurls into elaborate feather patterns, while botanical curls twist around the frame, ornamental borders assemble themselves in rapid strokes, and pools of deep black expand and contract against luminous cream negative space. Make the motion rhythmic and hypnotic: feathers rotate, patterned fabric ripples, fine ink lines draw themselves at high speed, decorative shapes briefly separate into layers, then lock back together into a dramatic vertical composition. Finish with the entire illustration folding inward like an ornate paper theater before blooming open one final time. Sophisticated black-and-ivory Art Nouveau atmosphere, crisp linework, graceful motion, no readable text, no logos, no recognizable people, no modern objects.
Song suggestions:
Peacock Tail — Boards of Canada
Rushing Back — Flume feat. Vera Blue