Although AI may still feel like something new, the term itself was born more than seven decades ago, during a modest proposal for a summer research project at Dartmouth that carried a budget request of $13,500.
It was an academic document, not a manifesto, but it quietly laid the foundation for one of the most consequential technological movements in human history.
The sad irony is that the field’s most famous philosophical ancestor, Alan Turing, was already gone by this point.
Turing had asked the defining question years earlier — “can machines think?” — and designed what became known as the Turing Test, a method to judge whether a machine could convincingly imitate human thought.
His work framed the entire discussion, yet he died in 1954, two years before the Dartmouth meeting that officially named the field he had helped imagine.
Turing’s death followed his prosecution in the UK for homosexuality, then criminalized, and he died from cyanide poisoning in what was widely ruled a suicide — a loss that removed one of computing’s most original thinkers just before his ideas began reshaping science.
Imitation game
Long before artificial intelligence had a name, Turing had already come up with the question that would define it. In his 1950 paper Computing Machinery and Intelligence, he proposed what became known as the Turing Test, or “imitation game,” replacing abstract debates about whether machines could truly think with a simpler challenge: could a machine hold a written conversation well enough that a human judge would be unable to reliably tell it apart from another human?
By focusing on observable behavior instead of philosophy, Turing turned intelligence into something researchers could actually test.
The idea was strikingly forward-looking given the reality of computers at the time. Early machines were slow, expensive and limited to mathematical calculation, yet Turing suspected that intelligence might emerge from sufficiently complex symbol processing.
Rather than asking whether machines possessed a mind or consciousness, he asked whether they could convincingly imitate intelligent behavior — something that inspired later researchers to treat thinking as an engineering problem.
That conceptual leap directly influenced the group that gathered at Dartmouth just a few years later, even though the man who posed the question would never see the field formally named.
The Dartmouth Summer Research Project on Artificial Intelligence, organized by John McCarthy with Marvin Minsky, Claude Shannon, and Nathaniel Rochester, was small and ambitious.
According to the proposal, researchers hoped to prove that “every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it.” The goal sounded ambitious then and still does now: language, abstraction, reasoning, and self-improvement, all encoded into machines.
McCarthy would later become one of AI’s most influential voices. In a 1979 issue of ComputerWorld, he said bluntly that the computer revolution “hasn’t happened yet,” even while predicting that it eventually would.
He argued that computers had not yet impacted life in the way electricity or automobiles had, but he believed that applications in the coming decade would initiate a genuine revolution.
McCarthy’s realism often contrasted with the hype that surrounded the field, a tension that has followed AI ever since.
AI as a hot topic
By the early 1980s, interest in AI had surged again, but confusion about what it really meant was widespread.
Writing in a 1984 issue of InfoWorld, reporter Peggy Watt noted that artificial intelligence had become a “hot topic,” with shelves filled with books and software companies racing to label products as intelligent. Yet she warned that “the term is being used and abused widely, almost to the point of losing its usefulness as a description.”
The frustration among researchers was obvious. In that same InfoWorld report, Dr. S. Jerrold Kaplan of Teknowledge said, “Whenever anybody says, ‘I’m selling AI,’ I’m suspicious.”
Kaplan argued that AI was not a single program. “The science of AI is a set of techniques for programming,” he said, describing systems that represented “concepts and ideas, explanations and relationships,” rather than just numbers or words.
This tension between promise and reality also defined the work of Marvin Minsky, one of Dartmouth’s original architects. In a 1981 issue of ComputerWorld, covering the Data Training ’81 conference, Minsky described AI as fundamentally paradoxical: “Hard things are easy to do and easy things are hard to do.”
Computers excelled at calculations that challenged humans, but struggled with common sense, language ambiguity, and contextual understanding.
Minsky explained that “common sense is the most difficult thing to inculcate into a computer.”
Humans absorb countless exceptions and nuances over years of living, but machines require explicit instruction. A logical rule like “birds can fly” breaks down immediately when confronted with dead birds or flightless species — a simple example revealing why intelligence is more than pure logic.
Expert systems
The optimistic early years of AI had already produced striking milestones. The Lawrence Livermore National Laboratory later described how researchers in the 1960s developed programs such as SAINT, an early “expert system” capable of solving symbolic integration problems at the level of a college freshman.
The program solved nearly all the test problems it faced, hinting that machines could emulate specialist reasoning long before modern machine learning.
Yet progress came in waves. Funding boomed in the 1960s as government agencies backed ambitious research, then cooled massively in the 1970s.
The dream of building human-like intelligence proved far harder than expected. Even McCarthy admitted that “human-level” AI was still “several conceptual revolutions away.”
By the time AI returned to the spotlight in the 1980s, companies were marketing expert systems and natural-language tools as breakthroughs.
Some systems impressed users by tolerating spelling mistakes or translating plain English commands into database queries.
Others, however, leaned more on clever engineering than genuine reasoning. As one unnamed researcher quoted in InfoWorld warned, the real test of an expert system was whether it could explain its conclusions.
Still, the vision persisted. Industry observers imagined computers capable of understanding natural language, translating documents, and even correcting grammar automatically.
Kaplan predicted AI would change how people programmed because it was “much more natural to work with symbolic terms than math algorithms.” The idea that software could assist, advise, and collaborate with humans was already taking shape.
Looking back, what stands out is how many early predictions were both wrong and right. McCarthy thought the revolution had not yet arrived, but he believed it would come through practical applications. Minsky warned that common sense would remain stubbornly difficult.
Hmm
Today, as AI systems write text, generate images, and assist scientific discovery, the echoes of those early conversations remain.
The Dartmouth organizers imagined machines that could “use language, form abstractions and concepts, solve kinds of problems now reserved for humans, and improve themselves.” All of which are (mostly) true today.
The $13,500 proposal did not seem remarkable at the time. It was just one funding request among many. Yet it gave a name to an idea that continues to change society, shaped by optimism, frustration, paradox, and unresolved questions.
And perhaps that is the real legacy of artificial intelligence. It began not as a single invention, like the transistor or the microprocessor, but as a wager that intelligence itself could be understood, described, and eventually reproduced.
Seventy-one years later, humanity is still testing that idea, still arguing about definitions, and still pursuing the vision imagined by twentieth-century minds who believed thinking machines might one day become real.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds. Make sure to click the Follow button!
And of course you can also follow TechRadar on TikTok for news, reviews, unboxings in video form, and get regular updates from us on WhatsApp too.
https://cdn.mos.cms.futurecdn.net/daca50fd1567c1e9f28bb4e32cb6170d-640-80.jpg
Source link
waynewilliams@onmail.com (Wayne Williams)
