Artificial intelligence as a concept took shape in 1943, when neuroscientists Warren McCulloch and Walter Pitts proposed that networks of artificial neurons could perform logical functions.
If you’re curious about the history of AI, you're in the right place. In this article, we’re going to dive deep into the history of AI and explore when and how artificial intelligence was created.
1950s: When Early Conceptual Foundations of AI were created
In 1950, British mathematician Alan Turing published “Computing Machinery and Intelligence,” a landmark paper that asked “Can machines think?”. Turing outlined a test (later known as the Turing Test) to judge a machine’s intelligence by its ability to exhibit human-like conversation
The actual term “artificial intelligence” was first coined in 1955 by John McCarthy, a young professor at Dartmouth College. McCarthy, along with colleagues Marvin Minsky, Claude Shannon, and Nathaniel Rochester, drafted a proposal for a summer research workshop to explore how machines could simulate human intelligence.
This 1955 proposal envisioned a “2 month, 10 man study of artificial intelligence” and conjectured 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”
1956: Dartmouth Conference and AI as a Field
Dartmouth Summer Research Project is an event widely considered the birth of artificial intelligence as a field.
Organized by John McCarthy (Dartmouth), Marvin Minsky (Harvard), Claude Shannon (Bell Labs), and Nathaniel Rochester (IBM), this workshop gathered about a dozen pioneering scientists to discuss and program intelligent machines.
The meeting’s goal was to map out how to make machines “use language, form abstractions and concepts, [and] solve kinds of problems now reserved for humans”
Although no singular “thinking machine” emerged that summer, the Dartmouth conference crystallized AI as an academic field. Attendees at Dartmouth (who would become leaders in AI) included McCarthy, Minsky, Allen Newell, Herbert Simon, Arthur Samuel, Oliver Selfridge, and others.
1956–1958: Early AI Programs
Following Dartmouth, researchers created many programs to demonstrate machine intelligence:
- Logic Theorist (1956): A program by Allen Newell and Herbert A. Simon (with Cliff Shaw) that could prove mathematical theorems. It successfully proved 38 of 52 theorems from Principia Mathematica, in some cases finding new proofs.
- Samuel’s Checkers Program (1952–1956): Developed by Arthur Samuel at IBM, this checkers-playing program was a pioneering experiment in machine learning. Samuel’s program could play checkers and improve its performance over time by learning from experience. By 1955–1956 it was playing at a competitive level.
- Rosenblatt’s Perceptron (1957): The perceptron was an electronic device that learned to recognize simple patterns (like shapes) through trial and error, adjusting connection weights based on feedback. It showed that machines could learn to classify information in a way loosely inspired by brain neurons.
- LISP Programming Language (1958): To support AI research, John McCarthy created LISP (LISt Processing) in 1958, a high-level programming language for symbolic computation, which allowed programs to manipulate symbols and lists. Many early AI programs (including McCarthy’s own research on common-sense reasoning) were written in Lisp.
1960s: Expansion of AI Research
By the early 1960s, artificial intelligence was firmly established in universities and research institutes. In 1959, McCarthy and Minsky founded the MIT AI Lab — this was the start of coordinated AI research at MIT.
McCarthy then moved to Stanford and in 1962 founded the Stanford Artificial Intelligence Lab (SAIL), helping make AI a full-fledged academic discipline on the West Coast.
During the 1960s, several notable AI projects emerge:
- DENDRAL (1965): The first expert system, developed at Stanford University by Edward Feigenbaum, Bruce Buchanan, Joshua Lederberg, and colleagues. DENDRAL was designed to help organic chemists by analyzing mass spectral data to deduce molecular structures. DENDRAL’s success showed that AI could replicate specialized problem-solving skills.
- ELIZA (1966): An early chatbot developed at MIT by Joseph Weizenbaum. It simulated a psychotherapist by rephrasing user inputs and could hold a simple conversation.
- Shakey the Robot (1966–1972): This project at Stanford Research Institute (SRI) led by Charles Rosen, Nilsson, et al., was the first mobile robot to reason about its actions. Shakey could navigate a room, perceive objects, plan simple actions, and execute them (e.g. rolling forward and pushing a block).
Bottom line
By the end of the 1960s it became an active field of research, had a growing community of AI researchers, and hosted regular conferences and publications.
The first International Joint Conference on Artificial Intelligence (IJCAI), for example, took place in 1969, and AI researchers from around the world gathered there. Universities established AI research programs, and funding from agencies like ARPA in the US poured into projects.
The term “AI” itself was coined in 1955, but these formative years through the 1960s saw the creation of the fundamental ideas and systems that modern Chatbots like ChatGPT operate on.