Artificial intelligence feels like the newest of technologies, but the dream behind it is very old. Long before computers, people imagined machines that could move, speak and think. This course follows that dream from ancient myth to the chatbots of today, and asks what each generation got right, what it got wrong, and why progress so often came in sudden bursts after long frustration. Here the focus is on how the field came to be, more than on the ethics of AI.
Myths and legends
Greek myth told of Talos, a giant bronze man made by the smith god Hephaestus to guard the island of Crete, who walked around the island three times a day and hurled rocks at invaders. In Homer's Iliad, Hephaestus is attended by golden servants made in the form of young women, who could speak and think. Jewish folklore told of the golem, a figure of clay brought to life by sacred words; a famous legend, recorded centuries later, credited Rabbi Judah Loew of Prague in the sixteenth century with making one to protect his community. These stories already raised questions that remain: can made things be trusted, and who controls them?
Clever machines
Real machines that imitated life also have a long history. Greek engineers in Alexandria built automata powered by water and air. In 1206, the engineer Ismail al-Jazari, working in what is now southeastern Turkey, completed his Book of Knowledge of Ingenious Mechanical Devices, describing water clocks, a musical band of automaton figures, and other programmable devices using cams and pegs. In eighteenth century Europe, craftsmen built mechanical figures that could write, draw or play music, amazing audiences.
The Mechanical Turk
In 1770, the Hungarian inventor Wolfgang von Kempelen unveiled a chess-playing machine dressed as a Turkish sorcerer. The Mechanical Turk toured Europe and America for decades and defeated many opponents, reportedly including Napoleon and Benjamin Franklin. It was a hoax: a skilled human chess player hid inside the cabinet. The machine was destroyed in a fire in 1854. Its story is a lasting warning that apparently intelligent machines may hide human labor, a point that still applies to AI systems that depend on armies of human workers labeling data.
Babbage and Lovelace
The first design for a general-purpose computer came from the English mathematician Charles Babbage. In the 1830s he designed the Analytical Engine, a mechanical machine that could be programmed with punched cards to carry out any calculation. It was never completed, partly because of cost and engineering difficulties. But it inspired Ada Lovelace, a mathematician and the daughter of the poet Lord Byron. In 1843, she published extensive notes on the engine, including a step-by-step method for calculating a sequence of numbers called Bernoulli numbers, often described as the first published computer program.
Lovelace saw that the engine could manipulate not just numbers but any symbols, and might even compose music. But she also set a limit that would be debated for over a century: “The Analytical Engine has no pretensions whatever to originate anything. It can do whatever we know how to order it to perform.” Alan Turing later called this “Lady Lovelace's objection.”
Logic as calculation
In 1854, the English mathematician George Boole published The Laws of Thought, showing that logical reasoning could be expressed in algebra using only true and false values. Boolean logic later became the foundation of digital circuits. The dream that reasoning itself might be mechanized had been expressed earlier by the German philosopher Gottfried Wilhelm Leibniz, who imagined a universal language in which disputes could be settled by calculation.
Robots
The word robot was introduced in 1920 by the Czech writer Karel Čapek in his play R.U.R. (Rossum's Universal Robots), from the Czech word robota, meaning forced labor. Čapek credited his brother Josef with the word. In the play, artificial workers manufactured to serve humanity eventually rebel and destroy their makers. The story set a pattern of fear about intelligent machines that science fiction has repeated ever since.
Worked example: is it really intelligent?
Imagine you are a spectator in 1783 watching the Mechanical Turk defeat a skilled chess player. How could you test whether the machine was truly thinking?
- Look inside. The exhibitor opened the cabinet doors to show gears, but in sequence, so the hidden player could shift position. Inspection can be staged.
- Test the edges. Ask it to do something outside chess. A real chess machine of the time could not; a human hidden inside might give himself away.
- Ask about mechanism. No known mechanism of the period could plan chess moves. An extraordinary claim needs an explanation.
- Follow the incentives. The exhibitor earned money from the show. That does not prove fraud, but it is a reason for skepticism.
These same questions remain useful when a company announces an amazing AI system: what is actually happening inside, what happens at the edges of its abilities, is there a plausible mechanism, and who benefits from the claim? The history of AI is full of real breakthroughs, and also of hype. Telling them apart is a skill this course will practice.