Vacuum tubes, room-sized computers, and a wild idea: that math could be made to think.
Imagine a time when computers were brand new, as big as a whole room, and people were just starting to wonder: "Could a machine ever THINK like a person?" That's exactly what scientists asked back in the 1940s and 1950s, and their ideas became the very beginning of Artificial Intelligence (AI).
It started in 1943 when two scientists named Warren McCulloch and Walter Pitts wrote a paper showing that brain cells (called neurons) work kind of like tiny on/off switches, just like the parts inside a computer. This was a HUGE idea because it meant maybe a machine could be built to think like a brain! Then in 1950, a brilliant British mathematician named Alan Turing asked a famous question: "Can machines think?" He invented a test (now called the Turing Test) where a person chats with both a human and a computer through messages. If you can't tell which one is the computer, the machine passes!
The biggest moment came in the summer of 1956, when a young professor named John McCarthy invited the smartest computer thinkers in America to Dartmouth College in New Hampshire for a special workshop. There, McCarthy made up the words "Artificial Intelligence" for the very first time. Around the same time, two scientists, Allen Newell and Herbert Simon, created a computer program called the Logic Theorist that could actually solve math problems all by itself. People called it the "first AI program" ever made!
After scientists invented the term "artificial intelligence" at the famous Dartmouth Conference in 1956, the next 18 years were super exciting! Researchers thought computers would soon be able to think just like people. This time is called the "Golden Years" because everyone was full of hope, and the U.S. government (especially a group called DARPA, which is part of the military) gave scientists tons of money to build smart machines.
During this time, the first robots rolled into factories, the first chatbot started "talking" to people, and computers learned to solve puzzles, play games, and even move blocks around on a screen. Famous scientists like Marvin Minsky and Herbert Simon made bold predictions, with Simon saying in 1965 that "machines will be capable, within twenty years, of doing any work a man can do." Spoiler alert: they were a little too optimistic!
By 1974, people realized AI was way harder than they thought, and funding got cut. But the cool inventions from this era are still the foundation of the AI we use today.
ELIZA worked by spotting keywords (like "depressed" or "boyfriend") and turning your sentences into questions. It didn't actually understand anything, but it sure felt like it did!
When the Logic Theorist found a BETTER way to prove a math theorem than the famous mathematicians Russell and Whitehead, Herbert Simon excitedly tried to publish it in a math journal with the computer listed as a co-author... and the journal said NO!
Marvin Minsky's 1951 SNARC machine had 40 fake "brain cells" built from vacuum tubes and could learn how to escape a maze, just like a rat.
The phrase "Artificial Intelligence" didn't even exist until 1955, when 28-year-old John McCarthy made it up for his Dartmouth proposal.
ELIZA pretended to be a therapist, and Joseph Weizenbaum was shocked when his own secretary asked him to leave the room so she could "talk privately" with the computer. She knew it was just a program, but it still felt real!
Shakey got its name because it wobbled and shook a lot when it moved. It had a TV camera for eyes and was so slow that planning one move could take hours.
Unimate weighed 4,000 pounds and cost $25,000 in 1961 (about $250,000 today). It later appeared on The Tonight Show with Johnny Carson, where it putted a golf ball and poured a beer!
Operations per second — 1956 vs. today (logarithmic scale).
I propose to consider the question, "Can machines think?"— Alan Turing, 1950