History of AI
In 1950, Alan Turing developed the Turing test, or "Imitation Game" to measure when we can declare that machines can be intelligent. The original test involved a judge that would have text-only interaction between a machine and another human person. If the judge could not distinguish between the machine and the human, the machine was considered to have passed the test.
In 1951 the first neural network was developed to help a virtual rat solve a maze puzzle. The AI used vacuum tubes, motors, and clutches rather than microchips and transistors.
In the 1980's, researchers switched from a "rule-based" approach in neural networking to statistics and machine learning. This meant that instead of trying to imitate the rules that humans use to operate, they instead took a trial-and-error approach and adjusted probabilities based on feedback to teach machines to think.
In 1997, IBM's Deep Blue defeated the world chess champion Garry Kasparov. With every update on the chess board Deep Blue was able to process thousands of possible moves every second. Deep Blue's victory demonstrated the true power of artificial intelligence over human intelligence in such circumstances.
In 2002 the commercial use of AI was realized with products like iRobot's Roomba.
​
In the "Modern Implementation" section we will go into further depth about how AI is commonly being used today.
