Short answers, pulled from the story.
Artificial intelligence was founded as an academic discipline at a workshop at Dartmouth College in 1956. The first AI program, Logic Theorist, was presented there, created by Allen Newell and Herbert A. Simon in collaboration with J. C. Shaw.
An AI winter was a period when funding for AI projects was difficult to obtain. One followed 1974, when the U.S. and British governments cut off exploratory research after criticism from Sir James Lighthill and pressure from the U.S. Congress.
The sudden success of deep learning between 2012 and 2015 came from two factors, not a new theoretical breakthrough: a hundred-fold increase in speed from switching to GPUs, and vast curated training datasets such as ImageNet. Growth accelerated further after 2017 with the transformer architecture.
Deep Blue beat reigning world chess champion Garry Kasparov on the 11th of May 1997. AlphaGo won 4 out of 5 games of Go against Lee Sedol in March 2016, then defeated Ke Jie, the world's best Go player, in 2017.
A single ChatGPT search uses ten times the electrical energy of a Google search. The International Energy Agency warned that power demand for data centers, AI, and cryptocurrency might double by 2026, an increase equal to the entire power consumption of Japan.
Stephen Hawking, Geoffrey Hinton, Bill Gates, Elon Musk, and others warned that AI could pose an existential risk, with a 2023 statement ranking it alongside pandemics and nuclear war. Others disagreed, including Jürgen Schmidhuber, Andrew Ng, and Yann LeCun, who said intelligent machines will usher in a new renaissance for humanity.
Algorithmic bias occurs when machine learning systems learn from biased data. In 2015 Google Photos labeled two black people as gorillas, and in 2016 the COMPAS court program was found to overestimate the chance a black defendant would reoffend, despite never being told defendants' races.