Short answers, pulled from the story.
The AI winter began in 1984 when Roger Schank and Marvin Minsky warned business leaders about unchecked enthusiasm. This public debate triggered a chain reaction of pessimism that spread to the press and led to severe funding cuts.
Machine translation research ended abruptly after the National Research Council released its famous report on the 2nd of May 1966. The council concluded that machine translation was more expensive, less accurate, and slower than human translation following an expenditure of approximately 20 million dollars.
Neural network projects lost major funding because Marvin Minsky and Seymour Papert published their book Perceptrons in 1969 which emphasized the limits of perceptron capabilities. Nobody knew how to train multilayered perceptrons at that time so interest faded until backpropagation algorithms emerged years later.
Professor Sir James Lighthill delivered a report to the UK Parliament in 1973 that criticized the utter failure of AI to achieve grandiose objectives. This report led to the complete dismantling of AI research across most British universities until Alvey began funding again from a war chest of £350 million in 1983.
The market for specialized LISP-based AI hardware collapsed in 1987 three years after predictions made by Roger Schank and Marvin Minsky. An entire industry worth half a billion dollars was replaced in a single year as workstations offered powerful alternatives to LISP machines.