1956 In the summer, at an important meeting held at Dartmouth College in the United States, scientists led by McCarthy, Minsky, Rochester and Shannon studied and discussed a series of problems of simulating intelligence with machines.
The term "artificial intelligence" was put forward for the first time, which marked the official birth of artificial intelligence, a new discipline. Since then, artificial intelligence has experienced many highs and lows in its development history.
After artificial intelligence was put forward in 1956, researchers boldly made optimistic predictions. Herbert Simon, one of the participants in Dartmouth Conference, made a more specific prediction: in 10, computers will become chess champions and machines will prove an important mathematical theorem.
Simon and others are so confident that their predictions have not been realized within the predicted time, and they are far from being realized. These failures have caused great damage to the reputation of artificial intelligence.
197 1 year, James, a mathematician at Cambridge University, published a comprehensive report on artificial intelligence according to the wishes of the British government, claiming that "artificial intelligence research is nothing if it is not a scam".
Under the influence of this report, the British government cut the research funding of artificial intelligence and dissolved the research institutions of artificial intelligence. The research enthusiasm of artificial intelligence was poured cold water for the first time.
In 1990s, artificial intelligence entered the second winter, which was represented by the failure of the fifth generation robot research and development and the failure of neural network in Japan.
Until the beginning of 2 1 century, the combination of deep learning and internet big data made artificial intelligence once again usher in a new spring. After the emergence of a large number of breakthrough achievements such as Alpha Go, human beings have rekindled their enthusiasm for whether machines (AI) can surpass humans. Behind the fanatical emotions, there is even the artificial intelligence threat theory.
Ray Kurzweil, technical director of Google and author of Singularity Approaching, predicted that artificial intelligence would surpass human intelligence. He wrote in the book, "Due to the exponential growth of technological development, machines can simulate the neocortex of the brain.
By 2029, machines will reach the level of human intelligence; By 2045, man and machine will be deeply integrated, which will mark the arrival of the singularity moment. In addition, representatives who support the artificial intelligence threat theory include the famous physicist Hawking, Microsoft founder Bill Gates and Tesla CEO Musk.
20 14 12, Hawking said in an interview with BBC that using artificial intelligence technology to make machines that can think independently will threaten the survival of mankind. Hawking said: "It moves by itself and can redesign itself at an unprecedented speed. Humans, limited by slow biological evolution, are not competitive at all and will be surpassed. "
Tesla CEO Musk's attitude towards artificial intelligence is extreme. 20/KLOC-In August, 2004, when he recommended Nick Postrom's book "Super Intelligence: Roadmap, Danger and Coping Strategies" on Twitter, he wrote: "We need to pay attention to artificial intelligence, which is more dangerous than nuclear weapons."
20 17, 17 In June, Masayoshi Son, a famous Japanese venture capitalist, said at the 20 17 Mobile World Congress that he thought robots would become smarter than humans, and in about 30 years, the IQ of AI would be expected to exceed 10000. In contrast, the average IQ of human beings is 100, and genius may reach 200.
Sun Zhengyi said: "Singularity is the moment when the human brain will be surpassed, and it is a critical point and intersection. Artificial intelligence and computer intelligence will surpass the human brain, which will definitely happen in 2 1 century. What I want to say is that there is no need for more arguments and no need for more doubts. "
With the increasing popularity of artificial intelligence threat theory, scientists in the field of artificial intelligence have raised objections to it. 20 14 April, Yang Likun, director of Facebook artificial intelligence laboratory and professor of computer science at new york University, expressed his views on the threat theory of artificial intelligence in an interview with Pop magazine.
He believes that researchers of artificial intelligence have underestimated the difficulty of making intelligent machines for a long time. Every new wave of artificial intelligence will go through such a stage, from blindly optimistic to irrational, and finally to depression.
Yang Likun pointed out: Many people think that the progress of artificial intelligence is an exponential curve, but it is actually an S-shaped curve. At first, the S-shaped curve is very similar to the exponential curve, but due to developmental damping and friction factors, the S-shaped curve will infinitely approach rather than surpass the human IQ curve to some extent.
Futurists assume that these factors do not exist. They are naturally willing to make blind predictions, especially when they are particularly eager for this prediction to come true, which may be to realize their personal ambitions.