This is not an alarmist. In fact, the concept of artificial intelligence is very broad, so artificial intelligence can be roughly divided into three types:
First, weak artificial intelligence:? Weak artificial intelligence is artificial intelligence that is good at one aspect. For example, artificial intelligence can beat the world champion of chess, but it can only play chess.
Second, strong artificial intelligence: human-level artificial intelligence. Strong artificial intelligence refers to artificial intelligence that can compete with human beings in all aspects. It can do all the mental work that human beings can do.
Third, super artificial intelligence: Nick Bostrom, a philosopher and famous artificial intelligence thinker at Oxford University, defines super intelligence as "in almost all fields, including scientific innovation, general knowledge and social skills, it is much smarter than the smartest human brain." Super artificial intelligence can be a little stronger than human beings in all aspects, or it can be a trillion times stronger than human beings in all aspects.
At such a point, we can easily see that the current artificial intelligence is still in the stage of "mental retardation".
Back to the question, we will notice that the key point is: new version, scrolling. Another thing about this news is that I taught myself for three days. mention
Here, I want to introduce a very heavy concept: recursive self-improvement.
The concept is this: an artificial intelligence running at a specific level of intelligence, such as brain-dead human level, has a self-improvement mechanism. When it completes a self-improvement, it is smarter than before. We assume that we have reached Einstein's level. At this time, it continued to improve itself, but now it has Einstein's intelligence, so this promotion will be easier and more effective than the last one. The second improvement made him much smarter than Einstein, which also made his subsequent improvements more obvious. Repeatedly, the intelligence level of this strong artificial intelligence grows faster and faster until it reaches the level of super artificial intelligence-this is the intelligent explosion and the ultimate expression of the law of accelerated regression.
Imagine: it takes decades for an artificial intelligence system to reach the level of human brain-dead intelligence, and when this node happens, the computer's perception of the world is probably the same as that of a four-year-old child; One hour after this node, the computer immediately deduced the physical theory of unified general relativity and quantum mechanics; An hour and a half later, this strong artificial intelligence became super artificial intelligence, and its intelligence reached 654.38+0.7 million times that of ordinary human beings.
When a super artificial intelligence was born, it seemed to us that an almighty God came to the earth. ?