20210911 I have seen it. Today is the 20th anniversary of the 9/11 incident. I am very touched by clicking "Captain Sully".
In Antonio Damasio's "The Error of Descartes", he pointed out that the dichotomy of sensibility and reason is a wrong concept. People often describe a person as "very rational" because of his calmness, his logical ability, and his strong ability to do math problems. After the development of artificial intelligence, there has always been a bottleneck, and it cannot surpass humans because of its essential differences. For a simple example, a math test paper, a human spends an hour to get 95 points, and a "rational person" spends two days to get 99.9 points, which one is more rational?
Human reason is the dialectical unity of "logical deduction" and on-the-spot response. Gage in "The Error of Descartes" has a steel pipe pierced through his head, making him more "rational". Gage is like a different person. He is no different from ordinary people in some simple test questions, but in life, he is more indifferent to his relatives and friends who are enthusiastic, and he does not know anything about decision-making. The difference between human life and the simulated world is that there are variables in life. Some variables need to be solved through theoretical knowledge [on the premise that time is very sufficient], and most of them need to be through "experience", and these experiences are even internalized to In our muscles and brains, one can quickly choose the best option, even if it may not be the best.
There is an old saying, "It is human nature to seek advantages and avoid disadvantages." Human rationality is based on human nature. Humans are frightened when they see snakes, and fear can lead people to make decisions, such as running away immediately, or staying away from grass to avoid encounters with snakes. The same is true when people make decisions in life or do math problems. A math problem may have multiple steps. As a high-end mathematician, he may complete it in one go. He even feels that he is not using reason. At each step, he felt that this method or formula should be used, and the brain would produce "happy" chemical transmitters, and the problem could be solved quickly.
In the movie, the best solution for the computer simulation is that after the flock of birds hit the engine, the plane immediately goes to the two nearby airports to land, which is safe and feasible. But the actual situation is that the captain needs time to make a decision, and after 35 seconds of consideration time is included in the parameter, the landing plan to the airport becomes unfeasible, and the bridge or the New York building will be encountered on the way. The captain made a decision in only 35 seconds when the aircraft engine was scrapped. He was on the Hudson River, with high-rise buildings on both sides, and 40 years of driving experience prompted him to make a "crash landing on the river". decision making.
Another book, "Antifragile," by Taleb, discusses why "human driving" cannot be completely replaced by "autonomous driving." Let’s start with a case: After the rise of the auto insurance industry, the probability of accidents in cars in the United States is higher. Because the driver will think like this: "The insurance premium is fixed every year. Anyway, if the car has an accident, there will be insurance to cover the cost." The insurance industry itself has a good starting point, so that ordinary people can take risks, but in the end it leads to more many accidents. The reason is the question of attribution of responsibility. If drivers take full responsibility for casualties and vehicle maintenance, I believe the current road will be safer. The same is true in the aircraft industry. If there is an accident, the pilot will shift the responsibility to the problem of "autopilot" and completely get rid of the responsibility.
The work of a pilot is not only as simple as traveling, but behind it is the happiness of more than 100 families. The captain questioned the computer simulation and failed to take into account factors such as "human nature" and "actual situation". An emotionless computer may crash or freeze when encountering a situation that has not been "programmed". I don't mean to completely deny artificial intelligence here, artificial intelligence AlphaGo has beaten Lee Sedol for many years, but it seems like yesterday. But what most people don't know is that AlphaGo is not the strongest existence, and what is even more powerful is that top human chess players fight side by side with AlphaGo. A system built on human nature is a good system [such as pilots], and a system that forcibly corrects human nature [such as using artificial intelligence to supervise the work of workers] is a bad system, I hope you don’t have to work.
View more about Sully reviews
