Judging from the existing technology, it is still very difficult for a conversation robot like Siri to pretend to be a real person and talk to people for three to five minutes without being seen through. Take Siri for example. After receiving a sentence from you, the mobile phone will send it back to an analysis system consisting of thousands of computers. By searching the vast amount of information and databases on the internet in a short time, it will "try its best" to calculate a reasonable answer and send it back to your mobile phone. Although its designers use various advanced technologies, Siri often answers irrelevant questions and leads to jokes again and again. Siri-like voice chat robots and some automatic chat programs are mostly unsatisfactory, indicating that their "brains" are not "smart" enough. The famous "Turing Test" measures computer intelligence. The judge stands on the right side of the partition, the computer is on the left side of the partition, and a real person is on the left side of the partition. Then the judge communicates with computers and real people in various ways through partitions. If you can't tell whether he is communicating with the computer or with a real person, it means that the computer has passed the Turing test. If Siri and those automatic chat bots take the Turing test, they will all fail.

Although the current computer operation speed has reached billions of times per second, it is an instant thing for a computer to work out the arithmetic problem that a person can only work out with a pen and paper for an hour, it seems very stupid to talk to people with a computer with a fast brain. Computers have strong computing power, but their thinking ability and intelligence level are far behind the human brain. According to the idea of "Turing Test", from 65438 to 0990, there will be a World Lobna Prize for Artificial Intelligence every year. Top computer scientists from all over the world who study automatic chat robots get together to see if their computer programs can pass the Turing test. In the competition, these computers don't need to sound like Siri. As long as they type and chat, the task is less difficult. However, in the 20 years of competition, there has not been an automatic chat robot that can completely scare the judges. As long as the judges chat with the computer for a short time, they will immediately find that it is not a real person, but a computer program. Suzette, the best champion in the last two years, can only cheat the judges on the other side if the real person on the left side of the partition deliberately pretends to be a robot. According to the provisions of the Lobna Prize, the champion of the annual competition can only win the bronze prize; If you can fool all the judges so that they can't tell whether it's a robot or a real person, you can win a silver prize; If you can "cheat people successfully" not only by typing, but also by voice, you can win the gold medal.