Third World Intelligence Conference held as scheduled in Tianjin, around the conference: theme "Smart new era of progress, Strategies and Opportunities", there will be more than 1200 international guests from England, America, France, Germany, Switzerland, 30 countries and regions rushed to the scene, to discuss "intelligent transportation", "intelligent manufacturing", "smart city", "intelligent life", "big data" and "research and innovation" and other issues.
Compared to the topic on these tall, 360 president Zhou Hongyi has proposed a relatively sharp question: "whether the development of artificial intelligence to replace human beings, if replace humans, some of the work appears irreversible damage who is responsible if these techniques are criminals. master, there is no good way to directly control the terminal, as well as artificial intelligence law has not been perfect, requires the introduction of relevant laws to respond to changes as soon as possible. "
Zhou Hongyi talk about the three events to substantiate these ideas:
Zhou Hongyi, the first thing to talk about is the Boeing 737MAX crash recent events. "Although I am not aviation experts, but I have seen from two problems." Zhou Hongyi, analysts say, first, and then design a perfect system, once the sensor error, it will affect the judgment flight control system, and then "make mistakes." Second, when a computer error, what is to listen to the computer or listen to the driver? This is a question worth considering, in the future no car, no factories, no docks, and many other unmanned systems, under the false sensor data is contaminated, the final control and decision should be left to machines or people?
The second thing is Zhou Hongyi talk about "fugitive killer Jacky Cheung." It is understood that, because Jacky Cheung concert scene recognition technology, police arrested more than the presence of fugitives. *** But they found that as long as a strange stripes print down to clothes, you can "mask" off the face recognition system, free access. This will probably be the "best-selling clothing" era of artificial intelligence.
Zhou Hongyi talk about the third thing is the use of artificial intelligence to the United States was bad video where the protagonist into a star. This led to his thinking: "This technology is causing panic, the US government seized this technology was also made circle of friends I no longer believe in the beauty of this tool reduces false if cost per person, you. false news, videos pseudonym person may appear. this subversion of our past 'Seeing is believing' is inherent in the concept, and even cause confusion at a critical moment. "
"I say these three things is not to belittle artificial intelligence, on the contrary, concerned about the safety of artificial intelligence in favor of better development. Recently there are a lot of people mention the topic of telephone robot telephone robot is more of a problem really easy to use, although intelligent voice market we did not have a particularly large number of covered, but the same token, good AI must be to the service of mankind, "said Zhou.
Zhou said the same token, some of the driverless car might be some specific symbols to mislead, lead to serious accidents. He suggested that enterprises in the on-line unmanned systems, safety testing should be prepared to go to find the problem by actual ***.
"Singapore is developing appropriate legislation." Zhou called on China should also strengthen the importance of this kind of problem, to prevent such a "tool" spread, triggering a series of ethical questions.