Musk, the biggest skeptic of artificial intelligence, co-authored 116 open experts in the field of artificial intelligence and robotics in an open letter calling for a UN action to ban the 'killer robot.' The letter said: 'Once this Pandora's box has been opened It's hard to shut it down again. "
Musk also publicly said on Twitter: 'If you do not want to be concerned about the safety of AI, you have to do it from now on.'
Last week, physicist Hawking once again warned: 'Artificial intelligence may be the worst event of our time and its rise will be the end of human civilization.'
In this regard, the most important thing is to personally feel the power of artificial intelligence powerful human go champion Ke Jie. Recently Ke Ke told First Financial Times interview, said he has been convinced that no longer can not win AlphaGo. 'Indeed, when it ( AlphaGo) Just over Lee Seokshi the moment, I had a broken faith in Go.
Awe from chess king
In mid-October, the release of AlphaGoZero's 'No Teacher's Self Help' report, a new version of AlphaGo, detonated again the question of whether AI should or may not be aware of. If future AI could create self-awareness, would it not Does it mean that it will not be under the control of humans or vice versa?
Because experienced, Ke Jie helpless artificial intelligence, but also awe.
At the forum on artificial intelligence to help smart finance and 'knowledge map' held in Wuzhen, Ke Jie told CBN reporter: 'The Go competition has always been to worship the strong ones, and only people like us before will think of. But now you discover that you are going to worship, a machine that makes you feel horrified. Playing chess with him is so painful that I was in tears at the time.
There's no point in re-playing the machine, we're completely out of the same dimension, we used to have some hope of finding some bugs, but now AlphaGo comes out with a new version (AlphaGoZero) and we're sure we've won Without it, we have a brain, searching for neural networks without it, and without it. "
However, in the United States University of California at Berkeley computer professor Wang Qiang, AlphaGo and not so powerful.Qiang Wang told the first financial reporter: 'The development of artificial intelligence to go through a process: the first step is the perception of the machine; the second step is to recognize the machine The third step, the interaction of the machine and the addition of emotion, the fourth step, the complete perception. Now artificial intelligence is still in the second step.
Professor Wang Qiang further explained that the development trend of AlphaGoZero should draw people's attention. 'Ke Ke chess, will find out five points from the board, which is most likely to win an analysis, but when Ke Jie counted 20 steps AlphaGo has been able to reach 1,000 steps in depth, and it can play chess with itself, and logical thinking can not be done by humans.
Wang Qiang thinks that this aspect is a good thing. 'The deeper the network speed is, the stronger the computing power is. In the future, it will bring more knowledge search capabilities and computing power in face recognition and knowledge mapping.'
He also cited the hot data analytics trader and SmartBuyer as an example, arguing that the hardest part to analyze is not portfolio analysis, but rather the investor's risk appetite and how to align the algorithms to one Platform. "Including Baidu Apollo Project, Microsoft Research Institute, they have done some platform, the latest algorithms into the platform.This requires rational and a large number of logical operations, the possibility of substitution of human beings.
As Kejie said, in this era, there are unfortunate, but also lucky. 'Unfortunately, because it can not win the frustration, fortunate because of its power, so that we can see, we can even feel in advance of the confrontation between man and machine '
Ke Jie told CBN reporter: 'In the future, 90% of the industries will be hit by artificial intelligence and face unemployment and transformation. In the future, when humans encounter machine confrontation, they can look at the current situation in the Go community.'
Human brain into the 'artificial intelligence brain'
However, the most horrible thing is not robots will replace humans, but robots may rule humans .This worry is not non-existent.Scientists believe that the rapid development of artificial intelligence and machine learning is likely to be born a ' New human beings like God. "But humans can also resist the robots of the future by incorporating new layers of artificial intelligence in their brains.
Musk mentions an 'AI-humansymbiote.' In his view, it is best that humans and machines co-exist with a 'WearetheAI' way of life. At Under such a scenario, everyone is 'artificial intelligence', there is not an evil 'artificial dictator', but everyone who wants to participate in artificial intelligence becomes a 'human-machine symbiosis'.
Musk compares this situation with the cooperation of the limbic system and the brain-center system, which is the brain tissue evolved from the ancient cortex in the central nervous system. The limbic system is involved in the regulation of instinct and emotional behavior. The hippocampus Structure also plays an important role in learning and memory.
He argues that the limbic system and the central system control the 'instincts' and 'thoughts' of human beings, respectively, and that by incorporating artificial intelligence, the neural connections between the central nervous system and the human brain to digitally expand external devices can be improved. As a result, digitized The problem of bandwidth does not exist, so humans can become a more effective 'human-computer symbiosis' and, because of the popularity of this technology, there is no question of who controls who.
According to Musk's imagination, the human brain can become more powerful by implanting a microchip in the human brain that can upload and download information. His plan is that in the next four years, a new type of neural net Brain layers that help people treat severe brain damage, and the Matrix technology, described in The Matrix for the next eight years, can be used by all.
Chen Tianshi, the founder of the leading artificial intelligence chip company in China, told CBN reporter: 'The rapid development of emerging information technologies represented by artificial intelligence is defined as' brain-robot or' brain-computer interface New human being, like a mixture of machines and machines. "
However, according to Chen Yun-ting, a researcher at the Institute of Computing Technology under the Chinese Academy of Sciences, it is still difficult for Musk's extrinsic equipment to "digitalize the human brain." He previously told CBN reporter: 'This It is technically very difficult, and now humans can only copy some of the more superficial information in the brain. "But looking forward to the development of the brain-machine interface in the next decade, he said: 'It is possible that many people are now doing this Try
Ke Jie's revenge
Musk, who has always advocated "artificial intelligence threat theory," believes that the Brain-Machine Interface (BMI) he created will be able to counter the so-called 'God of the future' New human 'best way.
Musk's theory of this technique is that there are many ideas in the human brain that compress to very low frequency data output, that is to say or written, and if you have two brain interfaces, the brain There is no need to compress these ideas so that they can be transmitted directly to others, the so-called 'electrocardiogram'.
For this' electrocardiogram ', Artificial Intelligence, Tom Mitchell, director of machine learning at the Computer Science School of Carnegie Mellon University, told CBN reporter:' Computers and the human brain can think alike Process, the accuracy rate can reach more than 90%, but to teach the machine how to observe the brain, can not do now, Musk's neural network will be a good start.
At present, human scientists can do as soon as possible to build a human brain model.Recently in the first interview with Financial News, Stanford University professor, 2013 Nobel Prize laureate in Chemistry Michael Levitt (MichaelLevitt) on human brain expression Out of great interest.Levitt is working with Fudan University brain-like research team on multi-scale brain model construction and multi-scale data analysis carried out in-depth discussions, and look forward to the future in this area there are important new breakthroughs.
Guo Aike, a researcher at the Chinese Academy of Sciences Institute of Biophysics and a member of the Shanghai Academy of Science's Neuroscience Research Institute, said last month at an internal sharing session of the 'China's Brain Planning Overture': 'The human brain is an evolutionary process that can not be duplicated but completely There is the possibility of integration - there is a striking similarity between the plasticity of the brain and the evolution and choice of nature that applies to all levels of life and intelligent systems.
However, science fiction writer Liu Cixin, in his novel The Time Migration, describes an 'invisible era' of 1,000 years later, regarding the future of the 'man-machine symbiont' envisaged by the above-mentioned scientists.He wrote that at that time human beings and The chip has been integrated, everyone in the invisible world to create whatever you want, or even create a universe with hundreds of millions of galaxies in the invisible world, imagination and reality is one thing, of course, is a quantum chip In fact, humanity did not need it anymore.
Although I did not really want to play chess with artificial intelligence anymore, because it was too painful, I would love to suffer again if everyone likes to see me continue to be abused by AI. "Ke Jie finally joked with reporters.
But for the 'human-computer community', Ke Jie has expectations.He told the First Financial reporter: 'I also have the illusion that if one day the super-chip can make the brain's computing power to explode, I will not hesitate Choose to implant chips, and AlphaGo fight, to revenge.