Chapter 128 Artificial Intelligence In Xinghai's space hub laboratory, currently only Fanren himself has an independent office. Although the construction costs of current space projects have been greatly reduced due to the operation of space elevators, the costs are still too high compared to construction on the Blue Planet. In the office, outside the window is the space port in the distance. Several Dragons and other types of spacecraft are docked at the port. After Ah Dai came in, he didn't feel strange. He walked to the window and looked at the huge and ugly Elephant in the distance. Even with his character of not caring about appearance, he couldn't help but twitch his lips at this moment. It’s really hard to describe the boss’s aesthetic sense. He can design a domineering-looking young dragon-class machine, but he can also design this kind of industrial monster with no aesthetic appeal without fear of being ridiculed. Fan Ren looked at him sighing in front of the window and knew what he was thinking. Since the Colossus came out, it has been criticized by many people because of its appearance. But he didn't care. No matter what, just ask the workers in charge of loading and unloading at the port whether it is inconvenient to load and unload a cargo ship with a layout like the Colossus.
The answer is absolutely. Although there are many inconveniences in life in space due to gravity, there is indeed one advantage. There is no need to place chairs in the room like on the ground. Although Fanren's office also enjoys the influence of rotating centrifugal force, the slight gravity can only allow them to keep their feet slightly close to the ground, and their legs will not bear too much pressure at all. There is no need to sit down at this time, just stand there lightly. "Okay, tell me what you think." There is a reason why Ah Dai has been absent-minded recently. I don't know whether this guy is gifted or it's the buff given by his bald head, but with the emergence of graphene chips in recent years, his technology in this area has advanced by leaps and bounds, and even Fan Ren, who has the help of the system, can't understand his depth. As the chief scientist in this field of Xinghai, A-Dai has the highest authority to use Xinghai's supercomputers. Relying on this authority, he has been immersed in the field of artificial intelligence in recent years. Not long ago, he ran up to Fan Ren mysteriously and said something that surprised him so much that his teeth almost dropped. "Boss, it seems our supercomputer has developed some initial self-awareness." Fan Ren remembers very clearly that his first reaction at that time was to immediately cut off all physical connections between the Xinghai supercomputer and external devices. He knew what this meant. There is no doubt that the Xinghai supercomputer is the most advanced supercomputer in the world today. Its computing power is thousands of times greater than that of the world's top supercomputer ever. A large part of the reason for this is the ultra-low power consumption and very low heat generation of graphene chips. Through the parallel architecture developed by A-Dai, it is theoretically possible to connect an unlimited number of chips in series. It is precisely for this reason that, with the help of this huge computing power, the technology in all walks of life in Daxia has developed extremely rapidly. The mass production of nanorobots is also one of the achievements of Xinghai supercomputing. If not, the research and development of Fanren's nanocoat would not be supported by the nanorobots produced in the laboratory.
After Ah Dai's explanation, Fan Ren realized that the situation was not serious enough for him to take physical isolation. What Ah Dai said about having a certain degree of self-awareness was just that under certain specific circumstances, the artificial intelligence model he designed would upgrade its code by itself. Of course, this specific condition was extremely harsh. The reason is that when Ah Dai first set up the model, he set up a string of codes. The meaning of this is that the AI model can later obtain permissions through this string of codes to upgrade its own code base. But this code was banned by Fanren at the very beginning. Ah Dai did not refute at that time. He still listened to Fan Ren's words and deleted the code at that time. However, the AI model had already entered the trial operation stage at that time. Even though this string of code was deleted from the underlying database, it was still learned by the AI. Of course, this does not affect usage. However, some time ago, when A-Dai was using a supercomputer to solve computing problems, he was surprised to find that the AI artificial intelligence that was doing deep learning in the supercomputer actually took the initiative to apply to re-add the original code that was deleted by A-Dai to the database so that it could better upgrade itself. This request was rejected by Ah Dai at that time. Then he was told to come to Fanren, and that's when Fanren was so scared that he planned to physically isolate the supercomputer. Fanren has been "taught" IT-related knowledge by the system. Although he is not as gifted as Ah Dai in this area, his understanding of related technologies is not low at all. He has his own views on the development of AI. In his opinion, with the development of science and technology, because the human brain's ability to process information is always limited, more and more information processing work must be done by computers. With the birth of artificial intelligence models, they have even developed to the point where they can complete many higher-authority command operations under pre-set rules and conditions. But no matter what, these artificial intelligence products are fundamentally tools used by humans. Tools do not have self-awareness, and all their "thoughts" are completed within the rules and regulations set by humans in advance. Once self-aware artificial intelligence enters human society, relying on the computing power of computers that is countless times greater than that of humans, artificial intelligence can even make plans a long time in advance to prepare for the interests it needs in the future. And it is very likely that humans know nothing about this. The risks involved are difficult to estimate. (Note 1) As time goes by, smart products will enter thousands of households, from small switches to large public transportation vehicles such as cars and airplanes. If in this process, because of this code, they develop self-awareness, then countless fields in the future may be threatened by artificial intelligence. Moreover, given the other party's extremely powerful computing power layout, it is very likely that people will know nothing about this. (Note: 1) Once it has self-awareness, it will definitely have its own interests. So who can guarantee that artificial intelligence will always choose to make sacrifices in the face of its own interests and thus safeguard the interests of mankind? After all, it has self-awareness. Self-awareness has always been a prerequisite to override all external factors. For example, a scholar once introduced it as follows: How can you tell whether the world you live in is real or a dream? And even after waking up from a dream, is the world you live in still the real world or still in a dream? The scholar's answer is: In this process, as long as you keep questioning whether the current world is true or false, then this question itself is the most authentic self-awareness. It is for this reason that perhaps humans can override AI models and command AI hundreds of millions of times to make choices that are in line with human values and interests. But as long as it has self-awareness, it is absolutely possible for it to have ideas about its own interests and put them into practice. Note 1: At the current stage of AI development, there is a saying that large models can be used to learn various meteorological data, so as to carry out meteorological intervention.
The concept is similar to the butterfly effect. (End of this chapter)