We know that in the past when designing machine intelligence systems, technicians always tried to make them simulate and restore human behavior. So we saw that in the process of replicating human beings, the machine learned how to watch, how to play chess, and even controlled the limbs to perform fancy gymnastics like a human. It seems that the capabilities of machines have advanced by leaps and bounds, but why are there so many arrogant robots in novels and movies, but none in life? This is because of the lack of self-awareness of the robot. It can only accomplish what humans tell it, and it’s also very strenuous. As for the self-simulation and self-modeling required by a strong intelligent system, it is a faraway fantasy. But now, the emergence of bionic swarm particle robots seems to have given a new answer to this question.
Unlike human-shaped intelligent robots with mechanical bodies and super-intelligent brains in people’s imagination, the machine consciousness mentioned by Hod Lipson is probably the first to be born from a cluster of particle robots.
In March 2019, the emerging particle robot invented by the Hod Lipson team appeared on the cover of “Nature” magazine, which loosened the “machine’s self-awareness”. This discovery also caused an uproar in the academic world. Its innovation mainly comes from two aspects: First, these mechanical systems composed of “particles” (that is, disc-shaped units) do not require strong centralized control by humans, and can also rely on autonomous expansion and random movement to complete complex tasks; The second is that this kind of system is likely to generate awareness, and then make self-simulation and self-replication through thinking “why am I here”, so that it can successfully complete tasks in equipment failures and special operating environments.
A small step that simulates the human mind, a big step for robots
We know that artificial intelligence research is based on how the human brain works, and consciousness is a very advanced part of it. The progress of the particle robot cluster shows which stage of robotics has come to?
First of all, before worrying about machines threatening human society, we will first taste the sweet moments brought about by technological progress. The reason for saying this is because the previous discoveries are pushing the mental capacity that machines can simulate to a higher dimension. According to current scientific research, the human mind mainly includes four major abilities, of which visual perception and reasoning abilities have been copied to the real world by neural network models. Computer vision and AI brains are becoming new productivity engines. Emotions and social relationships, for example, human beings in different cultural circles can feel the connected feelings of love, hatred, sadness and joy, and can generate motivation and predict behavior based on different social relationships such as cooperation, competition, friendship, and family.
The latter two abilities are the advanced intelligence that robots and scientific researchers cannot find. What would happen if the machine had them? At least the future within the eye can be very exciting.
Take particle robots as an example. This kind of very rudimentary “most likely to have consciousness” machine system can bring tremendous social value.
On the one hand, if robots can simulate other humans and robots, they will also initiate social behaviors between robots, and then they can recognize the environment, predict the future and take active actions.
If a part of the machine is broken, it can be repaired by itself, which greatly improves the utilization efficiency and scope of the machine. For example, the machine in the factory, the vacuum cleaner at home, etc., if some parts are broken, the whole machine will not work. In the future, machines composed of these nano-scale robots will be able to circulate continuously, and they can still be used after a single one is broken, helping users continue to save costs.
On the other hand, it can open up new imagination and possibilities for future technology. For example, space technology, biomedicine, etc. With the evolution of particle robotics technology to millimeters and nanometers, it will be applied to human blood vessels to help deal with lesions in the future, or robots on the moon can change their shape according to the needs of tasks. Any automated intelligent system, including unmanned driving that is struggling to find a way out of technical bottlenecks, can improve faster once they are self-aware and recover from damage. This is not far from the future of machine intelligence.
Who has the final say on the good and evil of the machine?
Next, let’s talk about the second question-if the machine is conscious, will “he” be good or evil?
Before answering this question, we might as well talk about the progress of computers. As we all know, today’s computers are smarter and more powerful than those in the last century, but there are fewer and fewer worries about the mainframe rebelling against humans in science fiction movies like “2001: A Space Odyssey”. Today, we may worry about the chaos caused by network vulnerabilities or maliciously created viruses, but we rarely imagine computer killers or secret organizations.
Why are humans so nervous about intelligent robots? In addition to the crushing of individual humans in terms of intelligence, ability, and physical strength, I am afraid that a big fear comes from unknown and uncertainty: once a machine has independent personality and cognition, will it harm humans?
The question is, why do robots want to do this? For more server resources, or hurting assembly line workers will make them happy? Will the machine go bad on its own without sufficient “reward”? Obviously, we have to ask humans themselves.
Many people believe that the robot awakening will be like the heroine Delores in “Westworld”, realizing that he is enslaved and played by humans, and then begin to revenge. Even if this day does come, I think the current relationship between humans and robots will not become tense.
Which “perverted scientist” would be willing to risk the world’s misfortune by throwing an “anti-human” data set such as an intelligent machine, without worrying about being abused by the machine first?
Furthermore, if such human “terrorists” appear, such as using AI to deceive and commit crimes, and transform robots into explosive devices, can they be detected and contained in time technically?
For this, intelligent science is learning from the network security community and introducing some mature prevention mechanisms, such as red teams in artificial intelligence development, formal verification in artificial intelligence systems, and responsible disclosure of artificial intelligence vulnerabilities.
In 2019, the “Malicious Use of Artificial Intelligence” report jointly released by 14 institutions and universities including OpenAI, Oxford University, Cambridge University, etc. also further called for the formation of a normative and ethical framework in the field of artificial intelligence. With such efforts in ethics, technology, and law, will the autonomous consciousness of robots still bring danger to humans? Strict, such as Hod Lipson, just answered—not that simple.
“It should be said that there is no simple answer to any such question, but we must embark on such a process and move forward step by step. We must now openly discuss what it can and cannot do. Keep researching.”
The worries about technology are like the fear of touching fire for the first time, hanging in human hearts forever. Today, most people talk about AI. They have no longer had the fear of seeing Deep Blue and Alpha Dog for the first time, and they have begun to focus on how it can release energy for the economy.
From this perspective, the “golden age” of intelligent robots has not yet arrived. Like all “predecessors”, they are experiencing fantasy, questioning, and chaos. Therefore, at this very preliminary stage of technology, it seems to be a metaphysical thing to discuss whether it is good or evil after awakening, but it is related to the future of each of us, and it is also bred in such thinking and connection with technological progress. Between torture.