Original High Tech & Industrialization High-Tech & Industrialization & Industrialization" can be subscribed
Ren Fuji
The emergence of big models is a catalyst for the development of the metaverse
The connotation of digitalization has gradually evolved from traditional informatization to intelligent, intelligent and digital civilization. The metaverse is the initiator of digital civilization, a network world born of virtual and reality, and defines it as the future that has come, which means that it is too advanced and "premature babies". However, the emergence of the meta-universe is in line with the trend of the times, and we can see some laws from the virtual world, the digital world, the consciousness world, the programmable world, and the inspectable world.
The metaverse and advanced intelligence could be at the heart of shaping the digital world. The eight emerging technology trends in the digital world predicted by Forbes in 2023 are all closely related to the metaverse, robots, and advanced intelligence, including the ubiquity of artificial intelligence, more humanized robots, the increasing popularity of gene editing, and the increasing popularity of blockchain technology.
This large model does not have much innovation in theory, but it does cause a sensation, and even detonates a vigorous artificial intelligence revolution. This lies in the "emergence" of a phenomenon that was not originally paid attention to, and the emergence ability of large models catalyzes the metaverse. We were originally pessimistic about the meta-universe, thinking that it had too many shortcomings, but the emergence of large models may accelerate the landing of the meta-universe in a certain sense.
Why can the emergence of big models catalyze the metaverse? The so-called "emergence" is simply something that is small-scale, invisible to individuals, and when deposited into large-scale, it will surpass human intelligence. Recent big models include ChatGPT, TPT4, and newer GPT families, which, while seemingly uninnovative in theory, are able to "turn the world upside down" and bring profound changes because of the emerging capabilities. In a sense, this is evolution that is not subject to human will. But the intelligence brought by the big model includes the emotions to be promoted next, and humans are completely different ways, and its emotional mechanism and intelligence mechanism are another channel.
If ChatGPT4 has interactive functions, the following six points are worth paying attention to: first, powerful natural language processing capabilities; Second, efficient text generation capabilities; Third, rich knowledge base and ability to provide world knowledge; Fourth, greater creativity and adaptability; Fifth, cross-language multimodal interaction capabilities; Sixth, a strong ability to learn and evolve itself.
The superworld model will be the ultimate destination of the metaverse
Why is the superworld model the ultimate destination of the metaverse? In the superworld model, not only can you be immersed and see his person, but he can also be immersed and like a person, and the robot will come to his aid. Human beings live in time, space, and human beings described by three variables, and the human world is divided into people and objects. In the superworld, it is no longer simply experiencing the binary world, but from positive time to negative time, from real space to virtual time, real thing to virtual thing, from natural person to virtual person. Traveling back and forth in the twenty-four dimensional world, we call it the "superworld".
Augmented reality in the hyper-world model is positive in time, space is real, and the characters in the world are also real people, but the physical object is virtual. What was the concept of virtual reality in the past? Time is negative. The content of the original science fiction novel can now be partially realized, such as reuniting with deceased loved ones, talking to historical characters, and turning physical objects into virtual objects. We predict the future of virtual reality, which will become positive in time and virtual in space.
Specifically, the metaverse is mainly about depiction, the physical world can be digitized, and the digital world can be realized, which is to create a fusion of virtual and reality. The big model discussed recently is actually an injection of intelligence empowerment, which promotes virtual systems to perceive and understand the physical world and promote virtual and real interaction. The superworld is a blend and symbiosis, shuttling and extending infinitely in the four-dimensional space of virtual and real to create a blend of virtual and real, which is the superworld model we proposed.
We have a general overview of the metaverse, and the current metaverse is the future that has come, which is moving towards the online world of virtual and real life. The superworld is the real world where the virtual and the real meet, and the main key factor is advanced intelligence.
Advanced intelligence is a research paradigm proposed by the Dartmouth Conference in 1956, the natural intelligence of hundreds of millions of years of biological evolution, and the fusion of affective computing in less than 30 years. It can be responsible for connecting the real world and the metaverse, giving the metaverse the ability to perceive, understand, and reason. Emotional intelligence is responsible for connecting the real world and the last mile of the metaverse, providing a natural and harmonious immersive experience and delivery.
THE TECHNICAL SYSTEM OF THE METAVERSE WAS ORIGINALLY CALLED THE "BIG ANT" BIGANT. This system is considered to be unable to support the "stable and far-reaching" of the metaverse, and there will inevitably be a bubble. The reason for this is that it equates artificial intelligence with several other supporting technologies. Artificial intelligence is the brain of the meta-universe, and in the new meta-universe technology system, artificial intelligence is put on the mind of the big ant, and artificial emotions and emotional computing are added to the original six plates. Why add affective computing and artificial emotion? Because as long as there is interaction, as long as it serves humanity, it is difficult to move without emotions.
The era of the "three" dance
What is a natural person? All of you here are natural people, real human beings in real space. Robots mainly refer to high-simulation robots, that is, robots made by imitating people, which not only simulate in appearance, but also have basic human behavior characteristics such as perception, decision-making and execution under computer-enabled technology. The definition of digital human is the aboriginal of the metaverse, and the virtual objects with digital appearance and existing in the non-physical world created only through computer graphics rendering, motion capture, deep learning, speech synthesis and other technologies can replicate human knowledge, emotions, memory, and thinking to a certain extent.
Natural people, robots, and digital humans "three people" on the same stage, from the perspective of human development process, there are very ancient legends. For example, the guide car, the buffalo streaming horse, the da Vinci surgical robot, and the emotional robots "Think", "Sophia" and so on. From natural person labor to robots with digital human partners, laying the foundation for the future "three people" on the same stage.
Why do digital humans, robots, and natural people dance together? According to the "Metaverse Development Research Report Version 3.0" released by Tsinghua University, the core industry scale of virtual digital humans will be about 33.6 billion yuan in 2021 and 99.8 billion yuan by 2025. These figures are somewhat different from what we have learned recently, but they are indeed increasing in line with this trend.
There is also a view that in the future, when natural persons, robots, and digital humans work together, the proportion of natural people's labor will only be 5% - we think that after 2049, this proportion will be even lower, with robots accounting for 15% and digital humans accounting for 80%. In addition, the cost of hardware such as robots and cars in the future cannot exceed 10% of the total cost.
What are the application scenarios where "threesome" dancing will appear? For example, in the field of health care, based on the emotional interaction ability of robots and digital human brains, it can accompany and care for the elderly, and will play a positive role in emotional management and psychotherapy. Our survey shows that 85% of seniors believe that loneliness is the biggest problem, and "three" on the same stage can alleviate this problem. In addition, including education, cultural tourism and health, etc., the "three-person" dance has a wide range of application scenarios, natural people, robots, digital people dance through thousands of industries, and even into thousands of households.
Emotional interaction is the "Mountain of Kings" of the "three-person" dance technique
Emotions are unique to human beings, and emotional computing is a worldwide problem. Originally thought that emotional research belongs to the liberal arts and cognitive sciences, in fact, emotional computing has a strong demand, including pension, companionship, child care, health care, science and education culture, national defense and military, smart communities, etc.
What should we do in the face of such a worldwide problem? This brings us back to artificial intelligence, which is essentially a roundabout of the essence of artificial intelligence. So for the next 30 years and 60 years, it is based on the paradigm of understanding of the brain, mind, and consciousness. Does the recent large model have a different intelligent mechanism and emotional mechanism than humans? It's all worth looking into. If we don't study this aspect of emotion, it is possible to have unemotionally resonant, low-immersion experiences, including weak human-machine systems.
In recent years, our research on affective computing mainly puts forward the theory of emotional interaction, including fusion perception, system reasoning and emotional interaction, which makes emotions transition from single-modal, static, passive, virtual, and non-evolutionary to multi-modal, dynamic, active, real, and evolvable trends. We create a "three-person" dance based on the brain-like multimodal perception and affective computing platform, including real people, media interaction, digital people, and of course, specific robots. Why has machine entities always been valued? Because we always believe that whether it is a big model, a metaverse or a superworld, the ultimate starting point is to stipulate that it serves humans, so we should study physical robots.
Specifically, to cross the emotional gap of "three-person" interaction, it is necessary to promote technological innovation in emotional perception, virtual and real integration, and emotional interaction, and create a natural "three-person" emotional interaction environment - natural interaction is emotional interaction.
In the future, in the global and even larger scale, natural people will also need robots. When the "three" are on the same stage, who pays the robot? After the advent of digital humans, who positioned it? What is its identity and what is the "machine power"? People have human rights, then robots also have "machine rights". After the emergence of "machine power", what will happen to the situation of "three people" dancing together? This includes a series of problems, but none of them have developed to the point where they need to be considered, and there are still many topics to be done to truly realize the "three-person" dance, and the main topic at present is emotional interaction. We always believe that if digital humans produce intelligence, they will be different from humans, and in the process of interaction, they must understand and assist each other, which is the expectation of the future "three people" on the same stage.
(According to the main forum report of the "2022 Chinese Industry Intelligent Industry Annual Conference", the content has been deleted)
