As early as the 1950s, Alan Turin, the father of computer science and artificial intelligence (AI) predicted that the development of AI would ultimately reach a point called the “singularity” at which machine intellect would surpass that of humans.

With regard to the future of human civilization in the post-singularity world, opinions tend to fall into two camps. One side argues that AI will merely augment human existence while the other side believes that the singularity could be humanity’s extinction event.

In fact, there is another possibility that is more realistic and needs to be pondered. We have no reason to believe that intelligence cannot possibly evolve in non-biological form, but we have reason to think that no other intelligence than that of humans can evolve in a natural way like biological things do. Since machine intelligence could only be made possible by relying on human intelligence and evolve on the basis of it, it is a type of secondary evolution of intelligence.

Machine intelligence was created and facilitated by human intelligence. However, once AI is capable of designing a higher level of intelligence than itself, the evolution of machine intelligence will begin. Secondary evolution based on the evolution of biological intelligence as it is, machines could possibly possess a higher level of intelligence than humans.

No matter what the post-singularity future will be, we have to deeply reflect upon the relationship between intelligent robots and humans, which is fundamentally the relationship between man and technology.

The progress of all humanity increasingly requires us to reflect on the past notions, ideas, theories, thinking patterns and behavioral modes of humans. However, the emergence of super AI prompts humans to ponder their way of existence as a prerequisite of progress. The evolution of AI is essentially the evolution of information. The primary form of the evolution of intelligence mimics the natural form of biological evolution while the senior form of evolution involves upgrading the information carrier. A biological organism is the elementary carrier of intelligence but as intelligence grows, increasingly advanced exploitation of materials and energy will create more favorable conditions for the development of intelligence.

The upgrading of the intelligence carrier may take two forms: One is that if AI and human intelligence develop in parallel, not relying on each other, the organism would be in a disadvantaged position as a carrier. The other is that the organism, as the carrier, keeps being replaced partially, which means that it is replaced progressively and alternatively. The former form would not have the merit that the latter form has, because the evolution of the information carrier, which possesses self-consciousness, occurs as a result of accumulated experiences, not only in a form as the pure intelligence does. For example, the computing capability can develop by itself without an information carrier that possesses consciousness. Therefore, intelligence with self-consciousness cannot finish the carrier replacement process all at once but only do so as the carrier gradually upgrades after accumulated experiences. This progressive model means that AI and human intelligence develop in an interactive way. It is a process of the biological intelligence being machined and the machinery intelligence being humanized.

Either of the aforementioned evolutionary forms is essentially the evolution of the information. Therefore, the development of AI continues the evolution of human intelligence rather than surpassing, replacing or even destroying it. The substance-and-energy based intelligence means that humanity may become obsolete as the carrier upgrades but information-based intelligence develops as an advanced form of humanity. The former takes population in the biological sense as solely an extension of humanity, while the latter regards reproduction in the sense of ideology and consciousness as a more fundamental way that humanity continues itself.

Diversity is the lifeline of ecology in a system in which diversity and dissimilarity runs counter to uniformity. Excessive uniformity is the functional cause of the end of ecology. Though human beings have confronted conflicts of interests caused by cultural differences among members of society, a rational man is fully aware that the diversity of biological genes is not only the most important and fundamental resource that helps the living things adapt to the natural environment but also is the valuable treasure that may bring new possibilities and new inventions for human development. Therefore, the consensus to protect the resource of biological genes in the sense of real super-intelligence tends to be more important than how human beings consider it.

In addition, the ethics of technology should not only provide ethical warning against the threat that AI poses to humans but also ethical support for the further advancement of humanity itself.

 

Wang Tianen is from the Division of Social Sciences at Shanghai University.