Can artificial intelligence become an ＂author＂
Author: Jiang Zai (young writer)
Do bionic people dream of electronic sheep? Is a science fiction novel by American writer Philip Dick. If bionic people can dream, do they dream of sheep and grasslands in the human world or an artificial intelligence world full of Cyberpunk elements? Consciousness, dreams and memories have repeatedly promoted our understanding of the boundary between human beings and other organisms.
In recent years, human beings have benefited from the development of artificial intelligence industry, and also felt new challenges because of its existence. As far as the literary industry is concerned, the potential of artificial intelligence in writing makes us see the possibility that it has self-awareness. Nowadays, artificial intelligence can obtain key information by deeply learning a large amount of data and semantic analysis technology of text, thus generating words with certain logic. However, as far as the current development technology is concerned, artificial intelligence still does not have the ability of original or structural output. Therefore, their creations will not pose too much threat to mankind for the time being.
But why should we oppose the creation of human beings and the creation of artificial intelligence? Why did it become an alternative situation? Although this seems to be a proposition about the future, this kind of problem is not as novel as we thought. The debate about whether the author should disappear has long existed.
This year, a novel created by artificial intelligence, Land of Machine Memories, won the second prize of Jiangsu Youth Science Fiction Works Competition. The judges voted on the novel without knowing it. In other words, under the premise of hiding the author’s identity, artificial intelligence creation has been recognized.
The disappearance of the author can make readers return to the pleasure of reading works. The writer’s life has become less important, the writer is no longer the object of attention of critics and readers, and language has become the only existence. Many writers are trying to become another person and escape from the only "I".
One way to accept the existence of artificial intelligence literature is that perhaps we can not talk about "who is the author?" And talk about "the function of the author." Why can’t artificial intelligence be an author? It satisfies the condition of countless "I", which is the combination of countless consciousness and countless data. Literature created by artificial intelligence represents the product of the era when artificial intelligence and human beings coexist. Must creation be related to self-awareness? The answer is probably no. There are many "works" in nature, which are the creation of nature’s "unconscious". So even if we can’t capture the consciousness of artificial intelligence, or assert that artificial intelligence is unconscious, it doesn’t prove that there is no concept of "creation" in artificial intelligence. In addition, I don’t think any author can claim that his creation is not influenced by his own subconscious or even unconscious. Therefore, from the level of consciousness, we cannot completely deny the possibility of the existence of artificial intelligence literature. Even, I think it is possible for artificial intelligence to have self-awareness in some ways. The embryonic form of its self-consciousness can be found in its "illusion" phenomenon.
The term "illusion" of artificial intelligence is borrowed from human psychology. This term describes the situation that an artificial intelligence system produces a response that seems to be unsupported by its training data. Also known as artificial illusion, fiction and delusion. Isn’t this the germination of self-awareness? They began to dream, and even Jeffrey Hinton, a Turing Prize winner with the reputation of "Nobel Prize in Computer Science", explained the logic that machine intelligence really dreams of electronic sheep. With the development of technology, emotion, love and consciousness can be reconstructed through neural network and acquired through deep learning.
If we accept that artificial intelligence can be an author, the literature it creates can be collectively called artificial intelligence literature. Then it can coexist with human literature and promote the new possibility of literary development.
However, new problems have emerged. For example, about the nature of literature. Does the essence of literature exist only for language and text? Or does writing exist for some higher and more mysterious value? Between the two, I prefer the latter, that is, writing is to make something fleeting immortal. The narrator of Arabian Nights makes death become distant and vague in one story after another, and the time of the story delays the time of death in the real world. The Peony Pavilion prayed that the living could die and the dead could live. So, what is the creation of artificial intelligence for? In order to meet the needs of human beings, do you keep giving instructions to it?
There is also the question of how the criticism of artificial intelligence literature will be carried out. When critics can no longer trace back to the "nurturing materials" on which artificial intelligence depends, how will critics connect seemingly unrelated things in series? How will their comments lead us to look at social culture, and how do artificial intelligence literary works influence other texts? How will the scale of similar studies be redefined? These problems are all human beings need to face directly.
Guangming Daily (December 9, 2023, 09 edition)