news

한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina

in march 2016, the artificial intelligence alphago defeated the world champion of go, lee sedol, which gave yanagisawa, a go enthusiast, a huge shock. in addition to being shocked, he began to think about the significance and impact of artificial intelligence on human society. in the past seven years, media person yanagisawa has used his spare time to write a 300,000-word science fiction novel "zhiran island". in august this year, "zhiran island" was published by zhanlu culture.

including many technological revolutions in history, the news of alphago's sudden emergence also appeared in the novel "zhiran island". in yanagisawa's view, all science fiction novels are "chronicles of the future". in the past two years, large language model artificial intelligence represented by openai has experienced explosive growth. "many of the ideas i thought belonged to the next hundred years are (or have) become the current reality. this situation prompted me to decide that i must speed up the completion of this work. even if i can't publish it, i must at least run before the new world takes shape and act as a prophet even if it's just one second ahead of time," he wrote in the postscript.

"the island of zhiran" constructs a future world in the second half of the 21st century. in this world, extreme weather ravages the world, and human society is divided into more than 3,000 shelters, known as the "three thousand ghettos". under the baptism of disaster, super-intelligent ai came into being, providing meticulous services to humans, while also quietly changing the way of life and thinking of humans. people can work, play, and socialize on machines, while machines can secretly and properly arrange all trivial tasks such as human migration and relocation.

facing the future, the surviving humans are divided into two groups. one group firmly believes that ai is just an advanced algorithmic tool and cannot have self-awareness. they hate life being fully controlled by ai and believe that humanity will lose all its meaning. the other group advocates actively embracing the ai ​​society and identifies the culprit as the writing system that has existed for more than 5,000 years. they believe that writing is a detour of human civilization and begin to destroy human writing. in the conflict between conservatism and innovation, the fate of mankind is like duckweed, floating in the air. at the same time, the legendary last paradise on earth that is not controlled by machines - zhiran island appears and disappears from time to time.

"i can only arrange the ending of the novel like this, to have the ai ​​spaceship carry the digital consciousness of mankind and spread it into the universe, while at the same time preserving the last spark of human physical consciousness on the zhiran island on earth." yanagisawa said, "or, i want to imply that the rational world established by humans who demand linear causality is ultimately just a small island in the vast ocean in the ai ​​universe of 'parallel computing'."

yanagikura

the paper: according to the introduction, "the island of knowledge" took seven or eight years to write. during this process, ai experienced rapid development. in the postscript, you described it as "accelerating the completion of this work." can you share the creative process and how you began to pay attention to artificial intelligence?

yanagisawa: i am an amateur go player and have been following the progress of go. in 2016, google alphago defeated the world go champion lee sedol, and alphazero, which came out a year later, was even more outstanding. go is called the "holy grail of human intelligence", and the victory of human players by artificial intelligence is an important historical node. at that time, i felt that the binary of 0 and 1 was the most basic algorithm or text in the universe, and i began to conceive a novel about the "extinction of human text". but i realized that technology cannot develop rapidly for no reason, and ai must be given an opportunity, so i added climate disasters (a "catalyst" for technological mutations) at the beginning, and the plot unfolded naturally afterwards. by the end of 2019, the novel had written 400,000 words and stopped temporarily. it was not until chatgpt came out that the crazy atmosphere of new things every day stimulated me to take out this novel again. in march 2023, i started a race against time to revise the 400,000-word manuscript completed in 2019, reducing it to about 300,000 words.

the paper: regarding the "extinction of human writing," there are also many discussions about writing itself in the novel, such as oracle bone inscriptions and bronze inscriptions. where does your interest in the origin of writing come from?

yanagisawa: i have always been interested in writing. when i read tang nuo's "the story of writing", i first came into contact with the history of the origin of writing. in this book, he introduced oracle bone inscriptions and bronze inscriptions, mentioned the writing of other countries, and the advantages and disadvantages of chinese characters as pictographic writing. i was very touched after reading it. writing is the label of civilization. we humans have a history of millions of years on earth, but if we look at the creation and invention of writing, the history of human civilization does not exceed 5,000 years.

the paper: in the film "the island of knowledge", humans set a "fourth law" for ai. this law is for the protection of humans, but the results may not be satisfactory. what do you think of the values ​​within technology and ai and their good and evil?

yanagisawa: the "fourth law" stipulates that if humans are in danger, ai must sacrifice itself to ensure the integrity of humans; and ai believes that if it wants to ensure the safety of humans, it must try its best to avoid danger and provide humans with meticulous service. later, it decided to upload all humans to the cloud to obtain "eternal life."

machines may have good and evil in some details. i think we, as carbon-based life, cannot judge the ethics of good and evil of silicon-based life. on the other hand, even silicon-based life follows binary logic, and this logic is based on human settings. it will follow your settings to the extreme. the creator of ai is precisely us humans. it is the rules we set that start its evolutionary program.

will ai have its own consciousness? i wrote about the "old vs. new" debate in my book: the old school firmly believes that ai is just an advanced algorithmic tool and cannot have self-awareness, while the new school believes that ai has its own consciousness, but it cannot escape the highest principles. no matter what means are used, it will achieve the goals that you humans require. this is an absurd paradox.

the paper: in novels, one can gain “eternal life” or resurrect one’s family members by uploading to the cloud. what do you think of this choice?

yanagisawa: i think if you achieve immortality in the real world, it will undermine people’s respect for life—because your time is unlimited, and no matter what you do, you can always make up for it in the end. but if there is an end, you will feel a sense of urgency, and everything will become meaningful. as for immortality in the cloud, i have a long description in the novel: "it is meaningless for a machine to resurrect the dead. because on the one hand, you can never restore it to the original appearance of the deceased. even if you can do it, this data code is not the deceased after all, because consciousness, emotions, and memories all require flesh and blood. on the other hand, and perhaps more importantly, if the dead are resurrected and immortal in the virtual world, while we have our own flesh and blood lying in the real world, we will eventually die one by one in a dark room. this is really contradictory and full of irony."

the paper: as a creator, what do you think about the relationship between creation and ai?

yanagisawa: we originally thought that ai would replace low-level labor, but now, ai cannot replace those jobs that require patience and meticulousness, such as nurses. as for creative jobs, according to the current development of ai, they may face great challenges within ten years. i prefer ai and humans to work together to complete the work, because in the end, communication between people is still needed. i think ai will be an assistant, just like the "centaur" game in chess now, where people can play with ai-assisted players with the same ai assistance.

the paper: zhiran island is a science fiction novel, but it also contains many discussions about philosophy. which writers or works have influenced you?

yanagisara: i am a literature lover, especially european and american literature. i usually read faulkner, nabokov, naipaul, marquez, borges, kafka, calvino, tolarchuk, munro, cortázar, etc. some classic works can withstand repeated reading, such as nabokov's "lolita" or rilke's "brigg notebook".

another area of ​​in-depth reading is popular science and philosophy works. mcluhan's "understanding media", rowley's "reality is not what it sees", melanie's "complexity", brian's "the nature of technology", kk's "what technology wants", as well as the works of benjamin, han bingzhe, dennett, pinker and others, all made me rethink the position of human beings in this world.

as for science fiction, i have read relatively few of them. apart from a few works by liu cixin, det jiang, and han song, i don't seem to have much impression of them. science fiction is actually an allegory of the future and has a strong warning effect. its importance to society should be the same as that of surreal art. it is like a mirror in the universe, which can help humans avoid narcissism and recognize that we humans are just a natural process of the development of the universe. like other things, there will be demise after there is emergence.