As smart technologies pervade our lives like never before, we turn the spotlight on some of Hong Kong’s most noteworthy achievers from across different creative fields who engage with Artificial Intelligence, or the idea of it. The five-part series explores what architects, artists, musicians, designers and writers make of the ways in which machines are impacting human society and how they plan to adjust and renegotiate their places in a changing world order where AIs outnumber humans by several folds. The fifth and last instalment features the novelist Dung Kai-cheung, who imagines AIs of the future as sentient, conscious, mindful beings, trying to make sense of the world by experiencing it. 

Novelist Dung Kai-cheung’s current preoccupation with the AI theme is an extension of his interest in the relationship between man and machines. (EDMOND TANG / CHINA DAILY)

• A human writer is like a machine in the sense that they often go through a series of trial-and-errors to arrive at arranging words in a sequence that clicks and is meaningful.

• The Open AI program GPT-2 can write letters and news reports logically and cohesively, continuing a piece started by a human writer.

• While AIs are producing what might loosely resemble stream-of-consciousness novels, they are yet to come up with well-defined plots, characters and a logical progression of events and situations to hold the story together. 

Kazuo Ishiguro, who won the Nobel prize for literature in 2017, has said his next work of fiction will be about Artificial Intelligence. AI-themed novels were trending in 2019. Ian McEwan’s Machines Like Me, for example,  puts a batch of sentient robots in an alternative history of the UK in the 80s. Frankissstein by Jeanette Winterson imagines a future in which it is possible to store millions of gigabytes’ worth of memories in a revived and fully-functional head of a dead scientist. In “The Lifecycle of Software Objects”, a short story in Ted Chiang’s much-lauded collection, Exhalation, people raise cute-looking artificial intelligence pets who can speak and return affection.

It may not be a coincidence that writers around the world are drawn to the idea of AI at this moment. Hong Kong writer Dung Kai-cheung, who published a two-volume AI-themed novel, Beloved Wife, this year, and is now writing a second, titled Post-Human Comedy, says this is a sign of writers responding to the exigencies of the present. 

“I think it’s inevitable that we need to deal with this, because we are actually living in an age in which AI and all sorts of new technologies are changing not only our ways of life, but also the way we define life,” says Dung. 

Such existential questions have always figured in Dung’s often highly meta-textual writings, drawing on sources as diverse as the 4th-century Daoist philosopher Zhuangzi and American sci-fi writer Philip K. Dick (1928—1982), with a nod to the tantalizingly mind-bending works by Italo Calvino (1923—1985) and Jorge Luis Borges (1899—1986). There is a sense of continuity in the way Dung has moved from maps (Atlas, 1996) to mechanic’s tools (The History and Adventures of Vivi and Vera, 2005) to artificial intelligence in Beloved Wife, using the frame of man-made appratus to tell stories in which archived history and imaginary landscapes often dovetail. 

Still, Beloved Wife and its spiritual companion, Post-Human Comedy, could be Dung’s most-strikingly experimental attempt yet at trying to renegotiate the relationship between man and machines. The story in Beloved Wife progresses through a successive building and shattering of fictional scenarios. There’s no telling until the very end if the anonymous AI appearing in a dialogue in the story could be traced back to the revived consciousness of the dead writer wife of a professor of Chinese literature or that of the professor himself who has been dead for years, or if they are both dead and their conscious minds have been uploaded on to the same body. 

In Post-Human Comedy, a cybernetics expert is recruited to fix certain issues with the software in the system of some cyborgs. The CPU in their system is called Kant Machine after the German philosopher Immanuel Kant (1724—1804), and their original setting is based on Kant’s thesis on the 12 categories of the mind. 

“There is also a politician trying to invent a kind of ideal citizen,” says Dung, adding that the character represents governments that are not comfortable letting people enjoy a great deal of freedom. “Hence they try to invent political systems they think will help maintain stability.”

Trying to develop cyborgs on the lines of the Kantian model is a tricky proposition, but probably worth giving a shot when it comes  to  developing future generations of real-life AIs, Dung believes. 

“In The Critique of Pure Reason (1787), Kant ponders on the question of how one could be free and conform to laws at the same time,” he says, the idea being, “because you are free, you are fully able to be a legislator of yourself — make laws that stop you from engaging in untoward actions. The politician thinks if we can create citizens having such a frame of mind, we need not be afraid of a completely open political system, as citizens will not deviate (from the rules they set themselves). They will not do anything immoral or politically dangerous.” 

But then, an AI’s mind could be as unpredictable and unknowable as that of a human, and such factors are expected to figure in Post-Human Comedy, whose ending Dung says he doesn’t know yet.   

The human machine

From Mary Shelley’s Frankenstein (1818) to McEwan’s Machines Like Me (2019), the theme of AIs rising in rebellion against their utilitarian, and insensitive, programmers has surfaced often enough in literature, and films. 

In real life the fears around the exponential growth of AIs who already outnumber the human population is not so much because one day they might want to be avenged against the human race that did not treat them well, but rather because AIs might actually replace humans by being better at what they do, says Dung. This is already in evidence as industrial robots replace human labor. Since 2013, China has been the world’s biggest importer of automatons to work in its manufacturing sector.      

However, the prospect of AIs taking over the planet and subjecting human society to a status of serfdom, he feels, is a bit outlandish. For “as long as they don’t have consciousness and autonomy, no matter how powerful they are, AIs are going to be working for others.”

A future generation of sentient, free-spirited AIs “might actually do a better job of running the world than we do,” Dung says. However, having a physical form and dimension is essential for AIs if they are to achieve those human faculties, he adds.

“I do believe in order to develop consciousness, machines must have a body, or sort of hardware,” says Dung. “It could be of human shape or other shapes. I don’t believe what exists digitally, without a body, can become conscious of its own existence one day. Without a body it’s just computation and data and things like that.”

“In my novels I always imagine AIs as having a human body in which the consciousness is connected to or replaced by an AI-like technology,” he adds.

He would like AIs to have direct experience of the world, and not just through the mediation of gazillions of data fed into their systems. “I think AIs will have a completely different world view if they are made to learn through experience,” says Dung. “This resonates with Martin Heidegger’s idea of ‘worlding’, (Being and Time, 1927). The world is not something out there, objectively there, or somewhere ideally there, and we set off to find it. The world is the thing that we live in and create by our experience. And so if we, human beings, get to learn things like this, I think AI could also be made to learn this way, and not by being fed information on some ideal things.” 

Unsurprisingly, the AIs in Dung’s own novels come in all sorts of shapes and sizes. In Post-Human Comedies, he took his cue from an apocryphal story about the French philosopher Rene Descartes (1596—1650) who supposedly took with him wherever he went an automaton sculpted after his daughter who had died at the age of five. The story is important, even if it isn’t exactly true, says Dung, “because it illustrates Descartes’ view of the human body as a machine.”

The idea ties in with Calvino’s view of writers as machines carrying out a series of trial and error experiments with words until they succeed in arranging them in a sequence that clicks, says Dung, alluding to the Italian writer’s 1967 lecture, Cybernetics and Ghosts.   

“Calvino reversed the way we look at the relationship between man and machines, saying it’s we human beings who have always been machines, very complicated ones. But can we create another machine that is nearly as complicated as we are? And is it necessary to do this?” says Dung. 

Learning through experience

The first AI-generated novel, 1 the Road, was published in the UK in 2018. The Chinese sci-fi writer Stanley Chen Qiufan used AI to write some of the dialogues in “Fear Machine”, a short story in the collection Algorithms for Life, published earlier this year. While AIs may still be some distance away from giving serious competition to human authors, Dung feels such an eventuality cannot be ruled out altogether. 

He mentions how the company OpenAI released a program called GPT-2, which has the ability to draft letters and reports by following the rule of thumb that every new word added to the opening paragraph, written by a human, follows the logic and theme of what precedes it. “I find this idea strikingly similar to Calvino’s thesis,” says Dung. 

As the prospect of writers and AIs collaborating to create novels of the future begins to look like a distinct possibility, Dung is in favor of opening up access to material — works by other writers, for example — to facilitate the development of AI writers. 

He is now especially aware that human writers cannot let themselves write with a sense of complete abandon in the same way that machines are doing. “While the basic thing about literature is freedom of imagination, at the same time we need to be aware that maybe there are some limits to this freedom. I don’t think poetic license is absolute,” Dung says.

He imagines AIs too will need to work within certain parameters if they are to produce material other than what could be loosely categorized as stream-of-consciousness writing. 

“What we need to work out probably is: how to allow AIs freedom, creativity, spontaneity, and at the same time make sure they know how to build rules of conduct and learn to tell right from wrong. I don’t know if one day we will be able to invent an AI that is able to do that.” 

Click here to read Dung Kai-cheung’s exclusive interview to China Daily

Contact the writer at basu@chinadailyhk.com