“I had about 20 technologies I hoped to cover,” says Lee, speaking from Shanghai. “Stanley had the great idea of doing it in 10 different countries. We weaved what type of story would tie in a country and industry and technologies […] Stanley gets full credit for the stories.” (Chen says Lee is too modest and contributed greatly to the fiction.)

AI 2041 risks massive predictions for AI: a transformational role in areas from health care to happiness. “Extrapolating the level of technology in 10 years is not hard,” says Lee. “I do that in my day job as a venture capitalist […] We have been investing in natural language [the ability of machines to process and understand human languages] over the last two or three years, and I envision that will blossom in the next seven or eight. We can make these predictions based on how fast industry is moving forward.”

“When you see a jump, you can extra­polate beyond that,” he says. “I think the predictions I make on technologies should be about 80 per cent accurate.”

“It’s very important to portray a future to give people hope,” says Chen, also speaking from Shanghai. “We always imagine AI and robots in a more dystopian way, but in this book we try to display a brighter future to our readers, because that’s how we’re going to create a future: you have to start from imagining it in a positive way.”

At 80 per cent accuracy, we humans are in for shake-ups from birth to beyond death. In the story “Twin Sparrows”, fully articulate AIs provide individualised, immersive education for prodigious twins, one with and one without autism. Both learn exceptionally well, but it is also clear that educational philosophy is changed in the process.

In Bedeviled: A Shadow History of Demons in Science (2020), author Jimena Canales notes that in the 1970s, attempts to get computers to “understand” children’s books led to strange re­conceptualisations of both AIs and children.

She quotes Eugene Charniak, professor of computer science at Brown University, in the United States: “AI research advanced by thinking of computers as simple and ignorant little children, and children, in turn, were increasingly considered to be like not-so-smart computers. Eventually the science of education would be reoriented toward a view of learning as a sort of programming of the human mind.”

At the other end of life – or, perhaps, beyond it – the story “My Haunting Idol” begins with a seance and leads to an experience so immersive that to its characters it feels entirely real.

But what about some kind of The Matrix-like right to the truth?

“My personal view is that deception is OK if it leads to a greater good,” says Lee. “If we responsibly take charge of the new technologies that we have, and if there are checks and balances, deliberation and debate, and also when there’s a catastrophic failure, there’s a way to unravel and undo the damage, provided these things …

“Job displacement will be a big issue […] knowing AI technology can do so many tasks, and that will cause people to become disillusioned, lose their meaning of life. How would you fix that if you’re unwilling to do some degree of deception? Maybe it’s not deception, maybe it is. Maybe it’s more like a white lie.”

Chen is more circumspect on the impact of AI-driven change. “In the future, people will give away their power, their freedom, to the algorithm, to the platform, to the tech giants, unless there is regulation, there is law enforcement, to restrict those tech giants. But that creates another question,” he says. Here Chen cites one of his favourite graphic novels, Watchmen (1986), by Alan Moore – “Who’s gonna watch the watchmen?”

The Membranes: eerily prescient novel on the terrors of technology

AI 2041, through its form, also predicts that AI will touch every region on Earth. Each story is set in a different place and culture – from the castes of India to Igbo-Yoruba ethnic rivalries in Nigeria, and data storage facilities in Iceland.

This geographical distribution has a universalising intent that, to some degree, is realised, but can be culturally homogenising and, in some cases, a little too gestural. For example, distances in Australia are not quite right, and references to the stolen generations (Aboriginal children forcibly removed from their families under racist government policies), while the dates are officially correct, do not ring true.

One effect of the stories’ focus on individual characters’ lives is that the broader ethics, policy framework and institutional uses of AI become less prominent.

Says Lee, “I don’t want to go into details about what is being done in any specific country – I’m not necessarily knowledgeable about that. In [the story] ‘The Isle of Happiness’ there is a contractual agreement that the people say it’s OK for you to track me and watch everything I do, because I think I trust you and you’ll use it for my good.

“And I think that is a valid way to think about how this could move forward […] I think it would be prudent to let different countries and cultures do what they feel is right.”

For Chen, though, the limit on our hopes for AI comes back to human arrogance and ego: “The most pessimistic part for me is, it’s so difficult to change the human being from inside. Something stops us becoming a better species to coexist with others on this planet. [That cannot] be solved by technology, not AI, not quantum computing […]

“I’m right now writing the sequel to [Chen’s 2013 debut novel] Waste Tide around ecology issues around the world, but also tapped into the area of spiritualism, how people reshape their consciousness and awaken to collaborate to fight against the [ecological] disaster together. So I think that’s the very high hope. On that part I’m just trying so hard to be optimistic.”

As part of the Hong Kong International Literary Festival, Kai-Fu Lee and Chen Quifan will appear at a hybrid event from 8.30pm to 9.30pm on Nov 13. Visit festival.com.org for more details.