The books arrived in my reading stack at the same time: AI 2041: 10 Visions for our Future and The Perfect Police State. So which would I read first? I chose The Perfect Police State. That order turned out to matter.
Investigative journalist Geoffrey Cain’s book is about what is happening now to ethnic minorities, primarily Muslim Uyghurs, in the Xinjiang region of Western China. Cain reports that in Xinjiang “The Situation” is what Uyghurs call this bald application to them of extreme surveillance techniques and authoritarian power. Their carefully neutral name for it captures the whole point of the exercise: to make them feel powerless and surrender their minds and culture to the larger political state. The book’s subtitle, An Undercover Odyssey into China’s Terrifying Surveillance Dystopia of the Future, promises a scary deep dive into a situation which – like all distressing human rights problems not happening to us – we’d prefer not to have to think about.
Thus, while I ordered The Perfect Police State with uneasiness and a sense of obligation, I anticipated AI 2041 with interest and excitement. This book by Chinese venture capitalist Kai-Fu Lee and science fiction author Chen Qiufan brings together 10 imaginative scenarios of how technology may evolve in 20 years, with relatively brief but highly informed explanations by one of today’s best known AI leaders of the technologies underlying those scenarios and their likelihood for development. Lee is not a mere “futurist”. An AI pioneer at Carnegie-Mellon University in the 1980s, his perspective rose from trench-level to birds-eye as he eventually held senior executive roles in the US and China at Microsoft, Apple and Google.
Reading Cain’s book first mattered. It cast a heavy shadow over our chances of achieving the largely positive opportunities portrayed in AI 2041. It left me feeling “never mind the future, Mr. Lee. What are you doing about the Situation that is actually happening now in your own country?” Lee’s last best-selling book, AI Superpowers, provided a terrific window on how differently Chinese and American technology titans, workers, governments and cultures approach the opportunities of an AI-driven world, and the battle they are locked in to dominate it. Now, he’s purposely exploring positive outcomes as he says in his introduction: “Amid what seems like a feedback loop of negative stories about AI, I believe it’s important to tell these other stories, too, and answer that question of ‘What happens next.’” Can this question be answered, however, without engaging the terrifying applications of AI technology happening right now in Xinjiang?
Under the AI Big Top
Circuses are politically incorrect these days but the image remains fresh (and even can still sell tickets if styled the right way – see the movie The Greatest Showman). It strikes me that a book such as AI 2041 may function very much like the Big Top’s Center Ring where a multiplicity of acts simultaneously thrill, frighten, and overwhelm.
Here is the titanic elephant act – Big Tech — that dominates the Ring and threatens to crush everything else if control cannot be maintained. Here, too, are half-tamed predatory animal acts – the Move Fast and Break Things beasts. Above are the high wire acts – billion dollar “unicorns” flying from trapezes and inching across tightropes where risk is great, success richly rewarded, and failure costly. Scattered throughout are the precision jugglers of data analytics; acrobats (robots) performing their seemingly impossible stunts; while be-sequined beauties on horseback (digital programming, CGI, gaming, AR and VR) entertain and delight. The ringmaster and assistants, the trainers and handlers (think regulators, ethicists, authors, pundits, mainstream and tech media, researchers, venture capitalists) are all very busy trying to keep everything moving under control and explain what is happening; while clowns (influencers and fringe media) divert the eye from slips, falls, and dropped balls; and whistle blowers perform their fire-breathing, death-defying acts.
It’s all very eye-catching, frenetic, and really hard to keep track of – in fact it’s exhausting. The Center Ring is a costly spectacle that grips our attention while vendors in the aisles (advertisers and retailers) hawk every form of food, drink and anything else we could want, financially sustaining much of the show.
In the Shadow of the Big Top
What if before entering the Big Top, though, a person visited the animal cages and carnival living quarters – the far less savory aspects behind the circus spectacle? Would this be akin to engaging issues such as the present-day test bed of advanced identification and surveillance technology deployed in Xinjiang and increasingly throughout China? Or what if our way into the spectacle of the Center Ring led through the shady and questionable “freak shows” of days gone by – even the seedier dark corners for gambling or illicit activity – AI technologies applied on the Dark Web, in Autonomous War, or toward the creation of future AI Overlords of General Intelligence?
The spectacle under the Big Top is far less enjoyable after coming face to face with the current wholesale captivation of an ethnic and religious minority and the destruction of its ancient culture, as painfully described in a book such as Perfect Police State. It further dims when we see the magnitude of the potential harms outside the Tent. And why shouldn’t this be the case? Circus spectacles have fallen out of favor for good reason – the remarkable feats of acrobatic skill and courage could not overcome a higher respect for human dignity violations in the Midway and evolving standards of animal cruelty within and outside the Big Tent. In the same way, a wholesale celebration of AI-powered possibilities becomes less acceptable as awareness grows of the potential and actual abuse of anything-goes data collection; the transmission of human bias and discrimination into screening algorithms; and the scary capability of always-on surveillance technology wielded without restraint.
Whose role is it to mediate the excitement of AI possibilities with the deep risks of AI empowerment? In a well-functioning democracy, the government could be expected to be a major mediator. But when the government is ineffectual, or worse – a critical part of the problem, it falls to other parties with sufficient knowledge and power to support the spectacular good that can come from AI-powered technology while denouncing unethical, unjust and potentially catastrophic practices occurring in the shadow of the spectacle.
Which takes me back to co-authors Lee and Quifan and how I should feel about their book because I read Cain’s book first. Both Lee and Quifan live in China, where Lee is now one of China’s most prominent venture capitalists. He has over 40 million social media followers. But it’s clear from The Perfect Police State that no matter your stature in business or culture, within China the Situation in Xinjiang is untouchable. If the Lee of AI Superpowers is to be free to act as an ongoing intermediary and translator of sorts in this continuing superpower fight, should we in the West cut him some slack in his current book for skirting the spreading dystopia for ethnic minorities and people of faith in his own country?
And broadening the question, how do all of us hold together conceptually and with integrity the challenges and opportunities of AI-powered applications, both in the present day and in our hopes and fears for the future? This is a fundamental question of AI ethics, and one which the ancient, systems-based beliefs and values of people of faith may prove uniquely helpful in answering.