The Chinese Singularity
Dr. Hugo de Garis, the father of evolvable hardware and a redoubtable AI researcher, moved to China several years ago, and is now leading the Artificial Brain Lab at Xiamen University. He is convinced a Singularity in the vein of Vinge and Kurzweil is likely to occur later this century — and that China is the most likely place for human-level Artificial General Intelligence (AGI) and the other critical technologies underlying the Singularity to arise.
As Hugo puts it: “China has a population of 1.3 billion. The US has a population of 0.3 billion. China has averaged an economic growth rate of about 10% over the past 3 decades. The US has averaged 3%. The Chinese government is strongly committed to heavy investment into high tech. From the above premises, one can virtually prove, as in a mathematical theorem, that China in a decade or so will be in a superior position to offer top salaries (in the rich Southeastern cities) to creative, brilliant Westerners to come to China to build artificial brains — much more than will be offered by the US and Europe. With the planet‘s most creative AI researchers in China, it is then almost certain that the planet‘s first artificial intellect to be built will have Chinese characteristics.”
Is he right?
(Full disclosure: I spent a month at Hugo‘s lab in Xiamen this summer, and Hugo and I recently received word that the Chinese National Science Foundation has approved a grant to fund his lab to pursue some of our joint research on cognitive robotics, aimed at enabling the Nao humanoid robot to learn, reason and communicate in English and Chinese.I‘ve even debated making a move to Xiamen myself.So I can‘t claim great objectivity on this topic… and indeed it was with some personal fascination that I asked a variety of individuals involved with AI research and software technology in China whether the Singularity will be Chinese.)
My first destination on my quest for wisdom about the Chinese Singularity was a visit to Temple University AGI researcher Dr. Pei Wang, a long-time US resident who visits his home country of China each summer. Pei expressed a milder version of Hugo‘s sentiments: “I think China is among the most likely places (though not the only one) where the first truly/generally intelligent system will be created… Given the population size and education level of China, its chance is quite large… there are profound intellectual resources to make AGI happen.”
Pei points out that “one of China‘s major advantages is the lack of strong skepticism about AGI resulting from past failures.” The US and Japan have spent large sums on AI research in past decades with disappointing results, and as a consequence are particularly skeptical of AI relative to other research areas. China never had that experience, and is making its first serious foray into AI in an era blessed with more powerful computers and deeper knowledge of cognition and computer science. Pei also noted that the research community in China tends to favor incremental research over riskier attempts at paradigm-shifting progress. This seems to have held true in the AI field, so far: Chinese AI researchers have made important innovations in multiple areas such as fuzzy systems, genetic algorithms, machine translation and spatiotemporal logic, but haven‘t yet launched any AI revolutions.
Dr. Min Jiang, an assistant professor in Hugo‘s Artificial Brain Lab specializing in AI cognition and formal logic, indicated a factor counterbalancing this conservatism: “In many fields, China today is a follower. But maybe this is part of the reason China wants to spend research money on innovative projects. It can be considered a ‘tuition fee‘ and an investment in the future. Even if some projects fail, we can learn lots of things from the experience.” The funding Hugo‘s lab has received seems to be evidence for this perspective. And this spirit of experimentation is precisely what will be needed to create AGI and other radical Singularity-enabling technologies.
If the Chinese fund an experimental Singularity-relevant project, and it yields sufficiently impressive results to excite the “power circle,” dramatic things might happen.
Min offered further insights into China: “I think the most important advantage (or disadvantage) is the [political and governmental] system. If the power circle thinks a project is crucial, we do that with all the strength of the country: for example — A-bomb, spacecraft.” Another example is the First Solar initiative launched in September 2009, a 10-year project aimed at blanketing 25 square miles of Inner Mongolia with solar panels, generating 2 billion watts of power, enough to light up three million homes. When the Chinese government really wants to do something, they think big.
This combination — a willingness to experiment with new ideas, and a willingness to put massive funding behind selected initiatives — is intriguing. If the Chinese fund an experimental Singularity-relevant project, and it yields sufficiently impressive results to excite the “power circle,” dramatic things might happen. This is exactly what Hugo has in mind with his “CABA” proposal, which he presented at the Oriental Technology Forum in Shanghai this October: “What I propose is that the Chinese government should create a ‘CABA‘ (Chinese Artificial Brain Administration) over the next 5-10 years, consisting of thousands of scientists and engineers to design artificial brains for the Chinese home-robot industry and other applications. CABA would do for artificial brains what the CNSA (Chinese National Space Administration) does for space, i.e. it employs thousands of scientists and engineers to design and control rockets for China‘s space applications.” Wildly ambitious? Perhaps. But so is covering 25 square miles of Mongolia with solar panels.
I found Western entrepreneurs operating technology firms in China to be the most skeptical voices regarding the possibility of a Chinese Singularity. I interviewed two such individuals in depth. Both are Singularity optimists, and both were concerned that their remarks be kept anonymous, to avoid potential harm to their Chinese business work. Both put the odds of a Chinese Singularity launch at less than 5%, and they gave similar reasons: they consider Chinese engineers on the whole “below average in problem solving and creative thinking,” “very conservative, unwilling to considering doing anything that is not established practice.” One of them also noted that “Local above-average talent insists on working for American, European, Japanese, or Korean (in that order) firms rather than Chinese firms. So, the best chance for AI breakthrough here is with a foreign research effort.”
I have heard this complaint about a “lack of creativity” before, but it runs counter to my own experience at Xiamen University. There, while I‘ve encountered some conservatism, I‘ve also met some extremely creative and individualistic young professors and students. In my experience, researchers in China are just as creative as anywhere else — but there are subtle sociocultural issues at play, with different implications in the corporate and university contexts. Chinese culture, in its current incarnation, tends to spawn social structures that suppress rather than encourage the expression of personal creativity. It also doesn‘t tend to support Western-style teamwork. There‘s a proverb to the effect that “a lone Chinese is as powerful as a dragon; but three Chinese together can‘t match a bug.” These are real issues, yet ones that can be worked around with care, using different methods depending on the context.
It must be understood that, regarding personal creativity as other matters, Chinese history has been powerfully cyclical. In his controversial recent book 1434, Gavin Menzies argues that the Italian Renaissance was launched by a fleet of Chinese ships that sailed to Italy and distributed advanced knowledge including encyclopedias from which Leonardo da Vinci indirectly derived many of his celebrated illustrations of mechanical devices, flying machines, and so forth. Whether or not this thesis is true, Menzies presents compelling evidence regarding the advanced level of Chinese engineering and science during that time period, before a change of administration in Beijing ended the period of wild invention and exploration and brought a new era of conservatism to China. My point is that Chinese “cultural DNA” has plenty of innovation and creativity in it, and one must be careful to distinguish stable characteristics of Chinese culture from cyclically-shifting ones. The pendulum of Chinese culture swings in a wide arc.
In the corporate software development context, one strategy for working around counterproductive cultural tendencies and bringing out Chinese creativity is the adoption of “agile” software development methods. A 2008 article in InfoQ summarized the experiences of five Chinese software firms who adopted the “Scrum” development methodology — a very dynamic teamwork-based approach to making software that requires constant adaptive creativity on the part of the participants. Three found the approach successful; two did not. Those who didn‘t find success complained that the development teams or managers understood the formalities but not the essence of the agile approach — the cultural disconnect was too great. And this is surely related to the reason why Chinese universities are so eager to bring in Western professors, like Hugo de Garis. It‘s not just the research ideas the Westerners bring, it‘s the different intuitions, experiences and habits regarding directing a research lab and a research program. In this sense Hugo‘s emphasis on China bringing “creative brilliant Westerners … to China to build artificial brains” may be savvy. If China can leverage its economic growth and openness to innovative research directions to recruit a sufficient number of Western research mavericks, then powerful things may happen. Imagine a situation in which every Chinese city has a number of labs, focused on Singularity-relevant technologies, in which Western research leaders are hard at work bringing young Chinese scientists up to speed on Western ways of doing creative team R&D. In this quite plausible scenario, the prospect of a Chinese Singularity doesn‘t seem so farfetched.
As well as AGI, it‘s also worth noting the differences between Western and Chinese attitudes on another radical future technology: life extension. Westerners tend to greet talk of immortality with skepticism or even moral disapproval — after all, the standard Christian story is that God wants us to die and go to heaven. But the Chinese memeplex is stocked with thousands of years of Taoist tales of immortality. Traditional Chinese methods of achieving immortality are often arduous; for instance Taoist Yoga has techniques involving lifelong celibacy and meditation focused on eventually giving birth to one‘s immortal self through the top of one‘s head. Many Chinese would be very open to immortality or life extension pills that could deliver the same benefits at lower cost and with greater reliability. So far this attitude has not translated into dramatic funding for life extension research, but the potential certainly is there — as is the economic motivation, since China will face a severely aging population around 2025-2030, similar to what Europe is facing now.
The Chinese government should create a Chinese Artificial Brain Administration, consisting of thousands of scientists and engineers to design artificial brains for the Chinese home-robot industry.
David Chambers of the Methuselah Foundation, discussing the 2006 Tomorrow‘s People Forum at Oxford University, compared Western and Chinese attitudes on life extension technology as follows: “Europeans don‘t look forward to a better future — but rather a managed version of the present. There‘s a distrust of revolutionary ideas…. [But] while Euros and Americans might have their various hang-ups about the ethics and implications of the new biology, China doesn‘t. Pei Xuetao, of the Beijing Institute of Transfusion Medicine [a leading institution in stem cell research and regenerative medicine], made it very clear [in his talk at the Tomorrow‘s People Forum] that China is open for business.” Alongside research aimed at curing cancer and other diseases, Xuetao and his colleagues have made important discoveries involving cellular senescence and apoptosis, working toward an understanding of the genetic networks that make us age.
These differing attitudes toward immortality may be connected with attitudes toward AGI. Western skepticism about AI may not be entirely due to prior AI funding fiascos, but may also be tied to deep-seated cultural issues. The same Christian memes that tell us we‘re supposed to die and go to heaven also tell us that machines can never truly be conscious because they lack an immortal soul. Yet Changle Zhou, the dean who supervises de Garis‘s Artificial Brain Project, regularly refers to Hugo‘s work as the “Conscious Robot Project.” Chinese culture has little of the West‘s subliminal resistance to thinking machines or immortal people and this cultural difference may manifest itself in the next decades in subtle ways.
Another cultural difference to remember is that extrapolating progress in China by plotting linear or exponential curves often doesn‘t make sense. Progress in China often matches the biological notion of “punctuated equilibrium” — long periods of relative stability punctuated by surprising and sudden changes. The Cultural Revolution and the recent shift to market-oriented “Socialism with Chinese characteristics” illustrate this phenomenon — as do the sudden initiation and cessation of Chinese global seafaring in the 1400s, and dozens of other instances in China‘s long history. It‘s easy to imagine a single technological breakthrough catalyzing one of these sudden shifts in the near future. It could be intelligent robotics, it could be life extension or something else wild and unforeseen. While this article was in the editing process, I heard some fascinating talk about a very substantial amount of funding being allocated by Beijing to a project called the “head brain instrument” (three Chinese characters) intended to improve neural function and hence accelerate human learning. I don‘t know enough about it to assess the viability but if it works out, it sounds like the sort of thing that could punctuate any nation‘s equilibrium!
The possibility of a Chinese Singularity may strike fear into the hearts of American nationalists or Eurocentrists, but it‘s not clear that it will make a big difference which nation makes the crucial breakthroughs. In today‘s scientific world “information wants to be free” — and since the most likely path to a Chinese Singularity involves collaboration of Chinese and Western researchers, the odds of an insular Chinese Singularity uniquely serving Chinese national interests seem fairly low. The work in Hugo‘s lab in Xiamen centers on open-source software development. It‘s evolving cooperatively with work done by AI coders outside China and it‘s delivered freely to the international research community.
So what‘s the verdict? Given China‘s lack of hang-ups about AGI and life extension, its powerful economic growth, its large population of smart and hard-working young scientists, and its eagerness to import western research leaders — will the Singularity be launched in China? I‘ll give the last words to two creative young scientists from Xiamen University.
Min Jiang made a statement I found intriguing given China‘s ongoing obsession with its 5000-year-old culture: “Today‘s China is a young boy, and as you know, eighteen is the age full of curiosity and fantasy about the future!”
And Ruiting lian, a PhD student at the Artificial Brain lab focused on multilingual natural language comprehension, generation and dialogue, cut to the chase more directly: “In China, the best answer to every question is ‘maybe.‘”
Ben Goertzel is the CEO of AI companies Novamente and Biomind, a math Ph.D., writer, philosopher, musician, and all-around futurist maniac.