Why Japan’s Robotics Dream is Fading
The country is throwing big money at AI — so where is the promised tech utopia?
On a visit to Japan last year, I had the chance to look around Tokyo’s National Museum of Emerging Science and Innovation, also known as the Miraikan.
I was excited to go, because many of us who first encountered Japan via postwar pop culture have grown up associating the place with futuristic robotics — thanks to well-loved characters like Atom Boy and Doraemon.
To be brutally honest, I wasn’t terribly impressed.
I found there what felt like a retro vision of the future: humanoid and non-humanoid robots that seemed reminiscent of the 1990s.
It got me thinking about whether this particular area of Japanese innovation, touted across much of the twentieth century as central to the country’s future, might be stalling or else fading a little in Japan’s sense of itself as a country and a culture.
The potential for robotics certainly seems brighter than ever, as advances in artificial intelligence bring ever closer the day when robots — notably in Japan’s care sector — might be able to read and respond convincingly to human needs and emotions.
Already, some of these machines use AI to process things like tone of voice and choice of words, using these to work out what kind of mood their human friends might be in and what kinds of responses to offer.
Seen from the point of view of western AI development, Japanese engineers stand out as having long-ago realised that ‘intelligence’ includes both an emotional and an embodied component.
Early work on AI in places like the United States tended to operate with a definition of intelligence that was narrow, logical and disembodied. Only in more recent years have people realised that you will never get human-like AI that way.
And yet despite Japan’s success in robotics, to the point where the association of the two has become something of a cultural cliché, there do seem to be doubts out there about what the future may hold.
The Japanese government recently announced plans to expand the development of generative AI by twenty to thirty times by March 2028 (‘generative AI’ meaning systems, like ChatGPT, which produce content — text, images, etc). And all sorts of reasons have been put forward to explain why people in Japan seem more ready than many in the West to accept robot companions and carers.
One is the success of creations like Astro Boy, both in inspiring new generations of robotics engineers and in persuading the general public that robots can be kind, compassionate and unthreatening.
Philosophical explanations are offered, too. The Judaeo-Christian West, it is said, has tended to regard human beings as special and even central in the cosmos, leading people brought up there to imagine AI in terms of super-human competition.
Japanese culture, by contrast, owes to a combination of Shintō and Buddhism a sense that ‘life’ is distributed, in a meaningful and valuable way, throughout nature.
Hence the generally warm welcome for SoftBank’s humanoid Pepper robot a few years ago, when it was dressed up in Buddhist robes and programmed to recite sutras. Pepper had real life and presence — what some Buddhists call ‘Buddha-nature’.
Whatever the reasons for the comfort that people in Japan feel with robotics — some are sceptical of over-doing cultural explanations, preferring to focus on the impact of promotional efforts by government and industry — evidence now suggests that progress in at least one major area of AI and robotics may be stalling.
The care-bot industry is potentially very big business. With both Japan and its close trading partner China looking to a future in which their populations age and shrink, anyone able to produce a workable and affordable care-bot — keeping elderly people comfortable while tending to their care needs — could in theory make a great deal of money.
But production of Pepper has been paused, partly because of poor demand, and a new book by James Wright — Robots Won’t Save Japan (2023) — reveals that some care homes in Japan are struggling to adapt to the use of care-bots.
Those care-bots, it turns out, often need caring for themselves, from maintenance to being moved around from room to room.
Some of the AI challenges facing Japan are similar to those with which many other countries around the world are being forced to grapple: where to focus AI investment (Japan seems to be going for generative AI, while China claims to be working on cognitive AI: systems that replicate human cognition), and how to plan for a future labour market shaped by automation.
But Wright’s research seems to hint at even bigger questions for Japan, given its demographic challenges. What would be the ideal size of population for a country like Japan, bearing in mind the impact of human beings on the climate? And what sort of society does it want?
One possible future involves precariously-employed migrant workers with little opportunity to learn Japanese or other higher-level skills, looking after both elderly Japanese and care-bots.
The latter may not exactly steal their jobs, so much as condemn them to working in lower-skilled ones, for poor pay, since complex skills like understanding and speaking Japanese can be handled via AI-powered robotics.
These are highly emotional, deeply moral problems: concerning how we ought to care for the elderly, and how human beings and robots ought — one day — truly to work together. Japan, it seems, is pioneering not just the technology involved, but also the practicalities and the ethics of putting that technology into action.
