National / Media | BIG IN JAPAN

Advances in robotics present singular worry

by Michael Hoffman

Special To The Japan Times

“Singularity” is an odd word. Originally it meant peculiarity. Then 20th-century physicists got hold of it and situated it at the very boundary of space-time, to the eternal bafflement of the lay mind. A singularity became the inexplicable, infinitely charged nothing out of which the universe grew following the Big Bang.

A new “singularity” now looms — the moment, fast approaching, when our machines surpass us in intelligence. Conservative futurists — an oxymoron if ever there was one — see it coming around 2045. Others say 2030. What’s on the other side? Nobody knows. Good things in abundance, seems the majority view.

“Abundance,” in fact, is the name of a global movement promoting an “Internet of Things” — ultimately an “Internet of Everything” — to follow hard upon the “singularity.” Everything you own — phones, clothing, refrigerators, TVs — connected to the Internet, the better to serve you, meet your needs, satisfy your whims!

Imagine, as some of the most luminous brains on the planet do, 45 trillion sensors, all linked in one vast global network. Give it 20 or 30 years and it will happen, says, for example, Janusz Bryzek, vice-president of micro-electrical-mechanical systems (MEMS) and sensing solutions at the Silicon Valley firm Fairchild Semiconductor International.

Bryzek claims the world’s most intractable problems — hunger, disease, lack of clean water and air, lack of energy — are within a generation of being solved. Microsoft co-founder and philanthropist Bill Gates predicts “almost no poor countries” by 2035.

Artificial intelligence isn’t the only source of optimism, but it’s a big one. We’re talking, in the long run, about machines “billions of times more powerful” than anything we know, says the weekly Shukan Diamond. “There’s no imagining where this expansion of human abilities is taking us.”

Shukan Diamond celebrates “the robot and AI revolution.” “Singularity” seems the better word, but Prime Minister Shinzo Abe in May proclaimed “a new industrial revolution,” and so, in Japan at least, “revolution” it is. Robotics have long been Japan’s forte — in manga (beginning with “Tetsuwan Atom” or “Astro Boy” in the 1950s), real life (at least 250,00 industrial robots in use in Japan, more than anywhere else in the world), and the cross-over between the two represented by Asimo the humanoid, developed by Honda in 2000 and still going strong, as a curiosity if nothing else.

No challenge to Asimo in terms of charm but far outpacing it in terms of usefulness is a tropical-sea-blue humanoid, still unnamed, developed by Schaft Inc., a Japanese startup whose two co-founders, both former University of Tokyo professors, sold out to Google months before the robot they designed to function in disaster areas too dangerous for humans trounced the competition at the two-day Pentagon Robotics Challenge Trials in Florida in December 2013.

You look at this creature and think, “Do I want to share the planet with it and its like?” Not only the planet. Your home, your work place, your nursing home when the time comes — for it won’t be long, Shukan Diamond says, before robots are doing much of the heavy lifting involved in caring for the growing ranks of infirm elderly. Their immunity to fatigue and mental stress is an old story; their intelligence is new and as boundless as the future. Already they’ve evolved to the point of elementary conversation, reading facial expressions and the like. Keep it simple and you can interact with them as with friends. Soon you won’t have to keep it simple. Before long they’ll have to keep it simple for you.

Robots seem a natural fruit of the human imagination. Drudgery as a grievance is as old as the leisure civilization gave us to deplore our lot. Who wouldn’t long for escape? Aristotle, as sober a thinker as ever lived, imagined in the fourth century B.C. an “instrument that could do its own work, at a word of command” — the only conceivable alternative, he said, to institutionalized slavery.

The word “robot,” more precisely “robota,” in fact, means “slave labor” in Czech. Czech playwright Karel Capek (1890-1938) gave it its modern meaning in his play “R.U.R.” (1920). By then the first proto-robots already existed, ripe for a catchy noun, and there it was. Time passed. A Westinghouse Electric creation known as Electro, 2.1 meters tall and weighing 1,200 kg, delighted crowds at the New York World’s Fair in 1939 by smoking cigarettes and blowing up balloons. How many of those present foresaw in it the solution to all humanity’s ills? Probably not many.

Even much later, it would have taken considerable foresight. By the 1980s, machine intelligence had evolved to the point where the question arose: Could a smart machine defeat a human chess champion? It finally happened in 1996, when IBM’s Deep Blue beat Gary Kasparov. Japanese shōgi, being more complicated, proved a tougher nut to crack, but in 2003 that barrier was also breached.

Games with rules are one thing. But real life demands more than mathematics and strategic thinking. Language, facial expressions, human nuances, even the simplest — those can’t be programmed, you have to learn them as you go along. That was the divide that — eternally, one would have thought — separated us from them, humans from machines. We could learn; they could not. But they can, and they are. By 2021, Shukan Diamond predicts, the first robot will pass the entrance exam to Tokyo University. Will we be sharing our schools with them too?

At the singularity, robots will be smarter than we are. Fifty or 100 years past that, they’ll be much smarter; in 200 years, maybe infinitely smarter. Will they take us with them, or leave us behind? Will they really, as Bryzek and others suggest, solve the problems that have dogged us, held us back and shadowed our happiness all our civilized lives? If so, what then? What will become of us? Will uselessness kill us off, as it did horses in the automobile age? Or will the solutions to old problems give rise to new problems, now unimaginable, at worst dooming us and at best keeping us busy doing what probably makes us happiest — solving problems?