When robot history comes to be written, April 2009 will occupy a prominent place. Future robots will look back, perhaps with pride, at the events of this month. A robot has been created that has, for the first time, independently advanced scientific knowledge.
We’ve already got robots that build cars, robots that look like women, robots that help around the house and robot pets. Soon we’ll have robots that can baby-sit and nurse us. But robot scientists? It’s a moment that science-fiction writers have anticipated for years, and it’s exciting to see it happen in real life.
In fact, the news of the robo-scientist will encourage those who believe — and those who hope — that the next great turning point in human history is almost upon us.
What happened is that (human) scientists at Aberystwyth University and the University of Cambridge in Britain designed a robot able to carry out each stage of the scientific process — thinking, planning, assessing and testing — automatically, without the need for further human intervention.
Using artificial intelligence, the robot hypothesized that certain genes in baker’s yeast contain the instructions for making specific enzymes, which drive biochemical reactions in yeast cells. Beforehand, the function of those genes was not known.
The robot, given the predictable name of Adam, devised experiments to test its ideas, ran the experiments using laboratory robotics, interpreted the results and repeated the cycle. It ended up by “announcing” the function of the genes. Adam is the first machine to have independently discovered new scientific knowledge.
Ross King, who led the research, said that if science was more efficient it would be better placed to help solve society’s problems. “One way to make science more efficient is through automation,” he said. “Automation was the driving force behind much of 19th- and 20th-century progress, and this is likely to continue.”
King’s comments will thrill those futurists who talk about “technological singularity” — the next leap forward in progress. The American futurist Ray Kurzweil, among others, predicts that will occur when superhuman intelligence is created. The demonstration that a robot has reached the level of a graduate student in designing experiments and testing hypotheses suggests this could happen before too long.
If you look at human population size, you can measure economic growth. Each time there was a technological advance, the world economy grew. Kurzweil and others have plotted the changes in economic growth throughout history, starting with the Paleolithic period(c. 2 million to c. 10,000 B.C.), when stone tools were used.
What they have found is that the world economy doubled every 250,000 years until the Agricultural Revolution took off about 12,000 years ago. Then the economy doubled every 900 years.
However, starting with the Industrial Revolution just over 200 years ago, the world economy started doubling every 15 years. When the revolution caused by superhuman intelligence — the singularity — kicks in, the economy is predicted to double at least every few months, and perhaps every week. More to the point of this column, the line between human and machine will be blurred and we will enter the age of the “transhuman.”
What this will mean for natural selection is something I’ll have to come back to. For now, I want to end by thinking about how the idea of machine intelligence has been considered in Japan and the West.
The word “robot” comes from a Czech word meaning something like “drudgery,” “forced labor” or “hard work.” The word, introduced in a 1921 play by Karel Capek titled “R.U.R. (Rossum’s Universal Robots),” captures the idea that the machines may be being exploited by humans — that they are our slaves. That feeling, and the fear that the slaves will rise up and destroy us, is at the root of the portrayal of robots in many Western films and books.
In Japan, of course, robots are kawaii (cute). They are trusted to make sushi as well as cars, and to take care of the elderly. It has been said that the acceptance and even love of robots in Japan — compared with the West, where they are often portrayed negatively — is down to Shintoism and its lack of spiritual distinction between animate and inanimate objects.
In many ways it’s just as well that robots are so embraced by Japanese people. Given the graying population and the strict laws on immigration, before long there won’t be enough humans to do the work in the country. The Japanese government certainly seems to be planning for this sort of a future. In 2007 the trade ministry announced plans for1 million industrial robots to be installed in factories across the country by 2025.
Since one robot can replace at least 10 humans — and robots don’t have to stop for weekends or to sleep at night — a million-strong robot workforce could make a big impact on the economy. Put like this, you begin to see how the world economy could start growing really quickly.
Kurzweil says robots will reach human-level intelligence by 2029. I can’t wait to see it. But I can’t help wishing that the word for the automatic machine had been coined by someone in Japan, so we wouldn’t have the word “robot.” Perhaps then we wouldn’t worry about drudgery and slavery so much.
What Japanese word might work? The craftsman Hisashige Tanaka (1799-1881) built mechanical toys known as karakuri, some of which could write kanji characters. Others could serve tea, or fire arrows.
Karakuri means something like “a mechanical device to take people by surprise,” which could still have a sinister meaning, or it could mean “an object that delights and baffles and moves with mysterious power.”
Karakuri might have been a better word than robot, but after a bit of a think I came up with jidoki, meaning “automatic-motion spirit” — though I won’t hold my breath waiting for it to catch on.
The second volume of Natural Selections columns translated into Japanese is published by Shinchosha at ¥1,500. The title is “Hito wa Ima mo Shinka Shiteru (The Evolving Human: How New Biology Explains Your Journey Through Life).”