TOKYO -- At the 2015 Tokyo Motor Show in late October, Japanese automakers Nissan Motor and Toyota Motor showcased concept cars with autonomous driving technology, hoping to put them in showrooms by around 2020.
The cars are evocative of "2001: A Space Odyssey," a 1968 Stanley Kubrick film based on the novel by Arthur C. Clarke. One of the themes of the movie is the relationship between people and machines. Five astronauts are chosen to explore Jupiter in spacecraft controlled by HAL9000, a thinking machine.
Fourteen years later, humans are not yet exploring Jupiter. But the less ambitious challenge of putting self-driving cars on the road in large numbers is within reach. To make that happen, two obstacles -- one technological and one legal -- will have to be overcome.
The Geneva Convention on Road Traffic, an international agreement that took effect in 1952, requires any vehicle whose operation has the potential to threaten human lives to be controlled by a human driver or pilot. This is one reason why commercial flights are operated by two pilots, according to one aircraft maker, although the technology exists today to fully automate planes.
Who's behind the wheel?
So what are the prospects for having the world's 900 million cars drive themselves? The benefits are clear. Autonomous vehicles would give mobility to people who cannot drive due to age or disability. As the technology moves forward, the regulatory issues are being debated in the auto industry.
Ray Kurzweil, an American inventor and expert on artificial intelligence, predicted in his 2005 book "The Singularity is Near" that artificial intelligence will be much more powerful by 2045 "than all human intelligence today." In other words, most jobs that can be done only by humans at present will be done by intelligent machines or other types of technology in the future.
In that sense, the most visible manifestations of advanced technology today -- computers that outperform humans on TV game shows, autonomous cars, the Internet of Things -- could be a prelude to the singularity -- self-improving, super intelligent machines. It is unclear if or when this may happen, but many futurists predict a turning point within the next 30 years.
A really smart car
Cars may start thinking for themselves sooner than other kinds of machines, according to a senior engineer at Nissan, which seemed closer to realizing a self-driving car than its rivals with the IDS concept car that it showed off at the motor show.
The Nissan engineer predicts the singularity for cars could come as early as in 2020. One major driver of development in this area is rich tech companies like Google and Apple. Japanese and Western automakers are hard at work as well.
Just because it is technically possible to take human drivers out of the loop does not mean it will happen anytime soon. In the case of self-driving vehicles, the question of liability looms. Who should be held responsible when accidents occur? About 90% of fatal road accidents are caused by human error. If autonomous cars are designed to drastically reduce or eliminate this source of error, it would, perhaps, be difficult to hold the owner of the car or the passengers responsible for a crash. Instead, the car's manufacturer might be liable.
It is unclear whether and to what extent the "operator" of the car should be held responsible for the actions of an autonomous vehicle. There is also the question of privacy. More autonomous cars on the road means more data being gathered by automakers and police on the operator's whereabouts.
"A trade-off will arise between privacy and safety," said Susumu Hirano, a professor and legal expert on product liability and privacy issues at Chuo University in Tokyo. "Society will have to build a consensus."
Humans may soon face a future where machines can do most of the jobs people do, only better. In 2001, HAL suddenly turns hostile and kills the crew. Are we really prepared to put our lives in the hands of machines, however smart? As technology advances, that question is likely to become ever more urgent.