Via a friend, a video in which autonomous vehicles are already with us. And picking up fast food.
The thought that haunts me the most is that that human ethics themselves are only a work-in-progress. We still confront situations for which we don’t have well-developed codes (e.g., in the case of assisted suicide) and need not look far into the past to find cases where our own codes were dubious, or worse (e.g., laws that permitted slavery and segregation). What we really want are machines that can go a step further, endowed not only with the soundest codes of ethics that our best contemporary philosophers can devise, but also with the possibility of machines making their own moral progress, bringing them past our own limited early-twenty-first century idea of morality.
Building machines with a conscience is a big job, and one that will require the coordinated efforts of philosophers, computer scientists, legislators, and lawyers. And, as Colin Allen, a pioneer in machine ethics put it, “We don’t want to get to the point where we should have had this discussion twenty years ago.” As machines become faster, more intelligent, and more powerful, the need to endow them with a sense of morality becomes more and more urgent.
“Ethical subroutines” may sound like science fiction, but once upon a time, so did self-driving cars.
Below are some thoughts about this article I posted on Google+, reposted since I have a suspicion it’s a ghost town.
It’s somewhat fascinating that we’ll even have this problem, since we have no standardized “ethical subroutine” in our current social contract, even for driving. We generally just tell people to try not to crash. We have signs and rules, but I was never told in drivers ed which way to swerve when I encounter an oncoming vehicle in a narrow mountain pass in order to minimize loss of life (especially if that swerving meant self-sacrifice). Likewise, we have no “right” answer we teach in elementary school about how to solve the trolley problem. We’re ok with ambiguity and possible less utilitarian outcomes when humans do it, but not when our creations do it.
To take it a step further, suppose we developed an ethical subroutine that generally as a society we felt comfortable with for robot cars. Would we then enforce conformance with that subroutine on the remaining human drivers?
David Strickland, head of the National Highway Traffic Safety Administration, told an industry gathering sponsored by Swedish automaker Volvo and the Swedish Embassy in Washington,
“Automated vehicles offer an important and challenging method for reducing crash risk that we believes holds great promise,” Strickland said. He noted that human error was a factor in about 90 percent of the over 33,000 traffic deaths recorded in 2010. “We have the chance of … saving thousands and thousands of lives as” cars in use today are replaced with automated vehicles, he said.
Another interesting bit:
He declined to say when the government might propose safety standards for automated cars. Setting such standards would require the government to fundamentally rethink the way it evaluates auto safety, he said.
“We may have to depend on modeling and simulation of detailed traffic interactions that lead to crashes as opposed to the typical crash-testing model that we’ve used … over the past 40 years,” Strickland said.
Key questions will be whether the software in automated cars will be able to handle complicated driving situations and whether there will always need to be a human driver paying attention and ready to step in.
Also check out this story about Nissan implementing semi-robot cars without drivers even knowing about it. There will soon be a camera and computer controlling some motors in between the steering wheel you hold and the actual steering column.
Slate explores whether autonomy-loving Americans will be able to handle robot cars.
This American emphasis on the individual’s sovereignty poses a problem for new technologies designed precisely to deny personal agency. Autonomous technological agents—from military drones to the self-driving car—are increasingly prevalent. Their potential benefits and conveniences are immense. Yet as the currently cutting-edge becomes commonplace, these technologies could bump up against the prized American autonomy.
The United States drives more than any other society, and the self-driving car provides the glorious possibility of a hands-free cross-country road trip. But how will it harmonize with American drivers’ varied preferences for tailgating, conscientious speed-limit-monitoring, passive aggression toward walkway pedestrians, or highway-traversing pursuits of the fastest lane? General Motors, Volkswagen, Audi, BMW, Volvo, and Google are each currently testing driverless cars, with intentions to make the vehicle availablefor mass consumption by 2018. Recently, Nevada became the first state to pass legislation asking the Department of Motor Vehicles to formulate guidelines for driverless cars.
While engineers are perfecting the technology, they still must grapple with the drivers, who must both trust and enjoy the automated-car experience. Making a driver-free car safe and effective requires overlooking the uniqueness of each individual’s driver personality. Research suggests that autonomous technological agents like service robots and anthropomorphic computer interfaces can diminish users’ experiences of control. And we hate to give up control.
It is uncanny how many of these robot car issues have already been explored so thoroughly in Total Recall. (Start at the 17 second mark)
MN/DOT is conducting a process, which they’re calling Minnesota GO, to develop a 50-year vision for all types of transportation in the state . I was asked to present in April at an advisory group meeting, and I talked about 50-year trends from a planner’s perspective.
The advisory group video page has some other great presentations from folks who more accurately fit the title of “expert” like Michael Noble on energy and Michael Huber on health. The video interview page also has some good ones, including David Levinson and Frank Douma. While you’re at the site, don’t forget to put in your two cents.