Complications of Letting A.I. Take the Wheel

2016-06-29-terminator-car

Self-driving cars will likely soon be a common sight on roads throughout America and the rest of the world. But as these autonomous vehicles become a reality, their introduction to the market and to real-world driving situations raises safety and ethical questions.

The extent to which self-driving cars will fully replace human-driven ones remains unclear. While it is easy to imagine self-driving cars remaining nothing more than a luxury for higher income brackets, at least at first, a new survey apparently makes the opposite assumption, asking respondents if they “believe that driving manually is a luxury that must be preserved.”

A considerable majority (72%) said that the “luxury” of driving should be preserved, and an even larger percentage (79%) said that car manufacturers, not autonomous car owners, should be held responsible in the case of accidents.

But aside from concerns about which humans should be held responsible when self-driving cars kill people, another ethical question that has recently come up is who the cars should kill, when they inevitably have to make the choice. Should they be programmed to protect their passengers in case of an accident, or to protect innocent bystanders at the passengers’ expense, and how should they make that decision?

It’s difficult to find a satisfactory answer. There are other unanswered questions about self-driving cars, however, that deal not so much with ethics, but with basic issues of safety. The technology in cars today is easily hackable, concerns are currently widespread regarding the security of the so-called Internet of Things, and as cars become increasingly connected as they move towards driverlessness, concerns of killer car hacks are understandable.

In an article about the prospect of hacking self-driving cars vs. hacking the connected (but people-driven) cars currently on the road, Kevin Collier makes a compelling argument that car makers will actually get better at cybersecurity as their cars become driverless and the need for security becomes more obvious and integrated into the design and manufacturing process.

Collier notes, though, that “consumers have good reason not to trust that car manufacturers take your safety seriously. The industry certainly has made gross calculations of human life—just look at the times when it waited to recall a faulty part until enough people died that a recall would be cheaper than paying wrongful death lawsuits.”

If there is good reason to be skeptical about whether manufacturers are doing enough to secure their self-driving cars, there is also reason to be wary of government as it regulates this new technology, and of how intelligence and cyber-warfare agencies within governments, or malicious actors within these agencies, might manage to exploit it.

In 2013, journalist Michael Hastings died in what many deemed a suspicious a car accident – for reasons including the physical circumstances of the crash, Hastings’ discussion with his neighbor that he suspected the car had been tampered with, as well as his long list of enemies. And another reason that many suspected foul play came in the form of statements made by a former high-ranking Bush administration counter-terror official in an interview with the Huffington Post.

“I’m not a conspiracy guy,” said Richard Clarke, former U.S. National Coordinator for Security, Infrastructure Protection and Counter-terrorism. “But my rule has always been you don’t knock down a conspiracy theory until you can prove it (wrong). And in the case of Michael Hastings, what evidence is available publicly is consistent with a car cyber attack. And the problem with that is you can’t prove it.”

As researchers discuss the A.I. ethics of choosing who driverless cars will kill in the leading scientific journals, those with less concern for ethics – but perhaps an interest in killing under cover of plausible deniability – will no doubt be watching developments in the autonomous vehicle field, and taking careful notes. Those who oppose the well-known ethical flexibility of elements within the latter group might be wise to do so as well.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s