1
$\begingroup$

From self-driving cars, it's a well known fact that the public doesn't like the technology. According to an article in the New York Times, the Waymo car was attacked many times by angry humans who don't seem to like modern technology.[1] The first impression is, that this kind of anti-technology harassment is a typical sign of the neo-luddite movement who are against technology, especially if it has to do with Artificial Intelligence and robots.

On the other hand, it's surprising to see that no such attacks are reported against self-driving railway trains. Instead, most customers are happy with autonomous railroad vehicles.

quote: “Approximately three-quarters of the respondents (77.6%) can generally imagine using autonomous public transport regularly in the future.” [2]

Can this difference in public perception be explained somehow, especially under the constraint, that the software for controlling autonomous vehicles is in both cases the result of Artificial Intelligence research projects.

[1] Wielding Rocks and Knives, Arizonans Attack Self-Driving Cars, https://www.nytimes.com/2018/12/31/us/waymo-self-driving-cars-arizona-attacks.html

[2] Pakusch, Christina, and Paul Bossauer. "User Acceptance of Fully Autonomous Public Transport." ICE-B. 2017.

$\endgroup$
  • $\begingroup$ Can this be linked with the fact that in driving a train the enviroment is somewhat more "well-defined" than driving a car? I mean, you are literally rail-roaded. Perhaps people feel more sure and less "de-humanized" by this. $\endgroup$ – olinarr Oct 8 at 14:02
  • $\begingroup$ Trains take linear/semi-linear paths. Cars don't. $\endgroup$ – Tautological Revelations Oct 9 at 9:03
  • 1
    $\begingroup$ I've added the "social" tag because this question involves public perception of (and response to) AI applications. $\endgroup$ – DukeZhou Oct 10 at 20:42
-1
$\begingroup$

I'm going to use statistics from the US, where Waymo is based. Statistics from Europe would probably be a bit different, but I doubt they would be so different as to change the conclusions.

People commit crimes like attacking property because they have a motive and an opportunity to do so, and because they believe their attack will have an effect. If people lack a motive or an opportunity to have an effect, they will probably not attack a autonomous vehicle.

People have a much weaker motive to attack autonomous trains than autonomous cars, because autonomous trains threaten the employment of fewer workers. Consider:

  • Driving vehicles, especially heavy vehicles, is an extremely common form of low(er) skill employment, especially in the USA where Waymo. America employs around 2 million operators of these vehicles. Companies like Uber and Lyft likewise employ several million people. Like Cotton spinning in the past, driving is this a common occupation, and one that is disproportionately pursued by people who lack other marketable skills. Technologies that threaten driving as an occupation will thus be more likely to produce a backlash.
  • In contrast, trains employ just about 35,000 people in the US as operators. Other jobs related to trains, like maintenance workers, and conductors, might actually increase in number if trains are automated. Train engineers may have more skills, and thus more options for alternative employment, than truck drivers do.
    • Further, people generally do not have to share their local environment with autonomous trains (or at least, no more than a regular train), whereas many drivers may have to directly interact with autonomous cars. If these interactions generate friction, people may become more aware of autonomous cars, and more hostile towards them.

I think people also lack the opportunity to attack trains effectively.

  • Trains are also much larger than cars, so an ad-hoc attack is much less likely to do visible damage. You can cause serious damage to many parts of a car with a baseball bat in a few minutes. A train takes many times longer to deface, and is usually built of sturdier materials. A planned attack is much more complex to execute.
  • Train tracks are usually kept separated from the general public with fences and other infrastructure, especially when the trains are not in motion. You don't just walk out into the street and pass dozens of potential trains to attack, as you do with cars.
  • Trains almost always have other people or infrastructure protecting them. These could be security cameras, security guards, staff who tend to parts of the train other than its operation, or passengers. Cars are much more likely to be left in unsecured areas that no one is watching. All of this means you are much more likely to be detected and caught if you attack a train than you are if you attack a car.

So, I think trains present a weaker motive (because they cause less unemployment) and a weaker opportunity (because they are harder to reach and harder to damage) for people to attack. Since these factors are weaker than for cars, trains are attacked less often.

$\endgroup$
  • $\begingroup$ It's correct, that truck drivers don't need social skills on the freeway because the speed is fixed. The situation is different in the downtown area. To navigate a truck in small streets in the city, the human driver needs a lot of communication skills, and sometimes the language is different from English. $\endgroup$ – Manuel Rodriguez Oct 9 at 11:14
  • $\begingroup$ @ManuelRodriguez I'm not sure what you're getting at with social skills there. Truck driving is a low(er) skilled occupation whether you need social skills or not. Specifically, it requires less training then operating a train. $\endgroup$ – John Doucette Oct 9 at 12:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.