>(cf. Volkswagen emissions scandal)

I'm not sure that was the best example for your point, the few untrustworthy people in Volkswagen did cause cars to not work as advertised. That's an example for not trusting any of their products anymore.

Look, it wasn't everyone in the company. Just the CEO, a team of product designers, some developers, the entire management chain between the coder and the CEO, and maybe some people in Q&A.

That's no reason to paint the entire company with a broad brush.

Sure, I brought that up for a reason. Despite some underhanded maneuvering and cheating, the emissions-compromised cars still passed independent safety tests and I haven't heard of anyone being killed by them.

1,200 people will be killed in Europe by these cars:

http://news.mit.edu/2017/volkswagen-emissions-premature-deat...

They won't be killed by the cars physically running them over or otherwise going out of control, true, but that's because the untrustworthy people chose to be untrustworthy in emissions standards and not in handling or self-driving algorithms or anything. You can't rely on them always making that choice, because they're untrustworthy.

> The researchers [examining the health impact from the 2.6 million affected cars sold in Germany] estimate that 1,200 people in Europe will die early, each losing as much as a decade of their life [edited for brevity]

I have no doubt that the extra pollution is doing health damage. I strongly want more air pollution regulation. I don't trust Volkswagen, and I had one of the super-polluter TDIs, and sold it back to VW.

But this claim of 1,200 premature deaths by up to a decade seems goofy. If VW's 2.6M cars in Germany alone were taking a decade off anyone's life, it seems like we'd all be dead already. The world has over a billion polluting motor vehicles, and we're not even talking about airplanes, fires, factories, natural sources of NOx, or energy production. Car emissions are a minority of the total.

For sure, lots of people are getting sick from pollution in Europe and globally, but I wish the authors had stuck with person-years of life rather than try to state it as a specific number of people affected. Whatever actually happens, assuming it was even possible to pinpoint VW, which it's not, the probability that number of people affected in Europe is near 1,200 is almost zero. So, I guess at least consider calling it what it is: someone's estimate.

Sure, they passed a safety test and didn't run anybody over because the emissions monitoring system doesn't affect the handling of the car

Is there any reason whatsoever to believe that the same kind of wilful dishonesty around monitoring cars' ability to choose the appropriate speed, direction or reaction in certain road situations wouldn't result in fatal accidents?

...and every Jihadi is our government's wet dream to use as propaganda for endless war... that generates Jihadis. Think about the big picture next time.

Unless they are tightly locked down by the manufacturers. Perhaps jihadis will not be able to hack them in such a way as to take remote control, and perhaps the manufacturers will eventually make the automatic not-plowing-into-pedestrian system override driver controls.

Yes, no system is perfectly unhackable, but since terrorist cells tend to be small and relatively isolated I could see it being the case that the people who want to use cars as remote weapons would not be people who have the time and talent to make that happen.

State-sponsored groups may be a different story of course.

> Perhaps jihadis will not be able to hack them in such a way as to take remote control.

That is wishful thinking. Islamic State has been very successful in recruiting and radicalizing young english speaking hackers in Europe -- mostly in Britain (children of Pakistani immigrants), Belgium (2nd generation Moroccans) and France (descendants of french-algerian immigrants).

Here's one that I know of. British-Pakistani hacker Junnaid Hussain => See "Here’s how the military tracked down and killed the top hacker for ISIS" http://www.businessinsider.com/isis-hacker-trick-found-2016-...

> Hussain, a high profile recruiter for ISIS who hacked CENTCOM’s Twitter and YouTube accounts, was considered the second ranked British jihadi behind Mohammed Emwazi (Jihadi John). Hussain was killed in a U.S. airstrike in Syria in August 2015.

Source: http://www.foreigndesknews.com/world/middle-east/isis-hacker...

Junaid was know for his hacking skills, he used to publish "Kill Lists" https://en.wikipedia.org/wiki/Junaid_Hussain

So, he essentially tricked some social media manager into giving him a password (probably the same used for both). That's even weaker than I expected.

Read you some Kevin Mitnick. 'Social engineering' as he calls it is very effective as an intrusion technique.

I don't understand it but a large percentage of fundamentalist terrorists were engineering/science majors in college, so I wouldn't count on them not having expertise. Also batting for non Islamic homegrown terrorism, Timothy McVeigh was a programmer in his teens, and the Unabomber was a child prodigy, starting college two years early.

http://www.nytimes.com/2010/09/12/magazine/12FOB-IdeaLab-t.h...

Anyone care to explain the down votes? The threat of Jihadis using self-driving cars is a very real one.

Recently a plot was unearthed , and thwarted, in London. A state sponsored actor with ties to ISIS branch in Libya was trying to start a self-driving car rental agency.

why would they want a self driving car, and not just a car?

Not OP, but I think the idea is that the Jihadis no longer need to "sacrifice" their "martyrs" in car suicide bombing missions.

They could - in theory - pack a self-driving car or worse, a truck , with explosives and remote control drive it into a crowd of "infidels".

Rinse repeat in various cities.