I agree with #1 and #2, and support efforts to keep the probability of such events close to (but not equal to) zero.
You’re entitled to your preferences regarding the manner of your death, but I would accept a slight increase in risk of being murdered by a hacker if it meant a vastly decreased risk of being negligently killed by some jackass texting on his phone behind the wheel.
Besides, as Ravenman pointed out upthread, by driving a current model car, you have already accepted a non-zero risk of being hacked.
Going in a slightly different direction, I wonder what the consequences of an automated road/car system would or will be on car ownership.
How many people will be willing to forgo a few minutes of time and convience, as well as letting go of the enormous resource drain of car ownership once car automation reaches some level of saturation.
I see some add on effects, grocery shopping and delivery by remote, it will certainly change the nature of the pub crawl, will it be thhe death, revival, or no impact at all on the waning drive-in theater industry? Things like that
Computers not only don’t draw a salary, they don’t get tired, they don’t get distracted because their wife left them, they never get bored, and they don’t have to be transported to wherever the flight is leaving from. They can work 24/7, and they never go on strike.
They will never be perfect. Eventually, they will be better than humans, on average.
The issue is the same as always - people are bad at estimating long-term risk. They are more afraid of things with which they are not familiar, and in situations when they don’t feel any sense of control. That’s why people are afraid of nuclear reactors, and not swimming pools, even though swimming pools have killed more than a few people and nuclear reactors have not.
Or a self-driving car, even if it has a better safety record than your average Uber driver.
But they do cost money. In the case of the computers that might replace airline pilots, they cost a LOT of money.
They do wear out, become outdated, get broken and have to be replaced.
True… for now.
Actually, I have seen replacement parts flown in to fix a broken airplane. At great cost. I once needed a computer component replaced in a jet I was flying - it was transported, installed by a technician and all-in the expense was way, way beyond what a senior captain at a major airline makes in a year.
This response is not to quibble the details, but to make the point that the time it would take to achieve actual cost savings by replacing human airline pilots takes some figuring. That kind of automation is not cheap, nor is the expertise to keep it working. That said, we already have a lot of expensive stuff in airplanes and people to keep them going. So this is another reason I think achieving full automation in cars might be a better goal at this point in time.
BREAK (meaning, I’m responding to other posters now)
I think some people are construing “blame” for accidents in the personal sense, not the legal sense. From what I’ve read it’s a very thorny legal problem sorting out who or what entities might be responsible when automation goes awry and causes damage or death. That’s a different problem than people literally having someone to point a finger at.
That’s certainly true. OTOH, I work in IT for an airline, and the computer systems to support the pilots and flight crew already cost a lot. So do I.
So do people.
Also quite true. But we will get there eventually. I don’t know if self-driving cars will be first, but there will be synergy. Once we figure out how to drive a car with a bot, that will help figure out how to pilot an airplane with a bot. We will still probably require there to be a human pilot for backup long after the bot can do better overall, if for no other reason than the pilots’ union will require it, and that is unlikely to be the case with cars, but eventually it will be cheaper/safer/more efficient.
