Hacker News Reader

Comment by caddemon | Hacker News Reader

Yes drunk driver is not the right bar, but I think he still has a point. If the accident rate is less than overall human accident rate we should want to use the self driving cars, but the accident rate might still be high enough that a car company wouldn't want to claim 100% liability for every single accident.

Car accidents are one of the top causes of premature death in the US. Self driving cars aren't there yet, but when they are an improvement on human driving they should be encouraged, even if the absolute numbers are still not amazing.

View on HN

who do you think should take the other portion of responsibility? Certainly not the person sleeping in th back seat with no control?

Well, if they're the ones who are nominally in control of the self driving car, because they activated it and are occupying the car, I would say that they have a portion of the responsibility.

Then we're back to the Tesla model - you don't have a self-driving car, you have a car that claims it is such, but a human needs to be 100% aware and in control at all times.

Honestly, I think that's fair. If you are the one who pushes the button and sends the device out into the world, even if you didn't program it, I think you deserve some part of responsibility of it malfunctioning.

My biggest problem with this approach is humans are terrible at paying attention to things that don't demand 100% focus.

A car that drives itself 99% of the time, but fails catastrophically the other 1% is doomed to failure. The human operator won't be engaged enough to take over that 1%. At least not without airline levels of squawks and beeps and wheel shakers and even that might not be enough - airlines don't have to worry about children chasing balls into busy streets, etc. And the pilots are highly trained - drivers are not.

I would think for a self-driving car to reach an acceptable level, it should be like a train. If I get on a subway or a bus and it hits someone, I am not held responsible, even if it stopped specifically to pick me up first.

worth noting the manufacturer doesn't have complete control over the car either, once it's "in the wild". maintenance is an important and frequently overlooked part of driving safety. a sophisticated ai driver (or human for that matter) can't do much with a set of bald tires or bad brakes.

Taxis are liable for their accidents.

The problem is that it's a different set of people who will die with self driving cars. Imagine the trolley problem but instead of 5 vs 1 it's 30,000 vs 29,000. The 30,000 people get hit by the trolley and die if you don't pull the lever, and a completely different set of 29,000 people die. People who did not do anything wrong. People who had no chance to fix the situation. Sure, it's 1,000 fewer people, but none of those 29,000 people did anything wrong. They were all killed by the self driving systems made by car manufacturers.

Regardless of whether or not you personally would pull the lever, you have to admit it's not the same as saving 1,000 lives.

I just want to point out that drunk driving also kills innocent bystanders: pedestrians, cyclists, sober drivers around, etc.

This is completely absurd reasoning. If we knew who was going to die, we'd let them know beforehand, so they could pick a safer spot on the street at their "death time".

There's no such thing as a deterministic future, as you suggest.

We can't stop doing things because a theoretical future will be changed. This is absurd.

While it is a whiff, one grain of truth is that its possible that pedestrian deaths increase with self driving vehicles, even if overall deaths decrease.

Is it fair to trade lives of people who choose not to drive a car at all for those who do? As of now NHTSA weighs weighs them identically which I think is a huge mistake.

That said its hugely dependent on the technology involved. I think a Waymo is safer than your average driver already for pedestrians. FSD Beta if unmonitored would be drastically worse than your average driver for pedestrian safety.

I'm not suggesting it's deterministic.

People have been saying that if driverless cars are safer than humans by any amount we should allow them. I'm saying they should be much safer than humans before we allow them.

The public will not tolerate car manufacturers killing 10's of thousands of people per year just because they are very slightly statistically safer than the average human driver.

Reply on news.ycombinator.com