Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

New video of Tesla crash demonstrates the problem of semi-automated driving

theautopian.com

Post reply on HN
Print view View original

121–130 of 714 posts

Re: New video of Tesla crash demonstrates the problem of semi-automated driving

#121

Post by gfodor »

nanidin wrote: post #98
gfodor wrote: post #81

Earlier quoted context omitted.

Not sure what you mean. If autopilot has a bug then there is a good chance a lot of other cars do too, since (afaik) autopilot is not at its core Tesla proprietary tech like FSD.

My point is that as an end user, there is basically no difference between autopilot and FSD if the FSD beta is enabled on the car. It's a technical detail that doesn't seem significant. If the end user said they were driving on FSD, it likely means they are enrolled in the FSD beta and were using autopilot - in the car there is no distinction between autopilot and FSD. They are the same thing in the car if the FSD be…

Sure my point had nothing to do with the user experience but the post-mortem on what to make of what actually happened and what needs to be done about it. In that context it is highly relevant. The person I was responding to seemed to be asking about which actual system was being used to navigate the vehicle.

Re: New video of Tesla crash demonstrates the problem of semi-automated driving

#123

Post by epivosism »

not2b wrote: post #100
epivosism wrote: post #74

Earlier quoted context omitted.

I'm interested in understanding why you would want every subdivision of "type of crash" to strictly improve, rather than being content that the combined crash rate goes down? i.e. if Tesla causes 100 new crashes of type A but saves 200 of type B, isn't that a gain? Even if type A crashes are one which humans don't normally make. A way in which I agree with you is if type B crashes are more avoidable by conscientious…

Is there any evidence that you can point to that Teslas are saving more crashes than they are causing, compared to some other vehicle with basic collision avoidance features?

I've googled and found some very partisan sites. It seems hard to tell what's going on.

My overall analysis is: Tesla has some very serious enemies in other automakers, lawyers, etc. And yet, there are no clear studies or statistics saying they're worse, and Tesla claims they are safer. This isn't a very good analysis and I'd love to have a clearer more well-defined group for statistical comparison, but it at least seems plausible to me.

Re: New video of Tesla crash demonstrates the problem of semi-automated driving

#124

Post by amf12 »

jayd16 wrote: post #82

This is an honest question here but why is the pile up so bad if driving conditions were ideal? The next 6 drivers were not keeping a safe distance or paying attention?

Honestly, when you're driving, you sometimes only see the car in front of you. You might see it slowing down first, and you start slowing down, or continue because there is enough distance anyway. Then the car in front of you stops abruptly or crashes. Now you panic and brake hard. Maybe you still crash into the car in front. Even if you stop before crashing, this hard braking causes the car behind you to crash into you.

TL;DR - Sometimes the 3 second following distance is just not enough even if someone is paying attention because they can only see the car in front of them.

Re: New video of Tesla crash demonstrates the problem of semi-automated driving

#125

Post by hermitdev »

nkrisc wrote: post #95
ASalazarMX wrote: post #69

Earlier quoted context omitted.

IMO the car signaled, changed lane and stopped without slamming the brakes. It didn't collide with anything, the pileup was formed by human drivers driving too close and too fast for their reaction time. If it had been a human driver that saw a kitten or a rock in the road, it would probably slam the brakes. I'm not defending Tesla's FSD, which I think is inferior for Tesla's reluctance to use LIDAR in addition to ca…

If it caused a crash it stopped too quickly. Unless there was an immediate , life-threatening need to stop it should simply not have stopped and continued until it was safe to stop. Any halfway competent driver could plainly see there is no safe place to stop there and you should never stop there unless your car is literally on fire or the road bed is missing.

It also stopped in a bad spot for human eyes. From the rear facing shot, looks like was just past a brightly lit to dark tunnel. Human eyes need a bit to adjust to that. It's quite possible, especially as the backup grew that people literally couldn't see the brake lights.

Thats not to excuse anyone, but it may have been a contributing factor.

Re: New video of Tesla crash demonstrates the problem of semi-automated driving

#126

Post by ClumsyPilot »

epivosism wrote: post #35

The question that is never answered in this kind of story is: 1. How many fatalities per mile does tesla FSD have, compared to its best alternative if we ban it? 2. Focusing on specific failure cases is not relevant since the political decision to allow it is all or nothing. 3. It'd be perfectly logical to accept road testing FSD, even if it's significantly worse in some areas, as long as this is not exploitable, AND…

> It'd be perfectly logical to accept road testing FSD, even if it's significantly worse in some areas, as long as this is not exploitable, AND the net gain from allowing it is still overall positive

So are all 20 manufacturers going to get the right to endanger my life just so they can train their self-driving? Or is this spesific to Tesla?

Do I and my family get some compensation for this danger, or in case we are killed? Do we get a share of the profit this self driving car will generate for Tesla ?

What if Tesla never produces a good self deiving car, goes bancrupt, then I died for nothing?

Re: New video of Tesla crash demonstrates the problem of semi-automated driving

#127

Post by MBCook »

amf12 wrote: post #118
Winsaucerer wrote: post #108

Earlier quoted context omitted.

> Have you watched the video? Its freeway speeds, a decently busy freeway, at a tunnel entrench that’s curving. Plus the car switched to the passing lane and abruptly stopped. The Tesla car switching and abruptly stopping was a big problem, and I can understand that first car having trouble, but for cars behind the first one they should ideally all be keeping a distance ready for suddenly hitting the brakes in an eme…

In the police report, out of the 8 cars, only 3 cars were noted to have been driving unsafe for the stopped traffic ahead. The rest of the cars were propelled into the car in front by the car behind them hitting them.

That makes a lot of sense. Doesn’t matter how careful I am if the person behind me only gave themselves 10 feet to react and stop.

Re: New video of Tesla crash demonstrates the problem of semi-automated driving

#128

Post by epivosism »

MBCook wrote: post #117
epivosism wrote: post #35

The question that is never answered in this kind of story is: 1. How many fatalities per mile does tesla FSD have, compared to its best alternative if we ban it? 2. Focusing on specific failure cases is not relevant since the political decision to allow it is all or nothing. 3. It'd be perfectly logical to accept road testing FSD, even if it's significantly worse in some areas, as long as this is not exploitable, AND…

So if an automaker can make a system that is capable of handling any driving task safely but occasionally ran itself off the road on purpose for no reason, as long as it’s a net positive in lives saved it should be allowed? I don’t agree with that assessment. It needs to make fewer mistakes than humans at all times , not just most of the time. Great with a disaster here and there is terrible.

I see where you're going, but how do you actually justify your choices?

A: occasional random suicide (100 deaths/year), no other deaths B: no random suicide, 10k deaths per year from normal driving

You would be saying you choose B?

It's not really about the exact numbers here - just trying to get on the same page - would you really weigh your uncomfortableness with "random deaths" highly enough that you would accept increased actual deaths?

Another possibility is that the random death situation really is a signal of "actually hard-to-detect badness". If so, it should come out in statistics and we should easily be able to come to agreement in that case.

But, if the choice really is between accepting "weird mistakes+lower deaths" or "status quo+many deaths" I feel pretty uncomfortable accepting the latter knowing the real-world consequences.

Post reply on HN