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

A comparison to Waymo’s auto liability insurance claims at 25M miles

waymo.com

Post reply on HN
Print view View original

251–260 of 497 posts

Re: A comparison to Waymo’s auto liability insurance claims at 25M miles

#251

Post by mtts »

usrusr wrote: post #147
Brian_K_White wrote: post #93

I just heard someone mention in a podcast how as a cyclist they need to make eye contact with drivers to be safe, and there is no one to make eye contact with. Waymo would say "don't worry, it sees you" and maybe the numbers even show they do some percentage more than the average human driver, but that still does not reproduce the self-help aspect of you looking at a driver until you see that that driver sees you. I…

As a cyclist I avoid eye contact like the plague: it's only a small minority of drivers, but that small minority is dangerous enough to outweigh all benefits eye contact might have with others. Some drivers will inevitably read eye contact as "the cyclist has seen me and understood that my big and dangerous car will crush them without taking harm so they will yield no matter what traffic rules might say, so I will go…

Not just as a cyclist. As a pedestrian in one of those huge Asian metropolises with absolutely bonkers traffic, you should never ever make eye contact either. It’s the only way drivers will stop for you.

Re: A comparison to Waymo’s auto liability insurance claims at 25M miles

#253

Post by bmicraft »

Aeolun wrote: post #118
Brian_K_White wrote: post #93

I just heard someone mention in a podcast how as a cyclist they need to make eye contact with drivers to be safe, and there is no one to make eye contact with. Waymo would say "don't worry, it sees you" and maybe the numbers even show they do some percentage more than the average human driver, but that still does not reproduce the self-help aspect of you looking at a driver until you see that that driver sees you. I…

> Maybe they will prove to be so reliable that you don't need to worry about them. You can just reliably assume and predict their behavior as animate objects about the same as you do for inanimate objects. This seems to be what is happening. There was a comment here by some user not too long ago that said they deliberately drive in front of the Waymo’s because they trust them to slow down.

You can also do that with human drivers and it will work for a while. The question is how often it doesn't work. A single person driving in front of a Waymo a couple of times can't determine that with any statistical significance.

Re: A comparison to Waymo’s auto liability insurance claims at 25M miles

#254

Post by mentos »

Brian_K_White wrote: post #93

I just heard someone mention in a podcast how as a cyclist they need to make eye contact with drivers to be safe, and there is no one to make eye contact with. Waymo would say "don't worry, it sees you" and maybe the numbers even show they do some percentage more than the average human driver, but that still does not reproduce the self-help aspect of you looking at a driver until you see that that driver sees you. I…

I feel like I saw a self driving concept in the last few years where the headlights had eyes that could appear to be looking at you to signal that it was aware of your presence?

Re: A comparison to Waymo’s auto liability insurance claims at 25M miles

#255

Post by mulmen »

notahacker wrote: post #178
mulmen wrote: post #145

Earlier quoted context omitted.

Then shouldn’t the comparison be to Uber instead of an ambulance?

Ambulance drivers might be a step too far unless autonomous systems are expected to drive ambulances at high speed and run red lights. But my point is that they really do need to be a lot better than the median driver given their real world operational expectations, and especially significantly better than the mean driver they're being benchmarked against here, who probably loses their license for the bodily injuries…

Well Uber drivers don’t have any special qualifications. That’s kinda the point of Uber. Is there any reason to expect them to be safer than the average driver?

Re: A comparison to Waymo’s auto liability insurance claims at 25M miles

#256

Post by Mawr »

LoLFactor wrote: post #149
Brian_K_White wrote: post #93

I just heard someone mention in a podcast how as a cyclist they need to make eye contact with drivers to be safe, and there is no one to make eye contact with. Waymo would say "don't worry, it sees you" and maybe the numbers even show they do some percentage more than the average human driver, but that still does not reproduce the self-help aspect of you looking at a driver until you see that that driver sees you. I…

This. I'm always surprised how few people know as pedestrians to look at me, the driver and not my car. Some don't even look at the car. I developed a habit of looking for their gaze and if they don't look back, assume they're not fully aware and just am more cautious. This works because I, as a human, know this and can compensate when they just rush the crosswalk without being fully aware of their surroundings. How…

It is not possible for me to see you in your car with any sort of reliability; even if I could the benefits are dubious and again, unreliable.

Looking at your car is all I need anyway - I can tell if you've seen me by your behavior, you're either slowing down to yield to me or you're not. If you're not, the only possible outcome of knowing you're seeing me is being misled into stepping into your path of travel.

Re: A comparison to Waymo’s auto liability insurance claims at 25M miles

#257

Post by mulmen »

necovek wrote: post #247
mulmen wrote: post #231

Earlier quoted context omitted.

Ah I see, like merging on to a freeway. In Washington the law is actually to zipper merge (take turns) so I would expect Waymo to do that. It’s a good question and I’m not sure what Waymo would do there. Based on my limited experience I expect it would navigate that situation correctly. The bias is to go (safely), not wait for perfect conditions.

Thanks for the term as well ("zipper merge") — that's what I was referring to. Yeah, I really wonder how it'd fare in that case while differentiating between "there are just 3 cars there, no need to yield" and "it's an infinite inflow of traffic, better yield to one car"?

My Waymo ride was in LA. They aren’t operating on freeways there yet. Maybe this is one of the challenges.

Apparently they do operate on freeways in San Francisco so maybe someone who has taken a ride there can share their experience.

Re: A comparison to Waymo’s auto liability insurance claims at 25M miles

#258

Post by Mawr »

david-gpu wrote: post #186
necovek wrote: post #174

Earlier quoted context omitted.

It's probably equally simple for a machine (if not simpler) to figure out if a human has looked at the oncoming vehicle with sufficiently advanced cameras and computer vision. However, we use a lot more hints (change of pace, facial expressions that indicate presence and focus, nodding etc), which, while not outside the realm of AI CV, would surely need lots of training, yet it comes pretty naturally to humans.

The problem is not whether the machine can see the human, or even whether the machine knows that the human has seen the machine. The problem is that the human currently has no way to know whether the machine has seen the human.

Of course there is, the behavior of the machine should be the same as of a driver - the car shaped object starts to slow down in a way that will make it stop before hitting me at the pedestrian crossing.

Re: A comparison to Waymo’s auto liability insurance claims at 25M miles

#259

Post by jpgvm »

amelius wrote: post #246
jpgvm wrote: post #42

Earlier quoted context omitted.

Why? We aren't looking at introducing autonomous ambulances any time soon. They should be compared to taxi/Uber/Lyft drivers that is what is being replaced by this and as such where the net harm increase or reduction will occur.

Why -> because if someone caused an accident I'd want to look them in the eyes and call them names; as opposed to calling a support number. If these vehicles were among the best drivers, at least I could live with it maybe. Emotions are real, you cannot always rationalize them away. Average drivers sometimes end up in jail for their mistakes. If Waymo made those same mistakes, guess what will happen.

I'm not sure it's possible to solve for that with additional prowess.

If you are wronged by a machine you are always going to be emotional about it, i.e having your email account unceremoniously terminated by Google without recourse by an essentially automated process.

The thing to solve for is the lack of recourse. Very very difficult in American society given how skewed the relationship is between corporations and people (or the government for that matter).

If instead when a Waymo demonstratively caused an accident the payout was 10-100x higher than if a human was involved I suspect many peoples qualms would go away.

While I think just the "knowledge" (assuming Americans would accept facts...) that the driver is X% better than the best drivers on the roads would provide little comfort to most.

I think in societies outside of America these may be more successful actually. i.e most parts of Asia.

Re: A comparison to Waymo’s auto liability insurance claims at 25M miles

#260

Post by MyFedora »

CalRobert wrote: post #46
digitcatphd wrote: post #17

Good to see studies of incidents in comparison to human drivers. Historically what I have seen is 'self-driving car crashes into X, which seems to be concentrating exclusively on outliers to prove a point.

Funny enough, when drivers hit people that’s _also_ reported as “car crashes in to x”

Ah, the classics:

"Pedestrian involved in accident"

"Cyclist collides with car"

"Car brushes against cyclist"

"Vehicle loses control"

Amazing how responsibility shifts when a two-ton machine is involved.

Post reply on HN