S.F. says incidents by Cruise, Waymo driverless taxis are ‘skyrocketing.’

https://www.sfchronicle.com/sf/article/driverless-taxi-cruise-waymo-18157172.php

by mikhael • 3 years ago

388 875 3 years ago

carapace2• 3 years ago

First, where is the data?

Second, yeah, it's bananas to test large, fast, heavy robots on public streets. The only reason we let it happen is because the robots look like cars.

Third, it's bananas to have large, fast, heavy vehicles everywhere mixed in with all the other traffic. The only reason we let it happen is a mass-marketing campaign (see "The Real Reason Jaywalking Is A Crime" (Adam Ruins Everything) https://www.youtube.com/watch?v=vxopfjXkArM )

Here is a film "a Trip down Market Street" recorded in San Francisco on April 14, 1906, just before the Great Earthquake. The source of the problem is clear: cars accelerate much faster than horses. https://www.youtube.com/watch?v=VO_1AdYRGW8

peterisza3• 3 years ago

I think they are training and not testing. I might be wrong, but if I am right, the use of public roads is a necessity. This is how AI works, I guess.

peterisza3• 3 years ago

Anyone care to explain why they downvoted this?

carapace2• 3 years ago

What I think they should do (what I would do if I were in charge at a AI-car co.): There's a whole ersatz city out in the desert somewhere the whole purpose of which is to allow in situ modelling of new "smart" hardware. (This is a real thing, but I forgot what it's called and I'm too lazy to look it up, I apologize for my barbarism.) The people that "live" there are actually paid employees. That's where you test your self-driving cars.

Further, I would have started by making self-driving golf carts made out of nerf that can't go faster than, say, two miles per hour, and then iterated. It's reckless to immediately attempt to make Knight Industries Two Thousand, in my opinion. The potential for mayhem and death goes up with the kinetic energy, eh? Both speed and mass contribute to the "killer robot" aspect of these machines. Start small and light.

Also, let's call them "auto-autos", eh?

8f2ab37a-ed6c3• 3 years ago

They're not very fast at all, unfortunately, at least for the riders. And they seem to try to pick the longest, least congested paths to get to destination to avoid bumping into too many tough turns that might get them in trouble. Takes easily 2x as long as with an Uber.

wholien2• 3 years ago

they are only allowed to run from 9pm to 5/6am though. Rode a few and they seemed perfectly safe.

sbuttgereit• 3 years ago

I regularly (like daily) see empty cars (Waymo & Cruise) driving around residential areas by Lake Merced while I'm driving to pick up/drop off my kid from summer school... that's between the hours 8am and noon. Also see them in the Ingleside neighborhoods during the day: they have been completely empty when I've seen them. So maybe they can't carry passengers other than those other times, but they are definitely driving around outside of the hours you cite.

Having said that, I've not seen any bad driving from them. There are certainly far worse human drivers, motorcyclists, bicyclists, and pedestrians out there on a regular basis, so I'm not against the self-driving cars being on the street.

parl_match2• 3 years ago

I'm wary about using public roads to test these, but I think the way the data is presented is misleading. I'm not sure how it's misleading, but separating "incidents" into categories (safety, traffic, accident, etc) might be a good start.

For example, I could start coning cruise cars, and cause these numbers to skyrocket. While that's an inconvenience to other drivers, it's not a safety issue at all.

By the way, as a motorcyclist (and thus hyper annoyed at bad driving), I find Uber/Lyft/Food drivers to be both much more dangerous and inconveniencing than these self driving cars. We should have scrutiny on the safety of these machines, but they're hardly the biggest problem facing SF.

notakio3• 3 years ago

I moved out of SF ~2018, and only recently revisited, and the increased frequency with which I saw robocars was immediately concerning to me, largely based on the potential for things to go wrong, and with just how frequent they were.

My second thought was, after hearing friends and family still there echo nothing for disdain for them, "how long before someone starts setting these on fire?" as some sort of nuisance campaign against the things. Turns out, 1) you don't need fire, and 2) not that long, since the cone thing started happening about a week afterwards.

standardUser• 3 years ago

I know this is a tired argument, but it's not wrong: There is equal or greater cause for concern with cars driven by people, we're just used to those horrors and reluctant to trade them for a different set of horrors, even if they are conclusively less horrible.

I lived carless in downtown SF for 10 years, usually with a walking commute, and it was not pleasant. Packed streets full of angry, stressed drivers honking and swerving and practically revving their engines as lights change so they can gun it past cyclists and pedestrians down the next narrow street. Anyone trying to replace these agro drivers with marginally-safer robots has my support.

paxys3• 3 years ago

For example, I could start coning cruise cars, and cause these numbers to skyrocket

And why would that not be a valid data point? Cars driving on public roads need to successfully handle all regular day-to-day situations, not just the happy path. It seems like almost every day I run into some Cruise or Waymo (mostly Cruise) vehicle stuck in the middle of the street blocking traffic, and that should not be excused as "oh it's just learning".

Self driving car companies love to release reports comparing themselves against regular drivers when it comes to collisions and deaths, but conveniently skip over every other aspect of driving. The way things are today, if we continue to release more and more of them into San Francisco it will basically lead up to a total traffic deadlock, and yet people will still keep repeating "oh but they kill 0.1 fewer people per million miles so it's an improvement".

veec_cas_tant• 3 years ago

And why would that not be a valid data point? Cars driving on public roads need to successfully handle all regular day-to-day situations, not just the happy path

Are you considering someone putting a traffic cone on a vehicle a day-to-day situation?

jmoss20• 3 years ago

It seems like almost every day I run into some Cruise or Waymo (mostly Cruise) vehicle stuck in the middle of the street blocking traffic

...really?

I've seen this maybe once or twice, total. Is there really some part of town where this is a daily occurrence?

dekhn20• 3 years ago

During the height of Critical Mass, bicyclists would surround minivans and keep them from leaving. Drivers don't know what to do because if they push too hard they'll knock over and injure a cyclist. I don't expect self-driving cars to deal any better with cones, or any hostile action towards a car.

__loam7• 3 years ago

It's a lot higher than 0.1 people. 40 people died in SF last year due to traffic accidents. I've yet to hear about a self driving car killing someone despite how many there are in the city.

I also disagree that coning a self-driving car's sensor is a "regular day to day situation". Throwing paint on someone's windshield or slashing their tires is a crime.

E: making stat accurate

medellin• 3 years ago

Depending where someone put a cone on my car i would possibly just run them over or drive off unsafely if i felt that it was an attempt to rob me or steal my car.

Having things like that categorized differently makes sense because it’s caused by humans being dicks. But maybe you are right and that is the new reality we live in where none of us get to advance because of a small group of angry and stupid people.

mustacheemperor• 3 years ago

If you start putting cones on the hoods of human driven taxis at red lights, that will also cause traffic diversions. The tendency of humans to pester the driverless taxis more than the drivered ones doesn’t seem like any metric related to the technology’s performance. Someone putting a traffic cone on the hood of your car is more than a typical deviation from the happy path.

As to your personal anecdote, all I can say is it’s different from my own experience driving in SF or riding in my partner’s car. We see plenty of driverless vehicles, and generally they’re behaving smarter than the average SF Saturday morning driver, not worse. I don’t live in a cruise-populated neighborhood so I’m sure your experience may vary elsewhere.

On that note, my neighborhood doesn’t have much transit at all. It’s a 30m walk to a grocery store or 25 to the 3rd st muni. We only get limited bus service. So I would be very happy to see additional transit options in SF.

parl_match2• 3 years ago

And why would that not be a valid data point?

It is, I didn't say it wasn't. It's also not really in the same class as an accident with injury.

nova22033• 3 years ago

It seems like almost every day I run into

How are they expected to respond to this data point?

Animats• 3 years ago

For example, I could start coning cruise cars, and cause these numbers to skyrocket.

Google could get tough about that. If you have Google Play Services on your phone, they know who you are. "Your Google account has been terminated for violation of Google's terms of service."

akira25012• 3 years ago

I'm not sure how it's misleading, but separating "incidents" into categories

The premise is that these vehicles would be unilaterally "better" than human drivers and that would justify their creation and use. If that's not the case, then perhaps we need to fully reevaluate the value proposition here.

While that's an inconvenience to other drivers, it's not a safety issue at all.

It hasn't proved to be a safety issue _yet_. We don't actually have enough data to forecast with here. I see this as a safety signal with troubling implications.

as a motorcyclist (and thus hyper annoyed at bad driving)

This is another safety signal I would be wary of... unless "hyper annoyed" is just hyperbole for it's own sake.

We should have scrutiny on the safety of these machines, but they're hardly the biggest problem facing SF.

The problem with automation is small problems have a tendency to combine into massive emergent problems once you start scaling your fleet up.

jstummbillig2• 3 years ago

The problem with automation is small problems have a tendency to combine into massive emergent problems once you start scaling your fleet up.

Not in this case, no. Every driver has a small problem. Each ai driver is just another driver, replacing some driver.

Unless the ai driver problem is not small but actually quite big, I don't see how automation economics would worsen the situation. On the contrary, them all being identical and automated makes fixing the small problems much more viable.

damnesian• 3 years ago

The incidents mentioned in the story would have been shared via first responder dispatchers. Seems like the way forward here would be for the city to make this data available to all driverless cars- maybe some sort of data registration with the city- and for the vendors to make it possible to drop these new hazards into the cars' maps promptly. I don't know how these systems work, or if they can work together, but I can write a computer program and there is clearly a looping input that is lacking in this system.

alexpotato• 3 years ago

Mapping apps (Google/Wayze) already have a reporting feature for accidents and hazards, it doesn't seem like that big a leap to either providing first responders with something like or creating some kind of gamified crowd sourced version.

jiveturkey2• 3 years ago

My kneejerk was to say, yeah 1->3 is a tripling. In modern journalistic terms that is 'skyrocketing'.

Then I read TFA. Yep 30->90. 'skyrocketing'.

That's not to say it isn't concerning. The downed wire incident that is the highlight of the story seems quite bad. But human drivers do much, much worse things, with very high frequency. Weekly sideshows, for example. The article is very unfair.

imperialdrive2• 3 years ago

I'm on the streets of SF for hours a day, every day, mostly on bicycle. Often at night, when these AVs are at their peak output. The difference in their behavior is scary noticeable, especially within the last 30 days. I'm very curious about the programming reason behind it.

jiveturkey2• 3 years ago

i have no basis for this guess, but i'd suppose it's more miles being driven, not a significant programming update.

yep, TFA linkes to another article that there was a recent significant increase in miles driven.

i guess that doesn't necessarily jibe with your perception of a noticeable difference in singular behavior, but our brains do funny things. when you buy a certain new make of car, suddenly you wonder why so many of them are on the road. (they always were)

shagmin• 3 years ago

Stopped at the paywall. But is this likely a case of the number of driverless taxis skyrocketing itself, and the number of incidents per driverless taxi mile being more or less constant?

codampa013142• 3 years ago

In San Francisco, those problems can mean self-driving cars blocking traffic, transit and emergency responders, as well as erratic behavior resulting in close calls with cyclists, pedestrians or other vehicles.

I have mixed feelings about how to test self-driving vehicles (mostly stemming from ignorance). But at some point don't you have to get this stuff out in the wild to see how it behaves or else we are resigned to not making progress on this (or very very slow progress)? And considering that "no one was hurt" and "driverless taxis have never killed or seriously injured anyone in the millions of miles they’ve traveled" are were now there?

slashdev2• 3 years ago

Uber’s driverless vehicle killed someone jaywalking at night in another state. So you can’t say they’ve never killed anyone. Plus Teslas have killed their own occupants plenty of times.

bshipp2• 3 years ago

I'm debating about renewing my truck license for this reason. If it's taking this long for automobiles to be approved and accepted, it'll be 30 years before trucking is automated.

The two decades during which human oversight of automated systems will be mandatory would be long enough for me to finish off my career getting paid to drive while I sit in a cab writing code, periodically checking over the status of my lead truck and the two or three slaved trucks following me.

de_keyboard• 3 years ago

Not disagreeing, but you might find it interesting to learn about the history of the term "jaywalking"

codampa013142• 3 years ago

This is a good point. The statement I made about not killing anyone (in quotes) is from the article, which is only a subset of the overall domain we're talking about.

ketralnis3• 3 years ago

at some point don't you have to get this stuff out in the wild to see how it behaves or else we are resigned to not making progress on this

It's not my problem whether Waymo makes progress on their technology, but they make it my problem when they hit me with their car.

I do have sympathy for wanting to make the progress: we might make better pacemakers faster by testing out 20 designs on 20 people. But nobody wants to be the one that gets the one that doesn't work. And in this case the victims aren't even involved. They didn't opt-in to beta testing the road full of half-baked robots, they're just trying to get to work.

AlotOfReading2• 3 years ago

In the interest of having a discussion, let's assume that AVs are a meaningful goal to work towards for whatever reason. But given that, how do we get vehicles from the drawing board to actually being viable products without on-road testing? I've never seen a complex product that went from non-existent to perfect in the first deployment, so it doesn't seem realistic to expect that here.

Instead, they should be developed iteratively, with design prototypes that proceed from closed-course testing to supervised public testing to closed course autonomous testing, to on-road autonomy over the course of many years. This is what Waymo did. There's a reasonable argument to be made that they did this too quickly, but I can't reconcile that with your argument that they shouldn't have done it at all.

In an ideal world, there'd also be effective government oversight and public safety monitoring at every stage of the above process. Regulators haven't stepped up to do this, though AV companies have done quite a bit to stymie the oversight process as well.

abecedarius2• 3 years ago

Should student drivers not be allowed on the roads either?

minwcnt58• 3 years ago

They've been testing this before putting it on public roads for what, 12 years?

How much longer do you think they should keep testing it off of public roads beyond when they think they're ready?

CardenB• 3 years ago

It's not my problem whether Waymo makes progress on their technology, but they make it my problem when they hit me with their car.

Have they hit anyone with their car? If they have, have they done so at a rate higher than a human driver?

What if they make it less likely for you to be hit by a car? Doesn’t that give you a benefit rather than a problem?

jconley• 3 years ago

@garry did a rebuttal video to this. https://www.youtube.com/watch?v=rjgUPUKD-Sc

gonehome8• 3 years ago

Yeah - none of this reporting can be trusted from the chronicle, the SF politics are crazy and there’s a lot of bullshit/bad-faith.

I hope cruise overcomes this and the people putting traffic comes and such on the cars are stopped.

reaperducer6• 3 years ago

none of this reporting can be trusted from the chronicle

https://en.wikipedia.org/wiki/Shooting_the_messenger

lucisferre2• 3 years ago

While I was skeptical about this video and the tone it opens up with did nothing to help with that, it is actually worth a watch.

solardev2• 3 years ago

What's the summary?

gruez5• 3 years ago

https://www.summarize.tech/www.youtube.com/watch?v=rjgUPUKD-...

jeffbee14• 3 years ago

Kind of cringe when he says "Municipal Transportation Association". Comes off like a guy who has probably never boarded a bus.

DragonStrength3• 3 years ago

I left SF a year ago but had Cruise and Waymo cars in my neighborhood constantly. They were a total nuisance. They'd jerk towards pedestrians in crosswalks leading to this uncomfortable standoff where no one in the neighborhood would ever step in front of one. I'd wager any stats here are way lower than real incidences as I was nearly hit multiple times and never got to the "report it" stage as there are so many barriers to doing so.

laweijfmvo2• 3 years ago

I have similar experiences (as a pedestrian) crossing in front of Cruise cars, especially at 4-way stops when there aren't crosswalks explicitly drawn. It's odd because the cars are generally excruciatingly cautious, but for some reason as I'm crossing they seem to calculate exactly how fast they can take off such that they pass me right as I'm clear of their path -- whereas most drivers might wait for the pedestrian to reach the sidewalk before passing.

What would happen if I "dropped" my wallet and turned around to pick it up? But also, given that they have all those cameras, I wouldn't dare do anything "unusual" like that because it'd probably run me over and say it was my fault, and here's the proof.

I don't doubt the cars' ability to calculate it exactly, but it's a different experience for sure.

cozzyd12• 3 years ago

that sounds very harrowing. I don't think we have any of these cars in Chicago, but I'd be hesitant to push my daughter's stroller in front of one...

Tempest1981• 3 years ago

Sounds risky.

I thought California law was to let the pedestrian fully cross, but apparently not:

In California, the law does not state that a driver must wait for the pedestrian to fully exit the crosswalk or the street before they proceed on their way in their lane. A pedestrian must be safely out of the driver's path of travel for them to begin driving again.

tick_tock_tick2• 3 years ago

whereas most drivers might wait for the pedestrian to reach the sidewalk before passing.

You're joking right? No one wait for you to clear the street. Hell most driving start advancing when you're going to be clear by the time they make it to you.

What would happen if I "dropped" my wallet and turned around to pick it up?

Same thing a human would; hit you or slam on the brakes. They'll just do it faster and harder.

opportune2• 3 years ago

I have the opposite experience, I live by a bunch of 4way stops that drivers get very pushy about if you try to use as a pedestrian. The driverless cars never put me in a dangerous situation at those intersections but human driven cars do very frequently.

okdood642• 3 years ago

What would happen if I "dropped" my wallet and turned around to pick it up?

That would be an extremely stupid thing to do in San Francisco. I would not trust drivers here at all to pick it up. I would just wait until all the cars were gone.

talldatethrow2• 3 years ago

New sport.. anyone in need of cash throws caution to the wind and temps ai cars to hit them for the payout.

fishtoaster• 3 years ago

For what it's worth, my own anecdata is the opposite. My neighborhood (alamo square / wester addition) is rife with Cruise cars and I usually pass 2-3 every night while walking my dog for 30 minutes. They've been, without fail, polite and normal "drivers." I've not felt any concern walking in front of them in crosswalks or biking alongside them when leaving the neighborhood.

jb12• 3 years ago

I live in the same neighborhood and saw one driving on the sidewalk a few weeks ago.

DragonStrength3• 3 years ago

My issues were mostly between Divisadero and Masonic north of the Panhandle, so different strokes. I was doing two half-hour walks per day during most of the pandemic, either to GGP or Alamo Square, so that's the context of my anecdata. Of course, no discounting they're just better now :)

rubyron• 3 years ago

They need a fake hand to rise up and wave over the steering wheel so pedestrians will know they’re waiting for them.

btown• 3 years ago

There's a lot of research on possibilities here! eHMIs (external human-machine interfaces) are the term.

https://www.theturnsignalblog.com/blog/ehmi/

https://www.mdpi.com/1424-8220/21/9/2912

It's unclear that there's a consensus on something that reliably works across cultural contexts, though - which, to be fair, is also a problem with human-human interactions!

DragonStrength3• 3 years ago

A visual indicator the car sees a pedestrian would be helpful.

dmode• 3 years ago

I really think incidents by autonomous cars are the least of SF's problems

shtopointo• 3 years ago

Can anybody find the data set from which they created the graph?

I tried browsing https://data.sfgov.org/browse but there's nothing that immediately refers to "autonomous vehicles", "incidents" or the like...

(I'm trying to piece the data together for myself, and I'm a bit suspicious that their graph stops in April. I'd also want to cross reference to number of vehicles on the road, because if there's been a ramp up since e.g. January, you would expect more incidents to happen).

somewhereoutth• 3 years ago

Given that in any high traffic environment driving is as much a social problem as anything else, it should be no surprise that anything short of Artificial General Intelligence cannot cope and causes disruption.

Considering how little progress has been made with AGI to date, perhaps it's time to call off the self driving experiment.

wordsarelies• 3 years ago

So it's funny, when they first started rolling these out folks were interested, intrigued even. Folks walking would give them deference, mostly not run in front of them if they saw them, very unlike how they treat regular drivers. But now folks are getting sick of them.

So they were modeling all their behavior and driving around cooperative pedestrians and cars (mostly) now they have to model for an adversarial public.

Good fucking luck.

valine5• 3 years ago

Would be easier to change the paint job. Just disguise them as regular taxis.

imperialdrive2• 3 years ago

They are MUCH more aggressive now. It used to be that if you saw one, you could be 99% sure it would stop and wait for the path to be fully clear. Now, they will buzz right by you within a foot or two. Very alarming how quickly they progressed to being careless.

JoshTko• 3 years ago

Google will optimize minimizing deaths/accidents vs maximizing trips/revenue. So it seems that buzzing closer to people is optimal for Google.

__loam7• 3 years ago

Do you have anything beyond anecdotal evidence to support this?

laweijfmvo2• 3 years ago

What seems odd to me is that hardly anyone in SF seems to want them there (citizens, officials) yet it keeps getting pushed through?

gonehome8• 3 years ago

I’m in SF and want them, my friends want them - we aren’t represented by the local media or reactionary politics.

Same with doordash as an earlier example, endless articles about how terrible doordash and Uber are while a large percentage of people in SF use it.

yazaddaruvala8• 3 years ago

I'm so excited for them to come to Seattle!

There are many people that want them! There are even more than want 90% cheaper Ubers.

addisonl2• 3 years ago

Wouldn’t get your hopes up on it being 90% cheaper—typically Waymo cost the same or slightly more than Uber for me. Obviously the economics change as time goes on but I doubt they will make it that much cheaper.

omgmajk• 3 years ago

I'm not in SF, or even inte US - but the view I get from the internet reading stories, posts and comments is that only people in the NIMBY community really gets to voice their opinions in SF.

8f2ab37a-ed6c3• 3 years ago

I'm not sure how one would go about substantiating that "nobody wants them there". It's hard to get the pulse of it outside of the media bubble which tends to be tech-negative.

tick_tock_tick2• 3 years ago

Everyone I know in SF is excited it's really only the officials and some news stories. The wait list to join the App is crazy long demand outpaces the number of cars they have on the road. You're just seeing a very skewed narrative.

clsec2• 3 years ago

I guess we hang in different circles because I don't know anyone who supports them.

pound• 3 years ago

I'm in SF and I welcome them. Human drivers are much worse.

dekhn20• 3 years ago

You're just seeing the loudest folks who are the most extreme, not the actual mean opinion.

cwp• 3 years ago

San Francisco won't be satisfied until they've completely driven out all the tech companies and their employees.

HWR_142• 3 years ago

Maybe. I'm not sure that's a bad goal for the average SF resident.

__loam7• 3 years ago

The average SF resident probably works for a tech company considering how many natives have been driven out by high housing costs. Tech companies are an enormous source of tax revenue for the city and many public services quite literally depend on that revenue. One of the major issues in the city now is the city government's obstinate refusal to approve more housing development. The idea that tech workers are gentrifying one of the most expensive cities in the world is frankly absurd.

tinus_hn• 3 years ago

It would be more useful to have some actual statistics comparing this to human drivers instead of scare pieces. Because accidents did happen before.

MontyCarloHall• 3 years ago

Cruise and Waymo say city officials have mischaracterized their safety track records. Their driverless taxis, the companies say, have lower collision rates than human drivers and public transit.

This is comparing the mean driverless incident rate to the mean human driver incident rate. This is disingenuous, since a small minority of human drivers cause the vast majority of incidents, thereby severely inflating the mean human incident rate.

Today's self-driving cars may be safer than the mean human driver, but I would wager they are far from the median human driver, and absolutely nowhere close to the top 10% of human drivers. It may be possible (but very difficult) to beat the median human with dramatic improvements to current self-driving algorithms, but beating the top 10% of human drivers will require AGI.

scarmig6• 3 years ago

Mean is the relevant metric, though. If it's only the bottom 10% of human drivers who cause all crashes and driverless cars are as safe as the 20p human driver, you can replace all human drivers with autonomous ones and eliminate all crashes.

That's to say nothing of the fact that driverless cars will improve from their current level, and the fact that companies operating them can be held liable for crashes and forced to account for their actions in a way that the worst human drivers cannot (which provides a strong incentive for them to improve).

cozzyd12• 3 years ago

you can more easily prevent the bottom 10% from driving by more rigorous driving standards / more stringent revocation of licenses.

johnfn• 3 years ago

Hmm, I think that, in fact, this comparison is disingenuous! If you want to reduce net collisions, you want to compare to the mean, not median. In that case, once you've produced a self-driving car that's better than the mean, you're saving lives. You don't need to create a self-driving car which is better than every driver, ever.

scarmig6• 3 years ago

It doesn't even need to be at the mean. It just needs to drive more safely than the worst drivers who cause fatal crashes, and you're saving lives.

valine5• 3 years ago

Why does it require AGI? That's a bold, unsubstantiated claim. Let me make my own bold claim: Simple common sense reasoning is very close to being solved with our latest LLMs. You would need AGI if you wanted to replace the team developing your self-driving car. A decent LLM, with vision input that's been fine-tuned on driving scenarios, will be more than sufficient for Level 5 driving.

reso3• 3 years ago

You need AGI for good self-driving because driving requires predicting the actions of other human drivers at an extremely high level. This is second-nature for humans so we barely notice that we are doing it, but it is extraordinarily difficult for non-humans.

yazaddaruvala8• 3 years ago

I roughly agree. I think the only other piece, that is critical to mention, would be remote humans for support in extremely awkward situations to help get the car get back on track as RLHF.

This requires the car can always come to a safe stop, which I think the LLM-based driver should be very capable of doing.

prpl• 3 years ago

P75 would be nice

bluGill• 3 years ago

I care about the overall mean though. Do you know what we call the bottom 10% drivers? Drivers! So long as overall they are better, then get humans off the road, sure the top 10% are worse off, but overall I'm better off as I will never have to deal with that bottom 10%.

speedgoose• 3 years ago

Thankfully driving safer than the mean or median taxi driver seems very achievable.

paxys3• 3 years ago

They also conveniently ignore everything except collisions. For example number of traffic jams caused because your car got confused and decided to stop in the middle of the street.

bhuga• 3 years ago

I hadn't heard this statistic:

since a small minority of human drivers cause the vast majority of incidents

It's tough to find anything relevant on DDG or google, they come up with information about racial disparities in traffic stops...can you link anything to read about this?

abeppu2• 3 years ago

I think a key disconnect is that the 'incidents' being complained of most are not collisions, but they can still be disruptive, and indeed dangerous.

Suppose one were to just install brightly painted immobile bollards on streets, and insist they were "driving" just very very slowly. They wouldn't hit anyone. They wouldn't kill anyone. They would piss everyone off.

This disconnect is repeatedly part of where this conversation gets tripped up.

“Cruise’s safety record is publicly reported and includes having driven millions of miles in an extremely complex urban environment with zero life-threatening injuries or fatalities,” Cruise spokesperson Hannah Lindow told The Chronicle.
The city’s transportation agencies documented several incidents where driverless cars disrupted Muni service. During the night of Sept. 23, five Cruise cars blocked traffic lanes on Mission Street in Bernal Heights, stalling a Muni bus for 45 minutes. On at least three different occasions, Cruise cars stopped on Muni light-rail tracks, halting service.

https://www.sfchronicle.com/projects/2023/self-driving-cars/

Ok great, they didn't kill or injure people which is nice, but they were _disruptive_ in a way which a human driver would not have been. And critically, we can't _have_ a fact-based conversation around those non-collision incidents because these companies aren't even required to report them:

But officials said it’s been difficult to assess their effectiveness because companies aren’t required to report unplanned stop incidents — some of which have been captured on social media — when they happen.
San Francisco officials want the state to require that companies report incidents when they happen.

I've certainly seen cases where self-driving cars behaved in a way that I would consider reckless if a human was at the wheel, but were not collisions, did not result in injury, and I expect will not form part of any statistics reported to any government body ... but I think that's too low a bar.

reaperducer6• 3 years ago

A few weeks ago I saw one blow a red light. Nobody hurt or killed, so it wasn't reported to any responsible agency. It might not have even been tallied by the company, if the car thought there wasn't a problem. But its action was clearly unsafe.

It's hard to get a grip on the problem when the data is so faulty.

gruez5• 3 years ago

This is comparing the mean driverless incident rate to the mean human driver incident rate. This is disingenuous, since a small minority of human drivers cause the vast majority of incidents, thereby severely inflating the mean human incident rate.

Why is this disingenuous? Unless you're also for advocating for banning those top 10% of drivers from driving, you can't just cherry pick only the good drivers.

mtsr• 3 years ago

Banning those drivers does make sense, though. Repeat offenders (speeding, dangerous driving, etc) should just have their license revoked, right?

okdood642• 3 years ago

unless you're also for advocating for banning those top 10% of drivers from driving

Yes maybe they should be at least suspended from driving for some period of time. The problem that won't happen because we don't catch them all.

mdorazio• 3 years ago

a small minority of human drivers cause the vast majority of incidents

Citation needed.

lamontcg2• 3 years ago

beating the top 10% of human drivers will require AGI.

even if it is true, this doesn't matter at all.

if they can beat the median driver and you replaced all drivers with driverless vehicles, they would eliminate whatever bottom percentile driver is responsible for most of the accidents and the world would be vastly safer.

by your own assertion, the 10% best drivers don't matter because they already don't cause accidents.

dontreact2• 3 years ago

I personally don’t care that much the collision rate. I care about the fatality or injury rate (deaths or injuries per miles driven)

I don’t mind picking an Uber driver if it reduces my chances of getting hurt, even if we are slightly more likely to experience a fender bender.

civilized• 3 years ago

a small minority of human drivers cause the vast majority of incidents, thereby severely inflating the mean human incident rate.

Could you cite a source for this claim? I have tried many different searches and can't find any support for it.

minwcnt58• 3 years ago

I would wager they are far from the median human driver, and absolutely nowhere close to the top 10% of human drivers.

I think you would likely lose that wager. In the information Waymo has published about their safety methodology, they benchmark themselves against an always-attentive driver (this already rules out most accidents), which they still outperform by a large margin. Even a driver who is never distracted doesn't have constant 360 degree vision or near-instant reaction times.

tiahura• 3 years ago

GPT 4 can't be trusted to set the menu for dinner without supervision, but we aren't expecting all sorts of wacky carnage by setting pre-alpha KITT loose on the streets?

nashashmi• 3 years ago

At some point we will need separate emergency technologies to deal with automated vehicles as a fill in for exceptional cases. the situations described here are such cases

martythemaniak• 3 years ago

Beware: the State sent a letter accusing the city of manipulating data in order to make AVs looks more dangerous than they were: https://twitter.com/annatonger/status/1673403230804385813

In the 4 Waymo traffic collisions SF cited to prove driverless cars are less safe, 3 were the Waymos being rear-ended and 1 didn't even involve any cars touching, CA says. https://docs.cpuc.ca.gov/PublishedDocs/Published/G000/M512/K...

microtherion2• 3 years ago

3 were the Waymos being rear-ended

While technically this makes them not at fault, it may represent some evidence of erratic braking behavior. Anecdotally, drivers have reported Waymo cars having a propensity for "brake checking" — not with ill intent of course, but risking accidents anyway.

sgillen• 3 years ago

I went ahead and read the accident reports, just because I was curious.

In one case the AV was stopped at a red light when it was rear ended.

In the second case AV slowed down for a person pushing a shopping cart into the street.

In the third case the AV slowed down for pedestrians in a crosswalk.

All the reports are in the report the person you are responding to posted. I do wish we had video footage of the events, while the first one is pretty clear cut, the other two are really dependent on what "slowing down" looks like, and what the pedestrians were actually doing.

cameldrv3• 3 years ago

I've just got to say I'm so impressed with both Waymo and Cruise. Last year they did 4.5 million miles put together in California with zero fatalities. This is an incredible achievement.

saalweachter5• 3 years ago

The US rate is 0.57 fatalities per 100 million passenger miles.

talldatethrow2• 3 years ago

While OPs statement did make me laugh a bit (I've personally driven almost a million miles myself and I don't think my odds of killing someone were 1 in 4), we do have to realize that MOST miles that make up the 100 million mile stats are done at highway speeds far away from pedestrians. If I had to do 4 million miles in SF, I just might have hit someone by now.

dekhn20• 3 years ago

Americans drove 3.2 trillion miles in 2021 and 42,795 people died in traffic fatalities (both values are estimates, and likely do not include every single mile driven or fatality that occurred, depending on how you prefer to count).

So humans drive average of 74 million miles before a person gets killed, and Waymo's driven somewhere above 20M miles but probably fewer than 74M. At this point, I'd expect no self-driving car traffic fatalities, statistically, if the SDC fleet is an average driver.

In some sense Waymo and Cruise are basically waiting around for a fatality (not that they want one) so they start having a denominator that isn't zero.

cameldrv3• 3 years ago

Regardless of whatever the current fatal accident statistics are in the U.S., being able to build a robot that drives that far in traffic without killing anyone is an incredible technical accomplishment. The progress over the past ten years is simply astounding.

jeffbee14• 3 years ago

The national average rate has a denominator that is bloated up with easy freeway miles. Cruise and Waymo never leave SF city streets.

satisfice• 3 years ago

These companies cannot be trusted to be honest about incidents or near incidents. They like to crow about safety, but the safety they want credit for is entirely hypothetical.

There is no good reason to believe that autonomous vehicles are actually safer. They haven’t yet been tested enough, and the testing that has been done cannot be generalized to new situations.

I believe they would be safer if all roads were fully digitized, controlled, with no pedestrians and no other human drivers. We have something like that, now, called railroads.

Mass transit is the better investment.

__loam7• 3 years ago

They are already required to report every accident to a public agency in California.

And the safety is not hypothetical. Waymo is driving millions of miles a month in places like Phoenix. They have a ton of data to support the fact that their cars are significantly safer than human drivers.

jeffrallen• 3 years ago

They are not required to report incidents which have a negative impact on other road users. The SFFD is collecting that data, and it shows that these cars are not able to interact correctly with public safety vehicles and personnel.

pms4• 3 years ago

They are required to report every accident, but not every mile these cars drive and in what conditions, so this data is unfortunately insufficient to make valid comparisons with human drivers...

Really, the problem is the transparency of tech companies. If self-driving cars are really safer, then tech companies need to prove it by releasing more data than necessary, rather than less!

lucisferre2• 3 years ago

Bit of a false dichotomy here though.

jarjoura3• 3 years ago

Laws are written and designed for humans who constantly try to push it to the limit and get away with things. These cars are all over the place and I've been caught stuck on a street in Bernal Heights because it couldn't quite fit and just froze in place. I had to back out of the street because there wasn't any room for me to go around it.

Yes it may be following laws to the letter, but not only was I grumpy about having to carefully back out of a street, there wasn't any way to communicate with the car. It doesn't have any feedback mechanism for me, so it just looks like a confused dead robot in the middle of the street.

I think Waymo and Cruise are both happily working on hard problems in this space with limitless VC funding (or daddy GM money) thrown their way. So I support their efforts, because at the very least, they'll enrich our understanding of self-driving cars even if it were to fail in the marketplace long term.

I'm curious how profitable it will be in practice once the VC money dries up. Paying half a million dollar salaries to employees who need to keep these things up to date adds up. That, on top of hardware costs for precise sensors that will fail fast in harsh weather conditions. On top of normal wear and tear of people in and out of the cars all day for cars that won't need down time.

Lastly, I'll end my thoughts with this... why are we obsessed with the idea of self driving cars anyway?! What human problem does this solve? I am 100% on board with building autonomous cruise-control for my own car, so I'm not against the tech. I just find the taxi angle to be weird. All I see so far in my mind is a product that just takes away an entire class of job opportunities with no actual gain.

pawelmurias• 3 years ago

We can drive people and goods around without having a human waste his time driving stuff around. Not wasting money paying people to do stupid stuff is gigantic gain.

jarjoura3• 3 years ago

The person the tech is replacing will still need to work and earn a wage to afford the goods the system that replaced them is delivering. It's not replacing a relatively high risk job or allowing that driver to scale themselves and improve their efficiency in any way. It's just eliminating their job, because.

As a passenger in a driverless taxi, instead of having someone in the car who might improve the journey slightly by being an interesting person to chat with, I'll likely sit in the car all alone. Worse, I'll probably be shoved ads in my face. So, this new taxi doesn't really add anything new of value to me as a consumer.

I know automated systems have replaced all kinds of jobs, and the closest analogy I can think of is self-check out lines at grocery stores, or ATM machines before that. However, those actually did free up cashiers so they could wander the store and help out in other ways, or restock, or funny enough, help someone self-check out.

Taxis are kind of an island to themselves, and not only will this impact the driver who won't have anything related to jump to, it will hurt the businesses who rely on drivers waiting for the next passenger. Then there's the urban centers these cars are zipping around in. These cars have no employee to generate business tax revenue from, so the cars are consuming infrastructure for no benefit to the cities.

Anyway, thanks for reading my random thoughts. I'm sure things will balance out in the end, but if you work in the industry and have some cool insights, feel free to share.

waldohatesyou• 3 years ago

Isn’t the gain that those folks can find another job that’s more useful?

reaperducer6• 3 years ago

Isn’t the gain that those folks can find another job that’s more useful?

What makes you think that if finding another job that they are capable of doing was so easy they wouldn't have already done so? Because they were waiting for some tech billionaire to throw them on the unemployment line?

QuantumG2• 3 years ago

Then we'll take that one away too.

dekhn20• 3 years ago

"Skyrocket" is a word people use when they don't have the data to support their claim, but want to make it sound as extreme as possible.

When I hear "skyrocket" I think "several orders of magnitude in a very short time". But the source of the quote with skyrocket, basically admits that they didn't have any quantitative evidence to support a massive change in incidents.

Remember: reality is banal. Things are far better explained by sampling error, bias, and base rates, than they are by sudden, dramatic shifts caused by a single factor.

justsomehnguy• 3 years ago

Yep 200% increase is 'skyrocketing'.

Even if means what 1 became 3.

babl-yc• 3 years ago

The SF Chronicle has an unfortunate habit of headlines telling you "what to feel" as opposed to "what happened".

LispSporks22• 3 years ago

I'm with ya normally, but literally a few paragraphs in they have the chart source: the TA, going from 3 to 91 "incidents" over a year. If you turn it on its side, it kinda looks like rocket blasting off.

dekhn20• 3 years ago

That's not a rocket blasting off, it's a rocket failing to reach orbit by barely clearing the launch platform.

More importantly: that chart isn't normalized by miles driven. As the companies have been ramping up, you'd expect (under a random model) that incidents would go up.

It could also be explained by greater knowledge of SDCs, increased news coverage, and more awareness of how to report incidents.

More importantly, I wasn't considering the chart, I was considering what the source of the quote said:

Julia Friedlander, SFMTA’s senior manager of automated driving policy, told state regulators in late June that driverless taxi incidents began “skyrocketing” this year. Though city leaders suspect it coincides with a rise in driverless activity, Friedlander said the city can’t make definitive conclusions because it doesn’t have detailed data.

NickM4• 3 years ago

When I hear "skyrocket" I think "several orders of magnitude in a very short time".

The article actually has quantitative data that shows reports of "incidents where driverless cars disrupt traffic, transit and emergency responders" rising by a couple orders of magnitude in a year.

Yeah, the city officials have to cover their asses and make disclaimers about what they can and cannot conclude, because all they know is that they're getting more reports of incidents; they don't have access to Waymo's data. But given that we know driverless car activity has increased substantially, it seems silly to assume that it must just be a random coincidence that people are now reporting correspondingly more problems.

gruez5• 3 years ago

"incidents where driverless cars disrupt traffic, transit and emergency responders" rising by a couple orders of magnitude in a year.

Right, but how much is that in absolute terms? If last year you had 0 incidents, and this year you had 1, that's an infinity% increase.

dekhn20• 3 years ago

I dojn't think that chart represents anything informative. It's at least partly based on "social media reports" and they say it's "incomplete". Any number of alternative explanations for that chart (which isn't a "skyrocket") explain the results better, such as increased awareness of the cars, increased numbers of miles driven (so the complaint rate per mile is roughly constant), and negative press coverage of incidents.

kwhitefoot• 3 years ago

they don't have access to Waymo's data.

That seems odd. Surely the authority that licenses this sort of activity should have access to all the data that allows proper evaluation of its safety?

Access to such data should be a prerequisite for the agreement.

jeffbee14• 3 years ago

The city assigned a guy to go around and document every time an AV does something. That's why the "skyrocketing". They didn't count them before and now they do.

If some SFFD guy just stood around documenting all driver stupidity it would a significantly different report.

throwaway9274• 3 years ago

Absolutely right. To situate in terms of recent events, there have been three stages of trying to delay / influence the California CPUC robotaxi expansion approval vote:

1. Tried a politician scare campaign. Failed because it was too transparent it was an attempted distraction from their terrible public safety record.

2. Tried coning and disabling the vehicles. Failed because it was too transparently astroturfed, legal liability.

3. Now attempting leaks of cherry-picked vanity metrics so massaged they can’t fairly be called “statistics” to friendly anti-tech local media.

Will fail because national and international media no longer trust our local journalists, and will investigate independently.

The cars work. The incident data meets the standards set forth by the state of California.

Local politicians need to find another anti-tech boogeyman. This one is too vital to the revitalization of SF.

I am not talking my book, have zero stake here.

It’s just saddening to watch a nascent technology finally start working well and then get kneecapped by political games.

Loughla• 3 years ago

My take is that this isn't so much about privately owned cars and/or taxis, but the real politics is in over the road trucking, and these incremental steps are just part of that.

If self driving over the road trucking happens (and that's sort of inevitable), politicians seem to legitimately believe it will collapse our economy. It seems like all self-driving anything gets lumped in with that.

cashsterling• 3 years ago

I don't know why the other reply to this post got downvoted to hell. THe person has a point.

Fully autonomous cars are cool... don't get me wrong. But electric cars and autonomous electric/gas cars are probably not the right solution to our person transport needs. using energy to move a 4000 lb vehicle to transport one or two 200 lb humans is not energy-efficient... most of the energy is spent just moving the vehicle.

A 60lb e-bike can also transport a 200 lb person whilst using far less resources (both to make the vehicle and to operate it) and space.

Public transport also has better resource metrics than person car-based transport.

But here we are...

8note2• 3 years ago

This one is too vital to the revitalization of SF.

Is it? Won't it mean needing to make the city even more car oriented than building out better alternatives than car use?

jasonlotito2• 3 years ago

Since you aren't someone who lives locally to this or is affected by this, why do you think it's "too vital to the revitalization of SF."

twelve40• 3 years ago

sorry, confused

their terrible public safety record

whose public safety record? the politicians'? or, who would even be materially interested to lobby against robocars and Google? honest q.

automakers seem to be generally onboard with robocars and trying their own. Cabbies have been wiped out as a force by Uber. Even Uber seems generally onboard with robocars. So who is this mysterious powerful Luddite villain?

inferiorhuman5• 3 years ago

Two points of contention:

No, autonomous cars don't work particularly well. Anecdotally, I saw some crap stuck in the high tension wires atop a utility pole the other day. So I called PG&E and reported it. On my way back from the store I saw the lineman had arrived and was doing his thing. This street had two lanes and was a one way street, and the PG&E bucket truck occupied one of those lanes. Shortly after a Cruise car pulled up behind the truck and stopped. At various points the turn signals and four way flashers were on with no rhyme or reason. Eventually it decided to pass the truck so it navigated into the other lane and just stopped. So now both lanes were blocked, various turn signal incantations went on, and the blocked traffic started honking.

After that… I watched as another Cruise car go full tilt through another intersection and stop a few feet from a pedestrian in the crosswalk. Way too close for comfort. Autonomous cars aren't inherently bad but they shouldn't be beta tested on public roads.

Second point – no, more cars will not revitalize anything. More cars will contribute to more decay no matter what is driving. Take a look at the Embarcadero now versus when it was designed to prioritize cars over people and had an interstate running along it.

bostonsre• 3 years ago

3. Now attempting leaks of cherry-picked vanity metrics so massaged they can’t fairly be called “statistics” to friendly anti-tech local media.

The situations where they do stupid stuff are also relatively rare. But they are programs, won't they do the same stupid stuff in those situations when they come across them again? If they run over fire hoses in one case, shouldn't that be a show stopper? This isn't some silly app or website, the whole move fast and break things attitude shouldn't be allowed for driverless cars. They should be acknowledging their shortcomings and striving for perfection.

Local politicians need to find another anti-tech boogeyman. This one is too vital to the revitalization of SF.

How does it help the revitalization of SF? Doesn't it take away jobs for those in the lower rung of society? The only people I can possibly see benefiting are the people that work at the companies that make the cars.

scubbo• 3 years ago

I am not talking my book

Today I learned a phrase, thank you!

[Autonomous taxis are] too vital to the revitalization of SF

As an expat from London - Jesus Christ, what a fucking bleak situation. To be clear, I'm not disagreeing, I think you're probably correct - the best way to get people out and spending their tech salaries in SF's service industries is to give them a fast, cheap, convenient way to get directly from their door to the bar/restaurant/club without having to set foot on the street or (heaven forbid!) public transit - but God, what an staggering failure of public policy that this is the case.

gcheong• 3 years ago

“This one is too vital to the revitalization of SF.”

Nothing against the technology but can you explain this a bit more? I fail to see how driverless cars are going to be vital to SF’s revitalization when Uber/Lyft are readily available and haven’t seemed to move the needle much in that regard.

Tiktaalik2• 3 years ago

Will fail because national and international media no longer trust our local journalists, and will investigate independently.

I think it's more that the journalism industry has been so throughly destroyed that there is no real capacity to do investigative journalism.

Zetice3• 3 years ago

Am I reading the graph wrong? a 3x increase over one month that sustains into the next month qualifies as "skyrocketing" to me...

dekhn20• 3 years ago

I've commented elsewhere I don't think the chart is informative (and listed several good reasons why). But all you're describing is a 300% increase, not several orders of magnitude increase.

If you look at Google's QPS chart (maintained in crayon by the original engineers!) you will see that they frequently had to rescale the chart by factors of ten because their growth rate in the early days was exponential.

The chart is not an example of exponential growth.

hosh2• 3 years ago

There may not be data to support this, but on the other hand, informal perceptions -- even if unsupported by data -- matters a great deal with products and systems.

But I would ask the people living there what they perceive, not just relying on a news reporter.

dekhn20• 3 years ago

informal perceptions are about the last thing you want when making decisions about large-scale technology roll-outs. Instead, people should be informed with the highest quality data, and need to be reminded that informal perceptions are often biased and skewed.

wheatzies• 3 years ago

I live in SF and have experience taking cruise rides so I can provide some (obviously anecdotal) information. Overall I'm a big fan of the autonomous vehicles companies but there are downsides with their current abilities. Pros: - As mentioned in the article, the cars generally follow traffic laws. I see a lot of drivers run red lights around my apartment, but at this point I'm fairly confident that a Cruise won't accidentally run a light and hit me. The article starts with a story of a Cruise ignoring dangerous road conditions and caution tape, but that behavior could be hopefully be fixed by working with SDC companies to standardize how road hazards are marked. - They're electric. We've got some serious climate issues to deal with and if these companies can give people more non-ICE ride options, then I think we should be working to normalize them. I feel similarly about the electric scooters in the city. - There's no one in them but me. I definitely fall in the camp of people that prefer not to have to talk to my Uber/Lyft drivers so that's a plus in my book. I also like the tagline mentioned in the article that the cars never drive drunk, drowsy, or distracted. I don't have to worry about who my driver is or what state of mind they're in. - Cost. The rides are cheaper than equivalent Uber/Lyft rides in my experience. One could argue that they're going to make driving for ride-sharing companies unviable as a way to make a living, but that's true for most new automation in a given industry.

Cons: - I'm a cyclist, and I often make eye contact with drivers to ensure they're aware of me. Without a driver, there's not a good way to ensure the car knows I'm there. That being said, I've personally never had a close call with one on my bike. - As mentioned in the article, they can get in the way of first responders. I don't think that's justifiable and should be something that these companies prioritize before expanding their operating hours and range. So yeah, a couple anecdotes and thoughts from someone in the area. They're not perfect, but I think the upside potential is great and the city should be working to accommodate them and get human drivers off the roads as much as possible.

splonk2• 3 years ago

I have no particular data to support this, but anecdotally, my biggest issue with Cruise testing a couple years ago was that it was excessively cautious - randomly braking for no obvious reason, dithering at intersections trying to yield, that kind of thing. I could live with that.

Now (literally yesterday) a driverless Cruise car trying to make a left turn was yielding to me walking across the intersection, and then suddenly decided to go before I'd cleared the intersection. Would it have hit me if I hadn't scurried out of the way? I don't know, but it didn't inspire confidence, and it doesn't take many experiences like this to turn public opinion. Cruise in particular seems to have made their cars more aggressive in my small personal sample size.

BaculumMeumEst• 3 years ago

It’s also language that the mainstream media uses to install as much anxiety and terror into as many people as it possibly can. These publications want you to be permanently miserable, outraged, and most importantly constantly refreshing your phone.

II2II4• 3 years ago

The numbers they gave is more than a doubling, while they were quite clear about the limitations of their data. That includes not knowing how many AVs are on the road, which is something that the companies should be able to provide.

Also look at the claims that AVs have not caused death or serious injury. (I noticed that they did not claim that AVs have not caused injury.) That is great, except I would not expect any given corporate vehicle fleet, autonomous or human controlled, to have a record of causing death or serious injury over a short period of time. The sample size is just too small.

Either way, there's not enough data to prove anything. On the other hand, we have a group with a clear conflict of interest (the makers of AVs) up against a group which has noted concerning incidents but has not been provided with the data they need even when it should be available. Then we have a third group who are being asked to let an experiment on the general public proceed. I doubt that it would pass many academic ethics committees, but if you have money, well, go ahead!

AtlasBarfed• 3 years ago

Anyone seen the rushing dash cam videos where pedestrians are throwing themselves at cars in order to get legal liability payouts. Not that I'm accusing all of the reports here of being something like that, but I wonder if robotaxis are going to result in something like that and actually make the entire driving world a lot less ethical.

Then again it'll all be on camera so maybe not

sroussey2• 3 years ago

In February they hit 1,000,000 miles cumulative for all previous years. Last week they hit 3,000,000 miles.

Cars kill about as many Americans as the Vietnam War, but every year. Aside from guns, I can't think of anything else so deadly that we give to our children as they enter adulthood.

It is common, it is not part of the news cycle. We care more about the unlikely terrorist. People are bad with numbers.

Personally, I think it would a huge ethics violation to not be running tests of autonomous cars (not talking Telsa toy driving stuff).

yafbum2• 3 years ago

I think it's absolutely critical that public officials report what they see: a dramatic increase in reported incidents. Maybe - probably imo - this is "just" due to an increase in miles driven, but without access to data to support it, public officials can't draw conclusions and they should talk about it. The public cannot simply entrust safety to commercial taxi operators i.e. take Waymo's word for it (or the taxicab mafia's word, for that matter).

ipaddr• 3 years ago

3 to 91 a year later is skyrocketing.

barbazoo2• 3 years ago

When I hear "skyrocket" in a headline from a source like that I immediately assume it's something like "from 2 to 4" or something like that.

kelnos9• 3 years ago

The article includes a month-by-month graph of incident counts, broken down by company, that shows the "skyrocketing" behavior starting in March of this year.

The article notes that the data is incomplete, but that only means that there are more incidents than the graph shows.

And I'd also expect that part of the reason for the increase is that there are many more miles driven now by AVs than a year ago. But in a way that doesn't matter: an absolute increase in incidents is a problem, regardless of how much driving is going on.

This is especially the case when we're talking about things that human drivers are less likely to do, like driving through caution tape and snagging Muni wires. Not sure how often human drivers run over fire hoses or drive directly into active fire scenes, but I'd expect it's not often when compared with AV software that seems to just not know it's supposed to avoid those things.

codeka2• 3 years ago

The article notes that the data is incomplete, but that only means that there are more incidents than the graph shows.

But what it doesn't note is where the data is incomplete.

If they started actively looking for incidents in March of this year, and the previous months are just whatever they happened to notice on social media, then it could be both true that "the data is incomplete" and that "there has not been a skyrocketing of incidents".

Enginerrrd• 3 years ago

Your last paragraph perfectly describes what for me was the key epiphany of my adult life. When I gained the scientific maturity to read papers and I realized how shaky the ground was upon which most studies stand, it was pretty enlightening.

throwawhey4362• 3 years ago

"Self driving cars were born in San Francisco. Companies like Cruise and Waymo are creating something that will literally save lives. But some San Francisco politicians hate technology so much, they’re literally willing to make up statistics lying to the public to justify banning them. It’s a lesson in killing the golden goose." - Garry Tan https://youtu.be/rjgUPUKD-Sc

aluminum965• 3 years ago

Meanwhile, the self-driving cars:

- https://twitter.com/desertflyer/status/1677464706251128832

- https://twitter.com/flrent/status/1677483882109882368

dilyevsky• 3 years ago

Garry Tan put out a video about it - https://youtu.be/rjgUPUKD-Sc

onlyrealcuzzo• 3 years ago

I mean, it's easy to skyrocket if you start at 1.

1 crash to 8 crashes is 800%. In percentage terms, it's kind of accurate to say "skyrocket". But in absolute terms, it might not mean anything, especially if the number of taxis is up 1000%.

Nicksil• 3 years ago

You still work for Google?

dekhn20• 3 years ago

No, but I have tremendous respect for what Waymo has been doing.

ChatGTP2• 3 years ago

Unless you provide counter evidence, you're also just making baseless claims too?

aluminum965• 3 years ago

Waymo and Cruise won't actually release comprehensive data. So anecdotes are all we have to go by... and they don't look particularly good. These are egregious driving mistakes that a human driver would almost never make.

- https://twitter.com/desertflyer/status/1677464706251128832

- https://twitter.com/flrent/status/1677483882109882368

Everyone needs to experience these things for themselves before unleashing them on the public. I want them to work, but in their current state, they are an absolute menace.

Zanni• 3 years ago

As a sometime pedestrian, I can tell you that those are the sorts of mistakes human drivers make all the time.

parshimers• 3 years ago

That’s really awful. These kind of incidents need to be investigated. Whatever situation caused the car to just run into the path of a pedestrian in a crosswalk can’t be allowed.

divan• 3 years ago

In some countries that's the default mode of interaction between human drivers and human pedestrians.

heyoni• 3 years ago

I thought all incidents were tracked by the dmv and publicly available. Unless you mean it’s not comprehensive enough…if so, what does that mean?

bmitc• 3 years ago

Remember: reality is banal.

Isn't that just going in the other direction?

I doubt corporations doing open-ended research, testing, and development on public roads can truly be described as banal.

bitsage2• 3 years ago

Luckily the article does provide data…

elif• 3 years ago

In terms of 911 and 311 calls:

"in that month of March, dangerous self-driving car incidents increased more than 300% over the previous month (30 such incidents in February, 96 in March)."

Skyrocket does not seem hyperbolic.

ImaCake3• 3 years ago

The parent comment is saying that the 300% increase is tied to a 300% increase in amount of driveless car activity. In which case, of course incidents increased. If you add 3x as many normal cars on the road the traffic incidents would increase in the same way!

nycdatasci• 3 years ago

Data appears to be freely available here: https://www.dmv.ca.gov/portal/vehicle-industry-services/auto....

I reviewed a couple dozen incidents and in every case human drivers are to blame. The most extreme example of this was an incident in which a manually operated vehicle repeatedly rear-ended an autonomous vehicle out of spite, then followed it back to the dispatch lot and made a variety of verbal threats.

oblio11• 3 years ago

Self driving cars will share roads with regular cars for possibly decades.

amelius2• 3 years ago

So the autonomous vehicle pissed off some human driver, not a good sign either.

wdabney• 3 years ago

So many of these are human drivers at fault doing a hit-and-run. I did see a couple where the AV was at fault, and one poor unfortunate doggo.

Solvency• 3 years ago

Weird, the media used that specific term 24/7 during Covid.

saagarjha• 3 years ago

I have to say I am amused that when it’s news articles of cars running through construction we suddenly start thinking about statistics and when it’s crime in S.F. we jump straight to how the police need to crack down.

Tade03• 3 years ago

To me the term now, after plenty of articles overusing it, means "~40% increase, but the previous number was small to begin with".

notjulianjaynes• 3 years ago

I'm suspicous of your motives for this comment. The article reports a ~30x increase: 3 incidents in April 2022, and 91 incidents in April 2023.

seadan834• 3 years ago

For the data we have, year on year increase of 30x, from 3 to 90, is not enough orders of magnitude for you to object saying "sky rocket" - that you call it a scare campaign?

It is part of the point the city is missing data, these AI car companies have not been transparent with what they know.

seadan834• 3 years ago

Trying to engage more constructively... Do you think that this is essentially a scare-piece (and is no way supported by the relatively limited data that is available [which is a concern in of itself])?

Second question, what if the word 'skyrocket' were replaced with 'increased', would that change your perspective?

collegeburner• 3 years ago

okay now let's look at "per capita" stats. how many more cars are running? the fact that sf chronicle is even running this without actual info is kinda silly. maybe they're not safe but hard to assess without knowing that.

_xivi• 3 years ago

My stance have shifted a bit lately, from "concerns/what ifs" to "let's see", I began to look at it through "the sooner the better" lenses.

Advancements like these are inevitable and have been on the radar for a long time. The more people familiar and informed about the tech, the better they are equipped in order to put the right regulations.

alasarmas• 3 years ago

https://archive.li/jpVut

jhatemyjob• 3 years ago

I miss Uber pool. It was $5 for a ride across town and it was a great way to meet people. Why do we need this self driving Taxi shit? Would much prefer my own car to have that functionality than to rent it out on a per-ride basis

BirAdam3• 3 years ago

I think it depends upon the person. Owning a car is very expensive and many would rather not if they only need it once or twice each month. Others, like myself, need a vehicle almost every single day. I will definitely prefer to own one.

jacobsenscott• 3 years ago

Cruise and Waymo say city officials have mischaracterized their safety track records

Also, Cruise and Waymo refuse to release their safety records. I'll believe the city officials in this case, since they actually have some data. The car companies just have hand waves.

scarmig6• 3 years ago

Cruise and Waymo refuse to release their safety records.

This is not true. They monthly release accidents and disengagement reports to the state.

SF officials, on the other hand, have been caught more or less fabricating a story and misusing data. From the Public Utilities Commission, which oversees the process and AV companies are required to report all incidents to:

Regarding collision responsibility, San Francisco’s analysis appears to omit or overlook relevant facts present in the data and collision narratives that are critical for understanding the context of the cited incidents. The examples of two injury collisions upon which it seems San Francisco bases its analysis of Waymo’s relative injury collision rate (included below in Appendix A as entries for June 2022 and July 2022) are problematic in this regard. According to Waymo’s account as submitted to NHTSA, the June 2022 collision does not appear to involve any contact with the Waymo AV. indicates the Waymo AV was rear-ended by another vehicle, which immediately left the scene. Note that no determination of fault, of the AV or otherwise, is evident through these reports.

(https://docs.cpuc.ca.gov/PublishedDocs/Published/G000/M512/K...)

m00x2• 3 years ago

They include incidents where the car has caused traffic disruption. Recently, people have been throwing cones on top of the cars causing them to stop and impede traffic. This includes SF city officials, which is wild.

Look for #weekofcone on Twitter.

benatkin• 3 years ago

Incidents of me posting the 1337th prime to Hacker News are skyrocketing.

11027

what-no-tests3• 3 years ago

I worked as a software engineering tech lead on the "Ground Truth" team at GM Cruise for a short period in 2018.

During orientation, the founder (Kyle Vogt) told us all that the cars don't need to be better than a human driver, they only need to be "not worse" than a human driver. In fact, when questioned about it, he said that the sooner we could get robot cars on the streets of SF the sooner the software could improve.

The idea was that the more dangerous, "not worse than a human" era would be a necessary sacrifice of safety (compared to waiting forever until we had 100% perfect driving robot cars) -- so we could fast-forward to major improvements in the robot cars' capabilities.

That (IMO) cavalier attitude, a lack of rigor in the way software for the vehicles was being developed, and the fact that we had constant meetings about "diversity and inclusion" rather than on robustness, safety, ethics and quality pushed me to resign abruptly and go my own way.

Very disappointing experience - as I had hoped to see "The Future" before joining GM Cruise and even a fat paycheck and RSUs weren't enough to sooth my scruples.

x86x87• 3 years ago

The classic "some of you might dir but that's a risk I'm willong to take?"

minwcnt58• 3 years ago

Not surprising that Cruise seems to cause an order of magnitude more issues than Waymo. It starts from the top down.

efitz• 3 years ago

I am on my 3rd Tesla- been driving them since 2015- and frequently use FSD beta. I’ve watched autopilot develop into FSD.

The thing that baffles me about all current autonomous vehicle efforts is how resistant they are to having any sort of environmental feedback engineered specifically to help driverless cars.

For instance, in addition to flashing lights, what if fire trucks broadcast a “keep away” radio signal? Perhaps engineered specifically to be easy to determine distance and direction. Perhaps driverless car companies would have to pitch in to pay for the transponder refit for every fire truck. This is just an idea from a layperson and probably unworkable in this form but you should get the drift of my suggestion.

Also, reading this story, why didn’t the city place orange cones in addition to caution tape? I seriously doubt any driverless car company is training their vehicles on caution tape; as a driver for 40 years I have never encountered caution tape in a way to influence my driving. If the city is allowing driverless cars maybe they should have rudimentary training and standards for personnel to avoid unsafe situations.

Ditto for the firemen and fire hoses - ORANGE CONES are what tells me not to drive somewhere.

hedora3• 3 years ago

There are laws regarding traffic markers. Cities can’t just do arbitrarily obscure things and expect drivers to figure out what they want.

One could easily argue that the liability associated with hitting caution tape is just the cost of replacing the caution tape itself (and not additional fines for driving through whatever poorly marked area you went into).

93po6• 3 years ago

Anything other than vision from cameras for self-driving is the wrong way to approach it. The vision capabilities of these cars has to be advanced enough to recognize and respond to the unknown and unexpected, to work even where roads have changed overnight, to work in adverse conditions where it's hard to see, etc. Recognizing emergency vehicles is very far down on the list of difficult things to do, and is not difficult at all compared to many other challenges.

Having transponders is just another thing to do fail, or to forget to turn on, or to be impacted by other things, or to give false positives, or to be abused by drivers, etc. Really nothing can be relied upon other than what you see.

adamkf• 3 years ago

In my experience living in SF, these cars are much safer than the typical drivers in my neighborhood for pedestrians. Most human drivers around here don't even bother to stop at stop signs, and instead just slow slightly. With the cruise and Waymo cars, I feel like the risk is a bit lower when I'm on my bike or walking.

I have witnessed a Cruise car stopping in the middle of the road when faced with an oncoming emergency vehicle, so I totally buy that they aren't ready for prime time yet.

Honestly, I'd prefer if we prioritized enforcing existing traffic laws for regular vehicles.

kaycebasques2• 3 years ago

I live in "the outside lands" [1] so I see them all the time. I think they do a lot of testing out here. I can only think of one time that a Cruise acted erratically and put me on guard. But honestly they drive so gradually and predictably that I am confident I could have got out of the way. I don't recall a single sketchy moment with a Waymo, but I'm a Googler so maybe we just chalk that up as incentivized blindspot and render that as inadmissible in the court of HN.

On the other hand, every day I see human drivers needlessly endangering themselves and others. Every. Single. Day.

Maybe the equation changes with more robo-drivers and less human drivers. I would take the other side of that bet any day.

[1] https://en.m.wikipedia.org/wiki/Outside_Lands

runeb• 3 years ago

I moved here from Europe and I am absolutely shocked at how reckless the drivers are. In the old country I think I saw a car run a red light once in 20 years of driving. Here is seems I see it every other day.

6gvONxR4sf7o2• 3 years ago

Everyone's complaining that the number of total incidents is still small, but the change in those mirrors my own experience driving near those cars, which does not show up in the small number of worst incidents. Meaning the small, but rapidly growing data in the article reflects just the tip of the iceberg.

About the time that the graph shows incidents blowing up, the cruise at least went from a cool thing to drive near to a very unpredictable thing that would act up every time I was near one (which was daily). For each of those ~100/mo reported incidents, there's surely a ton more going unreported, and even more non-'incident' nuisances. Which seem to be skyrocketing as well.

stellalo• 3 years ago

I’m not sure looking at the absolute number of incidents is very informative. How did the total number of miles-per-day from autonomous vehicles vary in the same timespan?

fnord77• 3 years ago

the SF government has already been caught lying about this.

this is a threat to transit fiefdoms. this is what it boils down to

fragmede4• 3 years ago

I will provide a dissenting voice. I was riding a Cruise with a friend the other night, leaving a bar - they only serve a nighttime crowd right now. suddenly a drunk hippie on Haight street started banging on the windows. Rather than get on our way and get us out of there like a Homan driver would, the cruise stopped, powered down - as in unlocked the doors, and called customer service. I shudder to think what would have happened if the hippie hadn't been drunk/inconsequentially high and was on PCP or something. This was a real thing that happened. I reported it to customer support. I can only imagine the conversations internally of this incident. (maybe I should get a job as an sre there.)

AlexandrB2• 3 years ago

There's also the other possibility that it's a firefighter banging on the doors, trying to stop the car from driving into a dangerous area. I'd argue that's the more dangerous scenario too. PCP doesn't magically turn someone into a maniacal killer.

BirAdam3• 3 years ago

Now that you mention it, this makes me think of all kinds of edge cases and how SDCs should handle them, and how they would detect them... hard problems. Of course, it seems like we humans are requiring the SDCs to be better than we are and not be our equals. I am not sure why. I do not know if that standard is good or bad; just making an observation. For example, a person could be banging on a window screaming for help. Many people would just speed away, some would stop and try to help, and in a few places in this country such aggressive behavior could get the person screaming for help killed.

cypress66• 3 years ago

Yeah. That's why I much prefer an L4 car that you still own and sit in the driver seat with steering wheel, than a taxi I rent and have no control over.

(obviously there's no such L4 car yet, but I'm sure tesla/mobileye/etc will get there eventually)

If you're not going to have a steering wheel, then you should have a "keep summer safe" mode.

worik• 3 years ago

I lost count of how many times there was a comment about a lack of data

The city is left in the dark about the exact number of driverless taxis operating in its streets, and the miles they’ve traveled.
city transportation leaders this time around informally collect their own incident data using 911 and 311 calls
Friedlander said the city can’t make definitive conclusions because it doesn’t have detailed data.

Yet Google and GM say things like

“Every single day of delay in deploying this life-saving autonomous driving technology has critical impacts on road safety,” Waymo said in a statement.

On the face of it Google and GM are lying

km3r2• 3 years ago

Or Google and GM have the data, but aren't sharing it because people will overreact because despite it being an order of magnitude safer than human drivers, it still has some incidents that people will focus in on.

cowpig• 3 years ago

That doesn't sound plausible to me. If it were demonstrably safer they would surely be transparent about that

ramraj07• 3 years ago

Sounds like apologia to me. Share the data and stop telling us how to feel about it.

dontreact2• 3 years ago

Humans cause fatalities 1 per 100 million miles.

There is no way to truly evaluate safety being an order of magnitude better until we have 1 billion miles driven.

You can crash less often (reduce fender benders) but kill more people (have rare but egregious errors) so I am not interested in seeing statistics about incidents or collisions.

bhelkey• 3 years ago

California law requires all accidents of autonomous cars resulting in "property damage, bodily injury, or death" be reported within ten days.

All of this accident data is publicly available with details about every accident. For example, Waymo had four accidents in May and four accidents in June.

[1] https://www.dmv.ca.gov/portal/vehicle-industry-services/auto....

pms4• 3 years ago

Sounds like an extremely arrogant and biased comment showing the blind righteousness of tech and its followers. How else can tech build trust with the public than via transparency? Think about this for a moment.

cphajduk• 3 years ago

On the face of it, we don't know.

It's just he said she said.

pier25• 3 years ago

It's amazing after all the billions they've thrown into autonomous driving that it's still isn't a solved problem.

Must be an extremely hard problem.

29athrowaway• 3 years ago

People are putting cones on top of the cars so that they get stuck.

kristopolous• 3 years ago

Is it normalized data?

I'm guessing a huge no

lacrimacida• 3 years ago

I’ll get really pissed if i get run over by a robotaxy (and if I survive of course) but who gets the blame? And is this being equivalent to being harmed by robots? What the hell have we done and what the hell are we searching for, we already have people who could, need and should be doing those jobs. Most likely technology will get there but it better be tested and retested for decades before being widely adopted.

aluminum965• 3 years ago

I feel like I'm being gaslit by the AV companies, especially Cruise. If these cars are so safe, why do I constantly seem them doing sketchy stuff on public roads? Why are there so many videos on Twitter and Reddit of pedestrians literally dodging out of the way to avoid being run over? It doesn't pass the smell test.

ceejayoz• 3 years ago

Why are there so many videos on Twitter and Reddit of pedestrians literally dodging out of the way to avoid being run over?

Part of the answer here is probably "because no one bothers to film human drivers doing this, as it's commonplace".

aluminum965• 3 years ago

I'm sure there is a selection bias. But if AVs are really an improvement, Cruise should prove it to the CPUC with actual data. Otherwise, I'll continue to believe outrageous videos [like this](https://twitter.com/flrent/status/1677483882109882368) over Cruise's claims.

ChatGTP2• 3 years ago

I feel like this argument just needs to stop.

If you're designing planes, they shouldn't just be doing a bit of casual swerving towards the terminal during taxing. We're engineering solutions, not making kind of working stuff.

jononomo• 3 years ago

AI is going to cause human extinction by turning us all into paper clips even though it can't drive a car.

gnicholas• 3 years ago

Is this what Garry Tan's video (The Truth and Lies About Driverless Cars in SF [1]) was all about? I find him to be pretty credible in general, and he makes a strong case that there are some folks within the city government who are misrepresenting the safety record of self-driving cars. I don't know much about the situation, and I'm glad this testing isn't happening where I live (we do have tons of testing on the peninsula, but it's with humans in the vehicle). But I think it's important for companies and regulators to be honest about what is happening, whichever way the facts come out.

1: https://www.youtube.com/watch?v=rjgUPUKD-Sc

kaycebasques2• 3 years ago

Tangential: This article made me curious about who owns the Chronicle these days. Looks like it's the Hearst family now [1] which is amusing because I'm pretty sure the Examiner and the Chronicle used to be sworn archenemies back in the 1800s. The Hearsts owned the Examiner back then.

[1] https://en.m.wikipedia.org/wiki/SF_Chronicle

oars• 3 years ago

Thank you for sharing this video.

sailfast• 3 years ago

Now present the data in incidents per vehicle mile and compare it to human drivers please? Very difficult to take anything seriously that doesn't compare this technology to humans when it comes to safety. We're terrible drivers on the whole.

bandrami• 3 years ago

Is this one of those "400% increase from 2 to 10" things?

perryizgr8• 3 years ago

What's the problem that autonomous cars aim to solve? Almost everybody today is able to drive. Most people are safely navigating billions of kilometers everyday. People who are physically disabled or don't want to drive for whatever reason have the option to order an Uber. So what is it that will change if we attain a 100% human-like AI driver?

wskinner• 3 years ago

Almost 43,000 people died in car accidents in the USA in 2022. Many people hope autonomous cars can reduce that number. The goal is not a 100% human-like driver. It is a driver that is better than humans.

spaceman_20203• 3 years ago

Its okay if a few people die as long as Big Tech gets to gobble up more of our lives and make a few more billion in profits.

Ycdr4thfdd• 3 years ago

Yes, a few people will die. More than a few. But I would bet anything that even in the short-term, and especially in the long-term, far fewer will die than if there were human drivers in those cars.

spaceman_20203• 3 years ago

I’d still much rather be killed by a fellow human than a corporation controlled entity.

“Greater good” has been the excuse for some astonishingly bad ideas.

FreshStart• 3 years ago

Eh, if driverless vehicles are between rides and not parked, could police ticket them for cruising?

ekianjo4• 3 years ago

Could be summarized by a better title being SF says bullshit as usual

exabrial• 3 years ago

Move fast and break things?

CatWChainsaw6• 3 years ago

Move fast and break human bodies!

Roark66• 3 years ago

I'm definitely a skeptic when it comes to putting an AI in charge with no manual override of a 3 ton vehicle with humans on board, but having said that, this article didn't provide any meaningful information if these "incidents" per mile driven are higher compared with human drivers? They say they saw "huge increase in incidents". Perhaps there was a "huge" increase in miles driven by these cars?

It is possible likely hood of such incidents per car/mile increased due to a software update or tweaking of some parameters. I heard people often complain self driving cars drive too cautiously. Try to address this, it is pretty certain there will be more incidents. But is it what happened here? We have no idea.

In general I feel the quality of journalist output diminishes in time, or perhaps it was always horrible? How can one do research, spend time talking to people, write an article and don't provide basic numbers?

johanneskanybal• 3 years ago

Bizzare that they are not required to provide data to the city if that’s needed.

kotaKat• 3 years ago

The easiest solution is to just pull Cruise's licenses because they've been and continue to be the worst offender -- the only reason they got significantly faster approval than Waymo to do things even more autonomously was because they're GM and leveraged their massive lobbying power to get away with breaking shit in production.

KETpXDDzR• 3 years ago

Driverless cars are probably safer in the US, because ppl here suck at driving :) In Germany you have to take driving lessons with professional instructors, including highway and night drives. The driving test in the US is ridiculously easy. You don't even need to be able to reverse-park.

OhMeadhbh2• 3 years ago

Hmm. I always thought these self-driving taxis were competing with regular taxis. But Cruise is saying "They also never drive distracted, drowsy or drunk," which makes me think they don't think they're competing with taxis, but with random drivers coming home from the bar at 2AM. I'm not sure I quite understand their argument here.

93po6• 3 years ago

why not both

OhMeadhbh2• 3 years ago

Because they do not compete well against one of the two.

drunkbetch45• 3 years ago

Basically in an ideal world it will significantly expanding the existing niche biking population in American cities, taking existing riders further/long (ramping up average usage) and introducing tons of new riders to the streets. But still not enough to significantly take cars off roads in a macro-context, especially considering how many people live in suburbs or city neighbourhoods which are glorified suburbs.

thefurdrake• 3 years ago

I don't understand how these driverless cars don't have some sort of override for public safety officials to move them in an emergency. It seems like that would go a long way to reducing concern? Or is this just rabblerousing?

S.F. says incidents by Cruise, Waymo driverless taxis are ‘skyrocketing.’

Related Stories