Setting a precedent for releasing footage means fault is automatically assumed when they don’t voluntarily release it.
It’s a problem if they’re not even releasing it to the regulators, but that’s not happening here.
The regulators don't want the footage released to the general public any more than Waymo does—the last thing they need is a bunch of armchair traffic accident investigators telling them how to do their jobs.
If you’re not an expert, it’s best to let the regulators and insurance adjusters do their job.
Public trust requires building it first. As one of the first instances of a waymo crash, yes the public needs to see it. If after reading the footage in the first 99 crashes and in each time waymo’s assessment was valid, that’s when Waymo has public trust and can credibly not release every single video but only do it on a case by case basis.
Waymo is regulated by independent agencies (CA DMV and NHTSA). They are watching the videos and assessing if Waymo is telling the truth. Their permit is pulled if they get caught lying (like Cruise). How are you and thousands of SF residents more qualified than them? Why should I take your assessment more seriously than that of the regulators?
The public is allowing them to borrow govt-built roads to test their products and potentially endanger the public. That's more than enough reason for Waymo to be transparent here.
I'd like them to release the video, but just be clear what you want, and best not to act like it is required of them.
Waymo reports to the DMV, they do not report to you or me.
Anyway, I'm talking about a moral or PR obligation, not a legal one. Waymo is using roads that taxpayers pay for, and that raises the bar for transparency for them.
States and the Federal government are managed by an Elected government and their appointees, and are empowered to levy taxes.
Taxpayers foot the bill, and some taxpayers are even citizens with the right to vote in elections.
The whole “X belongs to the taxpayer” is crap rhetoric you and many many many Americans were raised on, but which does not accurately convey anything about your relationship with the roads and highways you use. We pay for them, we elect the people who are charged with managing the entity that does own them and employs people to build and maintain them, but we do not own them.
I'm saying that we pay for the roads, and the entities that legally own/maintain/control the roads are supposed to be under the control of our elected representatives. We own the roads in the sense that they are public spaces that we pay for.
There is a moral obligation to transparency when benefitting from public resources, whether or not there's a legal one.
I mean, if it reflects positively on them, they have huge reason; they want to put as much distance between their response and Cruise's as possible, and show that they're acting in good faith.
A large part of what killed Cruise was, well, either the cover-up or its employees' amazingly persistent internet connectivity problems, depending on how cynical you want to be.
If they want to put as much distance between themselves and Cruise, they just need to be transparent with the regulators and not play games with them.
I think it is better that we distinguish that this type of event does not actually inform us about the safety of driverless cars nor does it provide enough information to adequately update our positions in either direction. It is void of so much context that it is essentially just noise. These things are hard to measure and even harder to make accurate decisions on. In an extreme case, it is possible for driverless cars to be "better than humans" at driving in the notion of accidents per mile (even normalized) while also being "worse than humans" in the notion of striking cyclists or pedestrians. "Accidents per mile" marginalizes the type of accident, but this would be important information as to making effective policy about when and where a driverless car can deploy (e.g. on average fewer accidents but most miles are on the highways. If there's a high rate of failure in detecting children in suburban settings, I sure wouldn't want that car driving in residential areas. The nuance matters). I think it's fair to say that the vast majority of us here on HN have a poor concept as to what the actual nuanced data says here and personally I feel like a video would only further sensationalize the event. I would not be surprised if a video were released which showed that Waymo could have not performed any better (without x-ray vision or omniscience) yet it also be used to further cases against them. I'm sure most people would make an accurate and reasonable conclusion upon seeing the video, but I still don't have faith that that means it wouldn't end up being sensationalized.
Waymo has 360 degrees cameras. There's absolutely no lack of context if the video starts 2-3 minutes before the incident and stops when everything was sorted out at the scene. Bonus points if it has audio also.
See all the dashcam videos on YouTube.
There's a common misconception that there is not an expectation of privacy in public but this is not supported by the law. The notion does not help people but does help those that wish to surveil. Be careful with this claim.
IIRC they were sued about identifying people and car plates.
There are 43,000 US road deaths per year. If all driving was autonomous, even if autonomous driving was 100x safer than humans, that would still be 430 deaths per year.
And some of those accidents will be failures that a human would not make (sensor failure, bad software update, edge cases).
What you _can't_ do is what Cruise did and hide evidence from the public and regulators fearing that the truth would impact your business.
[0]captured in multi-angle 4k, replete with gore and blood splatter. leaked to the press, compiled into tic Tok reaction videos, etc.
What’s not public is the video footage. It’s still shown to the regulators when they want to examine, it’s just not publicly released.
I hope they’ve released the full video to investigators, but they’re in a tight spot — SF city government is looking for reasons to come down harshly on the company, and trying to sway public opinion ahead of the city’s verdict won’t help.
And on top of that, I'm far less interested in the footage than I am the telemetry. What was the state of the gas and brake, what was the steering angle, how long was it from detection to action. Those things concretely answer the question of "would a human have done better or worse in this situation?", which is ultimately the only question that matters.
You absolutely cannot answer that question without looking at the video and considering how easy or hard it would have been for a human to detect (or suspect) the cyclist in that situation.
It sounds like you're making a deeply unreasonable assumption that no human could have detected or anticipated the cyclist faster than the computer did.
Given the current state of the art, those times are frequently not the same, and the difference cannot be answered without seeing the human-consumable video footage of a prospective human driver.
It's your post that is nonsense, as it appears you've read about half of what I wrote before you replied.
In this quote, you are asserting that something is true::
> What was the state of the gas and brake, what was the steering angle, how long was it from detection to action. Those things concretely answer the question of "would a human have done better or worse in this situation?"
Unfortunately it is not true: "Those things" you just identified in "the telemetry" can not "concretely answer" the question of "would a human have done better."
In particular, that is because "done better" includes detecting the obstacle or risk sooner--not merely time between detection and braking. To answer that, the inputs that would have been available to a human must be scrutinized by other humans, which means the video is important.
That'll be my last reply though, I'm watching a youtube stream of a guy trying to get gentoo to boot off NTFS, it's not going well, but it is more interesting than providing detailed analysis of simple language that you've misunderstood.
> Did I at any point say the video wasn't important?
Sounds like Motte and Bailey time.
The expansive Bailey: "We don't need raw video to concretely answer whether the machine did better or worse than a human, the telemetry is sufficient."
The defensible Motte: "I never said the video was unimportant..."
> pedantic [...] detailed analysis
Step 1: Casually dismiss the other person as having not read your post.
Step 2: When they lay it out in detail, whine that they read your post too closely.
What most people want to compare is time from when the bicyclist could have reasonably been detected to action. This requires looking at the video frames to decide when we think a human would have figured it out.