Tesla seeks to block city of Austin from releasing records on robotaxi trial
reuters.com
reuters.com
You want the privilege of using public roads? Then you play by public rules.
Why would autonomous vehicles be treated any differently then human operated vehicles?
The article unfortunately does not say what data is being required to be made public, but it does suggest that it is proprietary information, not specifically all information.
Crash information, I believe, is public information anyway. Perhaps not the intimate details of each crash, but the location, speed, etc, for both human and autonomous vehicles is (or should be) publicly available. There are definitely datasets for crash incidents, and there are rules regarding reporting of crashes in autonomous vehicles, which Cruise (I believe) failed to abide by and eventually led to GM shutting it down.
If the details were around the methods Tesla is using to assign robotaxis to certain areas and manage charge levels, should that be public knowledge if they have some special proprietary algorithms they feel are valuable?
It’s new.
> Why would autonomous vehicles be treated any differently then human operated vehicles?
That's the entire argument. That's it. Crashes involving citizens or public property on public roads are public information.
No exceptions for having a lot more money than the guy your unsupervised car crashed into.
- Do autonomous vehicles go through licensing requirements and need to get a government ID which requires the sharing of a tremendous amount of information with the government?
- Can autonomous vehicles be taken to court and be saddled with punishments that can end what’s most valuable to them, their freedom?
Autonomous vehicles can and should be treated differently because our current system of incentives, punishments, licensing, information gathering, etc are all designed with humans and human preferences and tendencies in mind.
Humans have been driving on public roads for over a century. We know how they work. We don’t know how a given self-driving system will work, and are trusting the companies developing them to do the right thing, but to make sure we need to see their books. The alternative is requiring each version of driverless software to take a comprehensive regulatory driving test, which is a bureaucratic nightmare nobody wants.
For one thing, we don't feel the same about robots killing humans as we do about humans killing humans, which happens about 40,000 times a year in the US on our roads. I don't know why that is but that's how humans think.
Now ask me if I would like human piloted accident data to be as transparent as what people are demanding of AV operators.
The article says the information at issue is email communications, not trial results.
Regardless of where you stand, people need to know before they communicate whether their communications will be public. (It's already safe to say that people should understand any communications in furtherance of crime can be made public as part of prosecuting that crime - that even pierces Attorney-client privilege.)
Consumer relations, rapport and trust does not seem to be Musk's strong suit. Maybe he should stick with politics and government contracts.
Tesla seeks to guard crash data from public disclosure (reuters.com)
501 points | by kklisura | 1 day ago | 443 comments
They shouldn't have advance notice of Tesla's strategy and plans.
That is, by definition, not historic crash data, because there have been no Tesla robotaxis operating during the past two years.
Nice try but it's not about plans or strategy --- it is about accidents on public roads.
That is, by definition, not about accidents, because there have been no Tesla robotaxis operating at all in Austin during the past two years.
The only difference is I'll publish them on some .ai tld.