Tesla will ensure you are a good driver before giving you Full Self-Driving Beta
electrek.co
electrek.co
And then, maybe a government body could ensure that cars are sufficiently safe to use in public. So the car manufacturer isn't putting the individual or the public at risk through shoddy manufacturing or software development.
So then, anyone with this certificate could buy any vehicle, and we don't have to rely on private companies to come up with their own standards for who can be trusted to use their product correctly. Since, if their product isn't safe for all certified drivers, then it's not safe for public roads.
Tesla FSD is safer than most drivers. It’s just novel, so there is a focus on mistakes. To adhere to your principle, we’d have to ban all cars except Teslas.
How can you possibly think that after watching any of the videos on YouTube, uploaded by the biggest cheerleaders no less
First, how do we know that FSD is safer than most drivers? From Tesla's marketing material? Who are they comparing to: the general public, or a matched control group? How are defining safer: number of crashes, number of deaths, property damage, something else? I work as a data scientist, and I'm hugely skeptical of the claim that FSD is safer because I (and I'm sure you too) know how to scam statistics.
Second, I bought an M3 last year and have put 23k miles on it traveling through 18 states. I didn't buy FSD, but I regularly use autopilot. It does some crazy shit, including trying to drive me off the road in Wyoming, trying to drive into opposing traffic in Marin, phantom braking sometimes quite abruptly, etc. And while I agree with the assertion that 99% of the time it does reasonable things, that 1% is horrifying and dangerous. And it's definitely not truly context aware-- as a human driver, I'm aware when someone is tailgating me or when a driver appears to be distracted or if there's a hazard on the road. Tesla, on the other hand, forced me to drive over large debris on 101 because it thought my evasive maneuver was a lane departure. It's a nice thought that people might just be scared of new technology, but I really don't think this technology is actually ready to be doing anything without close human supervision.
Third, I disagree that we have to ban all cars. The current standard is how humans drive. Humans are culpable for their negligence while driving. When the laws catch up to make Tesla culpable for damage caused by its algorithms, and when regulations exist for standardizing algorithm behavior and testing/quality-control, then I'll be more supportive of tools like FSD.
Finally, it'd be nice if we could acknowledge that individual choices often have direct impacts on the safety and wellbeing of others [you didn't say this, but another common response I've encountered is "if you don't feel safe with FSD, then don't use it"]. Just because someone might feel more comfortable "driving" from the backseat of their Tesla[1] does not mean it's actually safe. If someone wants to do reckless things that could get themself killed, then on some level that's their prerogative. However, deciding to put other's lives at risk because your gadget is nifty and usually right is really misguided. Further, Tesla shifting the responsibility for the safety of their product onto the behaviors of its users is shameful.
1: https://www.latimes.com/california/story/2021-05-13/tesla-dr...
The CA Insurance Board should have some questions about this. According to the paperwork filed by State National (the actual insurance company, it's just branded as Tesla) information about driving habits (hard braking, etc) is not used for the insurance pricing.
Also, CA explicitly does not allow that information to be used for rating - so what algorithm are we discussing here? Prop 103 does not allow for any of this information to be used in rating. Seems like this is just a blatant lie trying to make it seem like Tesla Insurance is more than it is.
When I use auto-steer, it beeps at my every 30 seconds if my hands aren't sufficiently applying pressure to the wheel - even though it has control of the wheel, it detects whether there's slight pressure from the driver to determine whether they have their hands on the wheel.
Is this data being sent to them to say that I'm a bad driver?