Hacker News Reader

Comment by sidibe | Hacker News Reader

I have a feeling a lot of stuff is being swept under the rug at this stage. Watching how many near misses are on YouTube from a handful of beta testers, the only accidents I've seen are rims being destroyed by curbing. From the amount of dicy things it's doing I find it hard to believe there's no accidents. Especially before the rollback last week of 10.3 where a lot of people reported very dangerous behavior as soon as it rolled out

View on HN

Wait till you see the amount of car accidents from non autopilot vehicles! you'd want to ban cars all together if a youtube video was posted for every accident.

Come on, that’s blatant whataboutism. Surely we’re good enough here to realize that we can hold both Tesla and other drivers to a higher standard?

Yes, people are often bad drivers. But it doesn’t follow that we must ignore the faults of any self driving technology just because regular drivers crash too. If Tesla is pushing out updates that make their own super fans comment on the cars doing unsafe and erratic things, that’s something we should look into.

I don't think that's an example of whataboutism, given that one of the stated goals of autopilot is literally to provide a safer alternative to the status quo of humans driving manually; pointing out the riskiness of driving in general is not some unrelated thing that's being raised as a distraction from the real issue.

No, we can't. Accidents are going to happen no matter what because driving is a complex task with an infinite number of potential failure modes, and no threshold where the risk of serious injury or death can be considered zero. If something is safer than average, then it is objectively an improvement, and that's all it needs to be.

> If something is safer than average, then it is objectively an improvement, and that's all it needs to be.

GP is asserting that FSD is significantly less safe than previously known, and only driver intervention is preventing FSD from killing more people. Personally I’ve heard way too many “FSD regularly tries to drive into that bridge pillar” for me to believe an assertion that FSD is safer than drivers without significant evidence.

Second, we’re talking about a case where even Tesla owners acknowledge that an update made their cars noticeably less safe. Even in cases where FSD is better than human drivers, which I don’t think is the case yet, we should be quite concerned about the possibility of software updates making vehicles less safe.

My comment was made in response to GP's claim that it is whataboutism to compare the safety record of FSD to human drivers. It is entirely possible that FSD actually is less safe than drivers, and if so let the evidence show it. But the fact remains that you should very much be comparing the safety of FSD to human driving.

Humans are pretty good drivers compared to FSD beta. If you think that's not true after trying it or watching some videos I would appreciate it if you tore up your license and used Uber

> Humans are pretty good drivers compared to FSD beta. Lol! You haven't driven in Montana, have you? What I see in the latest youtube videos is an order of magnitude better than the average driver here.

So sad that the average non-Tesla driver is not as enthusiastic about latest technology as not to broadcast their driving journeys on YouTube for the rest of the world to criticize and make fun of the shady dumpster fire of a manufacturer that made their car when they get into a close call.

Well, a driver is a human having a lot of other things in their head, from work, family, personal issues, driving is just an add-on, the Tesla Autopilot is a system designed to drive and the only job it has is to drive safe and be better than any human driver, that's why most of us have different expectations.

Reply on news.ycombinator.com