Of course, they also report when they interceded and there wouldn’t have been a crash …
Agreed, I thought my post indicated that already. But I was trying to focus more on the distinction specifically of unsupervised resulting in crash vs supervised FSD where a crash can be prevented by a real driver intervening. On the crowd sourced FSDTracker a critical intervention can also include dangerous violations where a potential crash was not necessarily involved
1 Like
My point is that the human is in the loop for all driving on that site, whether the car asks for it or not. So, we really can’t interpret those numbers as to what FSD would have done on its own.
I agree which is why I said it is not an apples to apples comparison. While FSD appears very good at almost all times in the community sourced tracker, the tracker does provide a very good indirect indication how far away FSD is likely from a 99.9999% safety achievement.
3 Likes
Self selected surveys also need to be interpreted with caution. E.g., what motivation does a person who never has a problem have to go to the site and report?
I disagree strongly. Maybe some; however, it’s very clear that the majority of reports are those with no incidents
When an info void is created because there is a lack of communication, then others seek to fill the communication void in other ways. Tesla has gone out of their way to create such a void by not testing and reporting test results themselves to regulatory bodies as peers do. They are redacting and failing to be upfront with incident reporting to NHTSA. They created the void; others are responding.
I find the community tracker to be heavily dominated with pro-FSD reporting that presents a very realistic view of current status..
1 Like
So, there is the potential for under-representation of people who have bad experiences with FSD … including those who partially or totally stop using it. Point being that we don’t know.
I doubt it. I see no reason to claim either under or over representation. As many in the media have already said about the community FSDTracker in the Elon info vacuum - “..its the best we’ve got”.
The reason the FSDTracker is dominated by drives with no interventions or minor interventions because the product has very gradually improved over the many years since the late teens!!!
The tracker itself has been around only the last 3-4 years so many decent reports were being generated every day. And now release 14 continues an upward trend.
I would say the worst sources of info for the real status of FSD that have been made available to the public are all the worthless FSD videos of 10, 15, 20+ minutes promoting incident free drives. These are meaningless. If Tesla ever reaches a 99.9999+ safety achievement, we wont know it based on fanboy videos — but such an achievement might start showing up in the communty tracker. But for now it appears very far off.
2 Likes