He said 10x so many times, in so many places, that it was obviously just hype. It has been the target for the Enhanced autopilot, the threshold for removing "beta", and the threshold for engaging FSD.
In fact, the only solid data that we have is the delta for the AP1.0 cars because they operated for 1 year without AP, then another year with AP. Following the Brown decapitation, NHTSA showed that the vehicles were safer (1.4x? if I recall). This was limited to airbag deployments, so it says nothing about fender benders, etc.
The thing is...that doesn't mean much for AP2.0 / AP0.2 (a detail I suspect NHTSA would be surprised to learn) It's a different system at the sensor / path plotting level. It's conceivable that it's less safe in "EAP" mode than AP1...and that is hard to even evaluate with the constant updates.
In fact, the only solid data that we have is the delta for the AP1.0 cars because they operated for 1 year without AP, then another year with AP. Following the Brown decapitation, NHTSA showed that the vehicles were safer (1.4x? if I recall). This was limited to airbag deployments, so it says nothing about fender benders, etc.
The thing is...that doesn't mean much for AP2.0 / AP0.2 (a detail I suspect NHTSA would be surprised to learn) It's a different system at the sensor / path plotting level. It's conceivable that it's less safe in "EAP" mode than AP1...and that is hard to even evaluate with the constant updates.