[jsjohnst]

Correct, but how many cars does Goog/Cruise/Uber have on the road daily collecting this high data rate? 1k? Let’s be overly optimistic and say 10k. Remember, normal driver driven cars don’t apply here generally, just the research cars.

Tesla on the other hand has hundreds of thousands of cars on the road and is rapidly on their way to 1M cars. Data won’t be as raw and abundant, but that doesn’t mean it isn’t potentially very useful none the less.

[visarga]

The kind of data Tesla collects is useful for general self driving, less so for solving hard cases. For that we need simulations and real life enactment, like Google does. We already have the basic data, we need the special case data now, or we could use the driving data from Tesla to discover new special cases.

Simply putting more SDCs on the road does not guarantee they will focus on the data they need.

[jsjohnst]

> Simply putting more SDCs on the road does not guarantee they will focus on the data they need.

Agree completely

[jaimex2]

Tesla's are currently sending back images of things they have trouble identifying only. They are ironing out the last % of edge cases it seems.

Not sure how close this puts them in the race.