That would be an interesting statistics.
Autonomous driving is hard, even at low speed in private parking lots. "Move fast and break cars or sometimes people" isn't the right approach for deploying this technologyhttps://twitter.com/W7VOA/status/1179489907716366338 …
-
-
This Tweet is unavailable.
-
-
-
-
Object recognition, speech recognition...these are tasks we get infants & toddlers to do. As for L5 driving, we wait until mid-to-late teens to give licenses to humans. Most other ML tasks we work on are peanuts
compared to this. -
To be fair, it should be compared with that of human's. I am pretty sure human error rate is much higher than a well trained (not perfect) machine. When there's an option, not perfect though, that could save more lives, shouldn't we just jump on it immediately?
End of conversation
New conversation -
-
-
A technology could save 96% * 37000 annual deaths? Hmm, every month delay is at cost of 3500 American lives.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
It's especially difficult on parking lots. Driving complexity is usually inversely correlated to driving speed. We can drive fast on the highway bc complexity is reduced. Slow on a parking lot because of little standardization and people and objects moving in all directions
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
What do you think *is* the best path forward for quickly and safely developing self-driving cars? What are your thoughts on the current state of simulations for training self-driving cars?
-
Developing AGI is the shortest path.
End of conversation
New conversation -
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Even more difficult when Tesla refuse to use lidar.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.