I want a separate thread for this apart from the Elon thread. If you're testing this I'd like your thoughts.
I've been testing this,it handles most driving situations well but I've had to take over several times to avoid dangerous situations.
1.) a puddle- There was a puddle 10 feet long the width of a lane. It was in my lane, and about 30 yards from it FSD suddenly made the warning noise, braked, and reverted back to human driving. Fortunately there were no other cars around so nothing happened. But it didn't want to go through that puddle and didn't know to change lanes despite not having any cars around me for at least a hundred yards in either direction. Two hours later I was on the same road and approached the same puddle again on FSD and this time it didn't switch off but I cancelled it to go around the puddle at the last moment.
2.) lane changes - I think there's a difference in FSD versions on the highway vs surface roads. FSD does a great job on surface roads changing into the appropriate lanes* well before a turn. On the highways it waits too long, especially when there's heavy traffic.
3.) *for the most part - Last week on Parmer FSD tried changing lanes to the right in the middle of an intersection to pass a car. Only problem was, the lane it tried to change me into was the shoulder.
Questions:
1.) How does it handle stopped school buses on the opposite side of the road where there isn't an improved median?
2.) How does it handle emergency vehicles? So far I haven't seen FSD change its behavior to yield to an emergency vehicle. I haven't had one come up on me from behind thought.
With Elon lowering the price to $99 / month I'm wondering what the uptake will be. It's impressive enough to think a 20% adaption rate is reasonable. With a 40x or 50x multiple that'd be around $8 - 10 / sh. Curious what others are experiencing.