L3 is already commercially available for consumers: Honda in Japan and Mercedes EQS in Germany (awaiting USA L3 permit sales).
Beyond L3 for consumers is still an uncertainty.
China Zeekr and Xpeng can work as an L4 (their competing Tesla FSD beta versions), but they only sell them to consumers as L2.
Commercial L4 testings work pretty well as in Waymo and Cruise in the sense of zero fatalities.
Recent Cruise robotaxi autonomously created a swarm and blocked the traffic is a good reminder about robots:
A number of Cruise-branded self-driving robotaxis were stuck at an intersection for "hours," forcing Cruise employees to manually intervene.
www.thedrive.com
Even with advanced AI, machines still need human skills to get them out of trouble.
Intelligence is complex, so it might be many years or decades before the machine can attain the skill of human intelligence.
To simplify the environment and scenarios for the machine, with less intelligence requirement, a predictable unchangeable commercial route could be a reality: Truck depots, Waymo Robtaxi in the prepared geofence...
When talking about the Tesla FSD beta program, there's no hope that it will ever advance to L3 until Tesla can guarantee to take responsibility for L3 accidents.
Tesla FSD beta still can collide. It wouldn't take
Dan Odowd's challenge to pass his kid-size mannequin collision test. Thus, thinking that Tesla will ever advance to L3 without additional sensors and a better software team is wishful.
Is Tesla still suitable for an investor even when it can't advance beyond L2?
Forbes magazine seems to say yes!
Most should have seen that Tesla is bumping the FSD pre-order price to $15K. This is odd with a product which isn't there yet, and has a resale price of around $3K to $4K in the used market, with a declining global take rate now at 7%. I explore these price factors, and also consider the...
teslamotorsclub.com