It’s as Karpathy says, it’s better to sacrifice sensing for scale.
That makes no sense. If you sacrifice sensing, your car will have a harder time detecting objects which will make FSD less safe.
If we start fitting all the recent Waymo news together, it doesn't look good for the Waymo approach:
Long time CEO, CFO, partnerships guy, etc. leaving the company
No service expansion in the Chandler area
This big messy disengagement, where there was mass incompetence and people having no idea what's going on or how to stop the car?
If Waymo can't get their act together with HD maps, the best machine learning experts, and the most experience, how do we expect any of these other Waymo approach companies to succeed? Heck, can Tesla beat Waymo's performance in a general approach? Even if Tesla does beat Waymo's performance in a general approach, they'll need to be far better (than one disengagement every ~1000 miles) to start a robotaxi service.
I am not worried about Waymo.
Waymo's approach is fine:
- They have good perception, prediction and planning.
- They have 20M autonomous miles of experience. The Waymo Driver can handle tens of thousands of edge cases. The fleet is learning.
- They have ride-hailing in one location and poised to launch in SF soon.
- They are demonstrating driverless operation.
- I am sure Waymo has already learned from that bad experience in JJ's video and improved the software and the remote assistance protocols.
- The 5th gen hardware is even better and cheaper.
I am more worried about Tesla's approach:
In 5 years:
- They are still at L2.
- They are still doing rewrites on basic perception.
- They are still figuring out which basic sensors to use, ditching a major sensor like radar and needing to revalidate their entire stack to make sure it is still safe.
- FSD beta only has about 150,000 miles of experience.
- They have no robotaxis, not even in testing.
- No ride-hailing service.