A few factual comments (ex bullet 2):
- seems is an important qualifier, Tesla still only has about 10 human-supervised taxis
- expanding the area of the geofence for 10 existing taxis is certainly more artistic achievement than technical advance, although one might wonder about the artistic appeal of the new geofence to about 1/2 of the population (but what do I know about branding compared to some others?)
- the service is not generally available to the public, so those 52k students are only riders if they are accepted by Tesla’s invitation process
- 10s of millions of unsupervised miles are needed to estimate safety metrics
- Tesla has revealed exactly 0 such miles
- Waymo has over 55 million through 2024 and accumulates more than 2 million per week and now has 100 million driver-only miles
- Tesla will need to drive a lot more miles, a lot faster, to approach L4 autonomy in the US
Where is Tesla’s long-advertised data advantage?
How exactly will Tesla’s data advantage manifest itself, if not now, after years of collecting data?
No one can answer these questions and Tesla is certainly not providing any transparency.
Waymo’s leadership: