NBC nightly news with Lester Holt just had a story tonight about Waymo stating that there have been several close calls with pedestrians in crosswalks and crossing guards that often goes unreported. So, fortunately not deadly. Stated that dept of transportation will need to push for better reporting of these types of occurrences.I should clarify the location - SF Bay Area Peninsula, but close to Monterey but without the water. Naturally in software dev, management, and tech startups most my life, but spending more time surfing in SoCal now. Except when I have a nasty cold this week, which is why TMC is suffering through my posts recently
Anyway, getting to the points:
Data: Waymo has only .4 injury accidents per million miles, which is 7x better than humans. Tesla may need that data, but Waymo doesn't. It's necessary for Tesla, but not necessary for Waymo.
Sensors: If it's $6k today, it will be less every year. That's not a barrier. Energy consumption is similar - new GPUs use less power. It's not fixed, especially by the time it scales.
Maps: Waymo doesn't *need* HD maps even today, but they provide additional safety.
Although I don't *know* that they can completely eliminate HD maps (or whether they'd want to), but it would be silly to assume they can't when other companies already have. Just like saying you don't *know* Tesla can reduce errors by 4 orders of magnitude when companies like Waymo already have.
Tesla's accomplishment in E2E AI is amazing, but I think most people are underestimating the difficulty of taking it to production. Right now, Tesla has an FSD level 4/5 prototype, not even a robotaxi prototype.
However, I'm not arguing that Waymo or Tesla has the edge. The person I addressed said he felt Waymo was a dumb idea and thought what they were doing was crazy. My point was that a lot of his objections weren't true or that big a factor. Waymo is running a production service, expanding it to more cities this year, and presumably more the next. A few thousand dollars of sensor costs annually (or much less in future years) is not an "insane" burden.
Hopefully people realize that this can and will improve vs push to remove all driverless taxis from the roads.