Vantaggi
Waymo is doing this right and is definitely ahead of the pack. Fully driverless cars are very far away and many teams are making it to the 95% mark, but that last few percent is hard. We have a real big group of drivers and test systems and I do think waymo will be the first to get there, but not anytime soon.
Svantaggi
3 things: 1 - there is a really long time horizon between the work you are doing today and its impact. Both that these cars aren't going to make a real impact for another 20+ years, but the company is also moving slower and being very cautious these days, so your code doesn't ship for months to years now. 2- the remaining challenges are relatively boring. All the cool sims, sensors, and pipelines are running solid. We're deep in the land of tinkering. 3 - the best team is long gone. Few OG Googlers still on the team combined with the above makes life here more and more boring by the day.