Cruise's basic theory was that the best way to initially scale testing wasn't to build a bigger fleet, but instead to test in busier environments where long tai

Cruise's basic theory was that the best way to initially scale testing wasn't to build a bigger fleet, but instead to test in busier environments where long tail events are more frequent. There was still testing in a few other locales (primarily Arizona and Michigan), but a lot of the focus was on SF as a result.