The first mistake here it equating robo taxi drivers to "average drivers." Is that even per mile? Shouldn't the coparison be other cab drivers? There is some st

The first mistake here it equating robo taxi drivers to "average drivers." Is that even per mile? Shouldn't the coparison be other cab drivers? There is some statistical skulduggery afoot. Waymo has an advantage here carefully planning tests in optimal conditions and locations to bias the results from the outset.