Comment by ufmace

I don't think I'd want to take much from such a statistical result yet. A sample size of 1 accident just isn't enough information to get a real rate from, not that I want to see more collisions with children. Though this is also muddied by the fact that Waymo will most likely adjust their software to make this less likely, and we won't know exactly how or how many miles each version has. I'd also like to see the data for human incidents over just the temperate suburban areas like Waymo operates in.