The evidence is always nuanced and we should be wary of motivated reasoning from those who stand to profit / lose from decisions made from such evidence. /throa

The evidence is always nuanced and we should be wary of motivated reasoning from those who stand to profit / lose from decisions made from such evidence. /throatclearing BUT the thing that most people ignore is that self-driving cars share a learning network, where the entire network benefits from hitting and learning from edge cases. Humans tend to have to re-learn their mistakes on a per-human basis. And that doesn’t even begin to cover the common failure conditions of humans — stressed, exhausted, distracted, frustrated, greedy, emotionally unstable, unpredictable, imprecise, etc. So long as the cost of the damage is borne by the car manufacturer/owner and there was minimal harm done to life, I’m perfectly willing to allow self-driving car experiments to continue. … Not that I had any significant power to stop it in the first place.