For the billionth time, we only recognize human mistakes because we also recognize acts of negligence and mal intent. If we can't hold someone accountable in the same ways for avoidable AI harms, the notion of an "AI mistake" is not only meaningless, it's dangerous.

Reposted from Business Insider

Society accepts that humans make mistakes, but it will also need to decide an acceptable level of error from AI too, says Uber's CEO.

Reply