One thing I don’t see mentioned is the fact that there’s a limited capacity of attentiveness in our brains. We get tired. So, while the implication here is that

One thing I don’t see mentioned is the fact that there’s a limited capacity of attentiveness in our brains. We get tired. So, while the implication here is that the inattentive drivers are at greater risk — something best measures with overall crash statistics — it’s also a valid hypothesis that over enough miles, the hyperattentive are at greater risk due to fatigue. If Tesla is indeed reporting lower crashes per mile with autopilot engaged, it suggests to me that the fatigue from paying constant attention is a bigger overall risk?