[audunw]

1. A human has a lot more options to deal with things like sun glare. We can move our head, use shade, etc. And when it comes to certain aspects around dynamic range the human eyes are still better than cameras. And most of all, if we loose nearly all vision we are intelligent enough to simulate the behaviour of most objects around us to react safe for the next few seconds. 2. Human intelligence is much deeper than machine vision. We can predict a lot of things that machine visions have no hope to achieve without some kind of incredibly advanced multi-modal model which is probably many years out.

The most important thing is that Tesla/Elon absolutely had no way to know, and no reason to believe (other than as a way to rationalise a dangerously risky bet) that machine vision would be able to solve all these issues in time to make good on their promise.

[mcv]

Not only do we have options to deal with it, we understand that it's a vision artefact, and not something real. We understand objects don't vanish or appear out of nowhere. We understand the glare isn't reality but is obstructing our view of reality. We immediately understand we're therefore dealing with incomplete information and compensate for that. Including looking for other ways to look around the instruction or fill in the gaps. Without even thinking about it, often.