To be fair, autonomous flight is probably an order of magnitude easier than autonomous driving. Particularly because it's far more predictable and you can put beacons on drones to automatically avoid collisions.
The sky is mostly empty space and landing zones are clearly marked and purpose built. You don't have to worry about any pedestrians or unpredictable drivers or roads.
> You don't have to worry about any pedestrians or unpredictable drivers...
Actually you do. Most airspace is uncontrolled, where the collision avoidance algorithm is "see and avoid". There are plenty of airspace users who do not broadcast any kind of beacon apart from being visible to other pilots.
You could demand that they all wear electronic beacons. Such technologies do exist. However they aren't currently predicted to work well enough in the expected congestion created by large numbers of autonomous aircraft in the future. And it faces most of the same practical challenges as putting beacons on pedestrians and all other road users.
That is not fair. Autonomous flight is just as hard as autonomous driving due to the need to handle mechanical failures. If something breaks in the air you can't just pull over and stop. Current autopilots are generally unable to cope with failures and rely upon a qualified human pilot being ready to take back control within a few seconds.
Flying is an order of magnitude worse than driving energy-wise. Even with electric motors it's still going to be a problem. Tesla travels for about 300Wh/mile, I doubt any flying machine would be within 10x of that.
That's why I mentioned beacons for collision avoidance. They wouldn't be expensive. If we're about to build this industry from the ground up I would expect to see something like that done anyway.
The sky is mostly empty space and landing zones are clearly marked and purpose built. You don't have to worry about any pedestrians or unpredictable drivers or roads.