The complexity argument rings hollow to me. It’s a bit like saying distributed databases are complex because you have to deal with CAP guarantees. Yes, but people still develop them because it has real benefits.
It was maybe a valid argument 10 years ago, but in 2025 many companies have shown sensor fusion works just fine. I mean, Waymo has clocked 100M+ miles, so it works. The AV industry has moved on to more interesting problems, while Tesla and Musk are still stuck in the past arguing about sensor choices.
Well, it's more like sensor fusion plus extensive human remote intervention, it seems: https://www.nytimes.com/interactive/2024/09/03/technology/zo... . Mind you, if it takes both LiDAR and call-centre workers to make self-driving work in 2025 and for the foreseeable future, that makes Tesla's old ambition to achieve it with neither look all the more hopeless.
Tesla robotaxis have teleoperation [1], which is worse than remote assistance others use because the operators have direct control. Yet they cannot fully remove safety personnel from the car.
It was maybe a valid argument 10 years ago, but in 2025 many companies have shown sensor fusion works just fine. I mean, Waymo has clocked 100M+ miles, so it works. The AV industry has moved on to more interesting problems, while Tesla and Musk are still stuck in the past arguing about sensor choices.