I agree except too much is being learned by trial and error vs looking to other industries. Aerospace has known the L2/L3 doesn't work for decades. To many folks in AV, who are mostly Commercial IT folks, think they innovate everything and that no one else has done a thing. They are extremely wrong about that and wasting valuable time and money. Not only regarding L2/L3 but also in not using aerospace level simulation for AI vs public shadow driving and not using aerospace systems engineering practices.