Well put.
And frankly, the #ML community needs a serious (technical and ethical) wake-up call.
Many would gush over Tesla’s so-called “#AI Days” while neglecting to demand any safety case.
That advanced and legitimatized Tesla’s wrongdoings - wrongdoings where people have died avoidably.
These are #SafetyCritical systems and not “AIs”.
A robust systems safety process is the real, competitive value - not the data, NN architecture, sensors, compute, or whatever.
Tesla has no process.