Why safe Full Self-Driving can't ever happen

Like most AI nowadays, they are attempting to substitute data and pattern matching for reasoning and intelligence --- because this is what binary logic is good at.

And it often just does not work. A human doesn't need a stored pattern of every possible scenario in order to make judgments in real time. And they don't "freeze" or hit the brakes whenever they see something they don't have a pattern for.

Most accidents with current self driving cars are rear end collisions --- typically blamed (erroneously) on the other human driver involved. In reality, these accidents are precipitated by the self driving car hitting it's brakes for no reason that the surrounding human drivers can see, understand or anticipate.

Interpreted more reasonably (i.e. by someone other a self driving manufacturer), the available statistics show that self driving cars aren't nearly as safe as human drivers.

Navigating ethical challenges presents a significant obstacle on our path to achieving safe Full Self-Driving. Imagine a scenario where autonomous vehicles find themselves in situations demanding quick decisions that directly impact people's safety. Take, for instance, an inevitable accident where the FSD system must choose between protecting its passengers or pedestrians. Resolving these moral and ethical dilemmas is an intricate task that requires the creation of well-defined rules for the system's behavior. Deciding whose safety should come first raises profound questions about human values, ethics, and societal norms, making the quest for secure FSD an ethically complex journey.