These nannies are perhaps somewhat helpful if you want to tune out while on a freeway and using cruise control for miles or on corporate campus under non-complex circumstances and/or where humans are not in charge but they are not and I don't think ever will be helpful or safe in complicated circumstances and where humans are allowed to make choices. There is no doubt a modern car can more efficiently shift than a human, but it can't necessarily shift when I want it to (sometimes I want to short shift, for any number of reasons) or not shift when I don't want it do (sometimes I want to rev it out, for any number of reasons), much less behave like I want it to when I am trying to do something the algorithms don't understand (like accelerate with slower traffic in front to get around that and slower traffic in another lane and cars are accelerating behind me and the driver understands and can reasonably anticipate3 my choice) and can't anticipate or come close to getting right (like when the roads are curvy, there are parked cars, slow moving cars, and various pedestrians who make logical or illogical choices). No effing thanks to a computer trying and failing to think for me. This stuff is dangerous in real world circumstances at least until we succumb to letting the technology control everything and we are passive bystanders (no thanks).