I think you have to drive one for a weekend to truly be the judge. Your argument is the kind of argument people made about tactile keyboards when iPhone first c

I think you have to drive one for a weekend to truly be the judge. Your argument is the kind of argument people made about tactile keyboards when iPhone first came out with a completely software one. On the surface, it was a valid concern - physical keyboards were simply far better input devices than software based ones when considered in isolation. However keyboards are part of a system, and in the system the physical keyboard was a local maxima. For the system to grow past that local maxima, it had to ditch the physical keyboard. Yes there are tradeoffs to make, but it's a different set of tradeoffs - ones that we can adapt around given a better future for the system as a whole. That being said, you really have to live with it for a weekend before being able to judge it in full. I think people are right - purpose built physical buttons and switches will always be superior at some level to software buttons, especially when driving. However, it's worth considering if there is the promise of a better future for the whole system by going down the 100% software route. Maybe it's a bit ahead of its time, but consider a semi-autonomous future where drivers mostly only assist driving and spend 90% of their time free to do other things (1). In the mean time Tesla gets better at delivering software updates to hundreds of thousands of cars in the wild, improving their software processes as a competitive advantage. Other car companies simply aren't doing this. In the future, every car company will need to do this and learn to get good at doing it. (1) If you've tried Autopilot, you know it's good enough that engaging with the touch screen for extended periods is no problem at all.