Please make a dumb car

- Already in the pre-touchscreen, button-dominated area, plenty of more complicated use cases had terrible UX, and in many cases touchscreens have made them much easier. For example, do you remember how difficult it often was to program radio presets in 80s head-units? There were a lot of real clunkers there.

- There's plenty of newer use cases that would be hard or impossible to make easy with traditional hardware button UX, such as route and EV charge planning.

- Good software offers many possibilities to avoid interactions altogether, or make them single-step instead of multi-step. A good screen-based system will now often suggest interactions pro-actively and situationally at the top-level of the system and reduce interaction to a single multi-modal (touch/gesture/speech) confirmation on what used to be multi-step processes.

- Screens can offer tactile experiences - most recently with the introduction of screens that can do permanent haptic feedback via electrostatic friction coefficient modulation, emulating different surface feels as the finger moves.

- You already mentioned features such as 360 degree cameras: If nothing else, screens are great for showing information. Interactions aside, this alone has a lot of legs left and there's absolutely quality differences between good and bad implementations.

- There's absolutely nothing wrong with a good physical button (on the contrary - few things are more satisfying than a good, solid, weighty button) and having both for use in the appropriate situation. It doesn't mean there's no possible benefits from further investment into good car software.

I feel a lot of this is rooted in first-hand experience with bad systems - which I agree are plentiful, and just as awful as a bad, clunky, unrefined phone onscreen keyboard. Dare to imagine a good one.

Working on car systems means getting to own the form factor and the entire surroundings of a user, much more than working on software targeting rarely-changing phone and PC devices. It's an open field for software and hardware innovation and deserves a little more professional excitement and ambition to do it well.

Probably the one thing I disagree with the article on. Power seat adjustment is a heck of a lot safer in the driver seat than manual levers, and it's not even close.

If a passenger accidentally adjusts the seat too far forward or too far back while the car is in motion, fine, it's an inconvenience. If the driver does this, they become a road hazard.

> After a week of fun and wild-eyed frustration with the 2022 Volkswagen Golf R ... I was ready to scold VW’s product planners. Never has so good a car been so let down by its touchscreen, software and switchgear.

> But I just don’t have the heart. It’s not like the R’s deciders wanted to put a one-eyed, lowest-bidder monstrosity smack in the middle of their flagship Golf R .... But they had a number to hit, a belt to tighten, and nobody was going home until they did.

> ... The VW system’s boot-up is interminable; operators must wait many long seconds in the morning before they can adjust simple things like fan speed and seat heaters. The touchscreen response is like an ATM from the 1990s. The thin horizontal sliders for volume and cabin temperature are not illuminated so they are hard to see at night.

wsj.com/articles/2022-vw-golf-r-sporty-stylish-not-the-best-ux-11643317829