Also, this "no interface" way of thinking often leads to poor discoverability. If users struggle to find the features they expect, the interface is not "getting out of the way". It can get pretty frustrating.
Yes, the name of the manifesto is very click-baity and in my opinion presents a naive view on interaction design. But to play the devil's advocate: I've attended a workshop by Golden Krishna, and I think he purposely oversimplifies his point to get it across to people who are not professional interaction designers.
For a good rant on the many issues with touch screen-centred interface design, I suggest this essay by Bret Victor:
http://worrydream.com/ABriefRantOnTheFutureOfInteractionDesi...
Note that at no point he suggests abolishing having an interface altogether.
"Ideally, we’d like users to go to a site and find the answer they’re looking for right there, in front of their eyes. That would mean zero interaction cost and is the holy grail of usability as a field.
Unfortunately, zero interaction cost is rarely attainable, since most sites and apps offer many things that users may want to do."
[0] - http://www.nngroup.com/articles/interaction-cost-definition/
Depends how long term we're talking here. If we're talking in a 100 years then yeah 'no interface' is king because ideally we would have microchip implants that allows us to intuitively gather knowledge and do whatever computing we do today simply in our heads.
But today? I have a feeling what they REALLY mean is seamless integration is king over just another app to limit the amount of time used navigating user interfaces (and if we can get that to zero such as opening a door even better).
Putting in a motion sensor and removing the light switch forces the user of the room to meet the conditions the engineers of the room designed it for.
Providing a motion sensing light switch that has a manual toggle on it allows the user to correct for potential use cases that the automated sensor is not programmed to meet.
The best interface is the intuitive one, not a lack of one for the sake of saying it does not need to exist.
1. touch-handle to unlock is awesome and useful 2. touch-handle to lock is worse than useless. keep this on the fob and on the interior. nobody reaches for the exterior handle when they're walking away from the car. 3. the touch shouldn't be too sensitive, and should only be on the inside of the door handle (or under the door handle, whatever the layout is) - my 4. there should be a delay of about 500ms i.e. touch it for half a second before it unlocks. 5. no button, like you mentioned
It wouldn't be rocket science to add a layer that worked out whether you were walking towards or away from some specific hardware reference transceiver, checked some credentials, and responded with some action.
So - car "keys", hotel room "keys", house "keys", house lights, virtual boarding passes and other travel tickets, etc.
Watch is supposed to do some of this but the UIKit in iOS wasn't really designed for this kind of interaction. So the 'wave your watch around to unlock your room' idea still isn't as streamlined as it could be. NFC/RFID may not be the right technology for this.
It's a good bet that Apple and Google have either worked this out already or will work it out very soon. Disney already have.
It's true that current app UIs are non-ideal. But apps are a transitional solution, and I don't expect them to be around in their current form for more than another few years at most.
Actually all terminal, vin, emacs, got, etc users here probably agree: best interface for advanced interaction with complex system is still the console. I would configure and program and repair my futuristic car from a remote keyboard and black screen. Then it will open the doors by nfc when my phone is close enough. And if on the move, I'll reprogram it from a term app in my phone.
Take the author's critical example of a smartphone app to unlock your car. Yes, that's terrible. Now, let's look at the actual interface - the key. I've owned two Subarus now. I loved the keyless entry device on my 1999 Outback. I feel the keyless entry on my 2010 Forester was a step backwards. I accidentally trigger it when I don't want to. It's unreliable, often taking several presses to work. It adds an extra button I've never used that is redundant functionality.
How about another interface? We recently got a new turkey baster, made by Good Grips. It has a built-in rest. Turkey basters get greasy-dirty, and you keep using them off and on for hours. The rest keeps it from getting grease all over your counter. That's a great user interface!
The problem isn't "no interface". The problem is "there's an app for that". Not everything falls under smartphone convenience. That doesn't mean it doesn't have a UI.