Oh in the end? None of them. I believe that the GUI Desktop metaphor is coming close to the End Cycle. I forsee the replacement as something *very loosely* akin to – an immersive environment not tied down to Documents, Files, Folders, phones, movies, TV, cameras analogies/metaphors.
The line will blur and disappear between video, typing, voice, pictures, as I think it will be possible to receive information in the format you most desire.
I think the emphasis will be on “views”; what you receive will be customizable depending on your skill level so that there are no barriers to human/computer interaction.
Someone who is 7 years old will equally be able to customize their experience to their needs and skill levels and primary mode of communication equally well as a 42 year old computer programmer, or a 22 year old game designer, or a 70 year old grandparent.
a) You have a preferred mode of communication. Visual, auditory, mathematical, musical, tactile are the big three, but there are others. There’s logical thinkers, emotional thinkers, introspective thinkers, etc and they all interact with themselves and the world differently.
b) If your preferred output is spoken words, you would have the option to speak and a set of powerful translators will be able to not only save your content as you provided it, but it will be saved in all of its forms, some parts translated into different modes if it needs to complete the set.
If you are the recipient of your work, you will want it back in your preferred mode of reception. Example: You might like to speak to get things OUT of your mind, but you prefer to READ to put things in.
If someone else is the recipient, if they prefer video, then they could get a video of you talking – *if* you give permission for that, that is.
If you wish to convey your exact words, you can have that done.
But if you wish the *idea* to come across, the computer will be able to translate that into analogies that, with your approval, would be transmitted.
Example: If someone understands things better in a line drawing cartoon form, your ideas will be translated into a story told by line drawn characters conveying your ideas. It might not even contain any words; and of course you will have the ability to modify it before sending off to be sure it conveys you correctly.
You will also have the ability to place *your* style of logic into the system itself so it can perform tasks in the ways that make sense to you.
In other words: Programming willl be a key part; but something akin to a much upgraded future version of the MIT SCRATCH project; an extremely powerful visual programming environment, aimed at 9+ but being used even by people who are 5 and 6; and some of the things created rival commercial games; precisely because they were given a set of basic but powerful tools to automate the tasks necessary in a game, in a way that makes sense for their way of thinking.
I’m not saying that Minecraft itself is the model; or that things will be programmable blocks (although I imagine the first iterations of a concept like this will start out that way), but rather, the idea of have a combination of Object Oriented Logic style logic flow, ALONG WITH something they keep TRYING to get rid of out of programming but really works more like how our minds work:
Akin to a teleport, it allows you to jump to any part of a system at any time that is necessary; It is akin to a click on a hyperlink on the WWW; where you are not limited to containers within containers inheriting features but the ability to jump to ANY part of a system just as our mind jumps from analogy to analogy, almost seemingly randomly sometimes.
Immersive, virtual 3D, no limits to the kinds of control you can have over what you input, how you receive feedback from the system, and what the internal processes are that you wish the computer to perform, and when.
So “who will win?” I suspect a dark horse will emerge that will blow them ALL away AND they won’t see it coming.
It will be easy enough to for a 4 year old to use; sophisticated enough for anyone to be able to make the Interface look and feel the way they wish and powerful enough to do anything with anything.
Example: If I wanted my computer to behave like Tandy Deskmate from 1990, I could set it up easily like that.
If I wanted to use long forgotten VAX/VMS typed commands to operate the system, I’d be able to do that.
If I wanted to stick with a desktop metaphor because I’m comfortable with that, I could do it.
If I wanted to create the Inside of the TARDIS and operate my computer from within there, I could do that.
There would be no file saving per se; more like a state system which allows infinite undo and returns to any point within the system.
Too many constraints in all of the systems currently available.
Sorry I couldn’t give you a simple answer; I’m 42; got my 1st computer at 11 yrs old; and as awesome as the changes have been through the last 31 years, some of the changes haven’t been good; and a lot lot of it, ,basically hasn’t changed at all. Human/computer interface has bothered me from the beginning; I work within whatever I find myself facing; but there’s a lot that needs to happen.