I'm trying to reduce suffering (the delta between the real and the ideal). For that, we need to communicate better.
Today, we generally communicate through natural language ("Hi, can I get a taxi at XYZ?") or software (Uber's "REQUEST UBERX" button).
I'm taking the best of natural language (ubiquity, flexibility, generality) and software (unambiguity, conciseness, guidance, automation) to create a better communication paradigm.
First, I'm getting rid of action verbs. Everything is the description of a state. Instead of "I want to go to the park", think "I am at the park" or better "me.location = park".
Second, I'm reversing the flow. Instead of manually requesting things, you get things suggested to you. Most of communication is done by accepting or rejecting suggestions, like with Akinator or Tinder.
Third, I'm making the real world interactive by allowing physical things to act triggers for suggestions. With the press of a button, you can capture you current location, a QR code, an RFID tag, a picture of a book, some Chinese text, the song currently playing, etc. The captured entity will be processed, meaning will be extracted, yes/no questions will be asked, and you'll be offered ways to interact with it.
Basically, I want to be able to point at something, get information about it, and act upon it.
This is neither academic research or commercial enterprise. I'm currently doing this on my own, but anyone is free to join.
Next week, I will publish a blog post explaining the relationship between GTD (Getting Things Done methodology) and the future of UI. This should cover some of the ideas I have mentioned here.