Contextual Computing At Work
Here’s the rub: Our senses aren’t attuned to modern life.
In the coming years, there will be a shift toward what is now known as contextual computing, defined in large part by Georgia Tech researchers Anind Dey and Gregory Abowd about a decade ago.
Peter argues that we need four graphs to make contextual computing work:
- The Social Graph - how you connect to other people and how they are connected to one another, including the nature and emotional relevance of those connections.
- Your personal graph contains (gulp) all of your beliefs - data relating to a your deepest held beliefs, core values, and personality.
- The Interest graph - what you like - is about curiosity
- Your behavior graph - sensors that record what you actually do versus what you claim you do
I agree that one great value of Peter's contextual computing is to make agents like Apple's Siri or Google Now much more effective in answering questions, making recommendations, and delivering what you want based on how you express it in your own words or gestures, taking into account your current situation, recent requests and interests.
In the world of work, I believe it's incredibly valuable to capture and connect the natural objects of your attention and interest, including tasks, projects, work product, relevant discussion, related references even if you're standing in for Siri or Google Now.
The important requirement is making tasks, projects, pages, discussions and other work products first class sharable, named objects that can be connected to each other and what you're working on, discussed, tagged, tasked, and navigated as well as found using search.
The objects and connections made in the context of work are more reliable than connections that need to be inferred from your behavior - and they're available now, including the ability to connect tasks, projects, pages and discussion in TeamPage and files, discussion, email and SQL databases in your external systems of record.