If I Could Build Anything
“What you seek is what is seeking you.” — Rumi
“When you want something, all the universe conspires in helping you to achieve it.” — Paulo Coelho
“When the Student Is Ready the Teacher Will Appear” — Zen Proverb
The Problem
Think of your life. The path its taken. Along the way how much of who you are was left to chance. Random people doing random things. Sure, sometimes the scales were tilted by working with exceptional people, or having friends with similar interests, but everyone owes who they are to random experiences. What if it didn’t have to be random?
Everyone looks for happiness. At the boundary of what we know, we tinker with the new in search of something meaningful. So how does one search for happiness “better”. People have traveled the same roads we have, searched for the same things. Even if the search is new territory, usually there are many people looking and finding fragments. This is what Douglas Engelbart dreamt of when he wrote about “collective intelligence” at the dawn of computers. He invented the mouse, the spreadsheet, hyper-text, online meetings, and more. But before he died looking at the Internet he was disappointed at what it had become. The potential of a collective whole had not emerged. Instead it was filled with trivial viral content.
Why not Social Media?
Social media is a little bit there, but its a news stream, focused on “now”. If one asks a psychologist why people have a sense of self, they often say that a sense of self is what one needs to accomplish something long term. Your plans, how you would react, what you value, your style, its all part of a plan for where you want to end up, what will make you happy. When you change your future plans you are changing yourself. With social media, you get what other people “like” “right now”, not so much what you deeply want long term. Your “self” is more than who you follow. You are a trajectory of where you want to go in the world, and the huge catalog of everything you know and can do that the world potentially wants from you.
Pairs
When describing this problem to a very smart person at a conference they said, “that’s the marketplace problem”. Sellers find buyers. I think this is one example of a more fundamental model. Look at this list of pairs:
- business — buyer : seller
- knowledge — question : answer
- books — reader : author
- biology — environmental niche : adaptation to exploit niche
Each of these has a “void” and a “fill”. The void is what is searched for. The “fill” fills the “void”. However many fills and voids are like yin and yang and often interchangeable. Its easier to think of them as a “pair” each seeking the other.
The New Opportunity
The promise of AI and machine learning makes it possible to build systems capable of describing voids and fills across vast numbers of people and to predict which fills and voids go together. On top of that, a system can learn patterns that identify experts and teachers adept at creating fills and patterns that learners take as they discover one piece of information that then builds on another. For innovators, it would be possible to detect a group of “rapid mutual teachers” where a fragment of a discovery quickly propagates from a random member to everyone in the group.
Modeling “You”
The model of you would be all the void-fill pairs that represent both everything you seek and everything that the world could be seeking from you. Using the marketplace metaphor this is what you wish to buy and advertising what you can sell, however in reality the goal is not to make money but to maximize the collective happiness of the world. The model of you is also the history of success and failure at matching void-fill pairs. This history makes it easier for a machine to model future void-pair matches.
Matching Pairs
Language is the best representation for matching fills and voids.
Consists of:
- Is it a “fill” or a “void”, something you have or something you want?
- A text description
- Standard system “tags” that describe the fill or void. Things like “Looking for a job”, “Seeking Information”, “Seeking Person”, etc. The tags are actually a hierarchical taxonomy or semantic network of statements both general and specific but ultimately structured information like RDF used in the semantic web.
The system tags can be defined by anyone but the system learns the patterns between tags and text descriptions and suggests the most common tags to apply. Every tag comes with a question that might have multiple answers each being a tag. For example, “Are you looking for a person, place, thing, idea, or something else?”. As one applies tags the system uses the combined tags and text description to rank the most probable and discriminating questions to ask.
As one defines a void-fill pair the system starts bringing you matches. You know its well defined when the matches are for what you seek. The experience would be similar to having a conversation with a person about what you wanted. You just describe what you want or what you have and answer questions about it.
When done you flag/up-vote the matches that were found and either mark the pair as “Satisfied” and or mark it “Open” so that it continues to bring you updates as they are available.
Collective Models
When enough people seek and find the same thing over and over it would be possible to establish a collective void-fill pair in something resembling a FAQ or a wikipedia page. Questions like “What does one look for in a relationship?” or “How can one create a successful career?”, or “How can I be a good parent?
The Seed
This is my idea, but its only the start. There are many aspects to this ecosystem. Things like communities and their rules, personas, and anonymity. But this is the start. If I could build one thing, one big shoot-for-the-moon thing. It would be what I’ve described. A machine to maximize the collective happiness of the world and start an exponential acceleration of ideas and discoveries like the world has never seen, all from a shared space made of what you have and what you need.
Originally published at davidhagarblog.tumblr.com.
