Web Wide Matrix

For many years, I’ve been really suffering from the problem of information overload: I’m subscribed to many content feeds, which are very valuable to me, but don’t have any way to effectively consume them – even if I’d sit all day & read them, I won’t be able to cover all of them.

Thinking about this problem I understood 3 things:

  1. The problem starts with the medium – the interface in which we consume content feeds today is like a baggage carousel, where you need to look at a list of items, arriving in random order, until you see what you’re looking for. This interface is somewhat effective with few dozens items, but will never work with thousands. On the other hand, an interface such as a supermarket is effective for selecting a few items out of thousands of items, because it’s an organized space that you can walk into & see all items, & since you go there frequently you know exactly where to find what interests you & discover more along the way.
  2. There are many good technologies today to automatically process large amounts of information, in order to enable people to make value out of it, but to use these technologies you need a team of engineers working for weeks or months building a solution for some specific content feed & use-case. This is similar to the situation before the Web, where if you wanted to connect to some information you needed to build the software, protocols & infrastructure for accessing information remotely. The Web introduced a simple & powerful standard way to access information, that removed these barriers & enabled anyone to publish & access information from anywhere in the world. Similarly, we need a simple & powerful way to enable anyone to process any content feed & make value out of it. This should be as simple as writing a simple HTML file.
  3. Until today, every person needed just an internet connection & a device to access information – something to run a browser on & access information & services. However, by now we stretched to the limit our ability to consume & make value from the sea of information available to us – Slack/Email/Articles/Twitter/Data/Opportunities/Events/&c. The introduction of powerful machine-learning & autonomous software agents can now enable every person to have a team of bots working for him by processing information & creating value out of it. So, a browser & internet connection is not enough anymore – you need another software/service to cope with the never ending streams of information flooding you, that needs to be available & affordable for everyone, just like an internet connection & a browser.

So I set out to design an architecture & solution based on these ideas, which I call the Web Wide Matrix. Inspired by the Matrix movie, it is a

  • Virtual Reality interface for consuming content feeds as an organized space
  • generated by a personal team of hacker bots, working inside software hovercrafts to process your content feeds
  • using training courses that are as simple to write as HTML documents

To make this a reality, I’m building this as an open-source initiative, led by a non-profit organization (called the Wachowski groupoid, in honor of the Matrix creators), currently consisting of just myself. I’ve written an initial POC & put up an initial web site for the project. Check it out to learn more & drop me a note if you’d like to join me in solving this problem once & for all.

Screen Shot 2017-04-03 at 6.30.56 AM


Knowledge syndication

I got as a b-day present 2 great books, one is the must-read On Intelligence, & the other is a billingual collection of Confucius sayings. As there are many knowledge extraction posts on the 1st, e.g. Robert Burke's, but less of the 2nd, I decided I must share with the world the wise Confucius knowledge as a Listible list (the other option was a Moodle course on my homepage).

Then I saw in the browser status bar that my NetVibes tab is busy reading (RSS) content from various great web sites, & it suddenly hit me that instead of me listing Confucius sayings in this location, & Robert Burke listing Jeff Hawkins sayings in that location, what's needed is a standard format for syndicating knowledge beliefs. Not just pure knowledge, domain coneptualization as in the Semantic Web, but practical behavioral knowledge, executable specification, that can be used by beliefs-based engines to acheive goals.

I made recently an architecture of such an engine, that is driven by goals & an inventory of beliefs, but didn't thought about the idea that the inventory of beliefs would be the whole web, or at least what's encoded in the NSSKS format (Not So Simple Knowledge Syndication…).

Maybe it's just a subset/dialect of OWL, but with proper beliefs-based machines, it could be something very powerful. Then we'll need also standards for Goals syndication, & the agents will become completely autonomous, doing the work for us without any need to tell them what, when & how to do it.