Computing for the real world

Computing is currently based on taxonomic hierarchies. This has served us well in the past and has suited the Western penchant for classification and reductionism. Taxonomic classifications have allowed us to communicate with a reasonably common voice. They have allowed us to store large amounts of data in a organized fashion. All good-right? Well, yes and no. It is good up to a point and I doubt that Thomas Freidman's Flat World would be a reality if it had not been so. However, taxonomies by their very nature, bury data within a fixed structure that requires you have some idea of where to look before you can dig it back up again. Think of files on your PC. Write your words of wisdom, file it on Sunday, can't find it by Wednesday. Well, at least that how it is for me. I have to be very disciplined with how to file stuff or resign myself to some real quality time rummaging through my files, trying to remember in what frame of mind I was in when I filed it away. If this doesn't sound familiar then I am truly envious but suspect most folks out there are just like me. The more stuff you accumulate, the worse it gets. And imagine if you change jobs and have to re-classify everything to suit a new set of jargon for your new company? Eek. Inside big businesses, I know from personal experience, that days, weeks, months are spent re-assigning names, categories, data, etc, whenever there is a change in environment be it a merger or merely a new group formed. So, what can be done? Computing by taxonomic methods allows huge quantities of data to be stored in fixed hierarchies. The real world is not a fixed hierarchy. It is a flexible mish-mash of interconnected, interdependent networks that includes 'things' (like you and me, the computer I'm using, the desk at am sitting at, the cup I sip my coffee from etc), and relationships between the things. The things and their relationships constantly change depending on the context of the situation. Computing for the real world must account for this contextual variation. It must also account for the fact that the world just 'is'. No forced categorization required. Now we are in the philosophical realm where debates about the 'is-ness' of the world have raged for centuries. I will leave further discussion of the philosophical underpinnings of computing for another day except for the introduction of ontology as a concept. True ontological methods allow a characterization of things based on their relationships to other things, that creates a flexible knowledge network that in turn reflects the flexible nets of the real world. Ken Wilber's Mulitplex (www.integralinstitute.org) has an ontological look to it. The Semantic Web will require a step change in this direction in order to take the web to the next level. One small company in Cambridge UK is already developing ontology based approaches to health intelligence for the drug industry and beyond (www.biowisdom.com). These methods have to be seen to be appreciated. In our current world we have become so accustomed to filing in traditional taxonomic ways that it's hard to imagine how it could be different. However, ontological computing is here and it is growing. It will allow an unprecedented transparency to our real world in all fields which will yield a host of opportunities as well as challenges. I look forward to them all---bring it on!