Generally the web knows just two answers to complexity. On the one side simplification and on the other side information filtering. Both strategies are valid and useful but are they really appropriate to the emergent phenomena and challenges of today's global internet?Without a doubt, doubt is indicated for one specific reason: avoidance strategies often lead to defensive behaviour.  The "price" of efficiency of active avoidance within systems is always determined by the necessity of more activity  during learning of new habits. As a result every system has to put more and more efforts ( energy, information, ressources) into the development of innovation which makes innovation more and more unlikely.

Basically and essentially on the web simplification and information filtering are security technologies which should protect us of  "destructive" information. The same moment they are a signs of the inability to cope with complexity or at least they show us the lack of advanced tools to turn complexity into the driving force of innovation and creativity.

Where's the way out of this lose-lose situation?  First we need more information, then we filter more information, then we lose more and more our capability to work with information.  Do we have to build an algorithmic immune system to the digital world or do we have simply to avoid digital media at all? Neither, nor!

Not long ago we were used to call the web cyberspace. We expected it to be one huge data highway. So, in the beginning the web was not considered to become an abstract "literal space",  instead everybody expected it to become a 3D counterdraft to the world itself.  Today, whereever you look  the web is particularly about text, be it HTML, E-Mail, Instant Messaging, XML, RPC, RSS etc. The web is indeed so deeply based on text that it makes us wonder how  anybody ever could believe in the pure multimedia vision of the internet. All the important aspects of the web, content, protocols and algorithms are finally text based. At least this is the common collective impression. The web as a literal medium.

At this very moment and at a much deeper level of understanding of the medium internet we have to realize that these assumptions are simply false. What really happens is a continuous and ongoing process of transcoding semantics into a geometrical space; a constant change of semantic entities into mathematical ones. How many people know that most of the sophisticated search-, similarity- and profiling tools they use everyday are based on term frequencies that are mapped into vectorial space, where relationships are represented by trigonometric functions, especially the cosine. Paradoxically the more the representatives of the semantic web talk about topologies and ontologies, what they are really reflecting on are the topographical aspects of binary data.

I'm convinced that another ongoing paradigm shift is going to make this visible. It's the increasing importance of graph structures which are going to be the central nervous system of the upcoming intelligent semantic web.

The intelligent reasoning proactively done by machines then becomes  some kind of  geometric combinatorics, where computer and algorithms travel along the simulated edges and nodes of polytopical structures to find and create meaning.