Interactions, collaborations and breakdowns

What is the point of modelling anything - why don't we simply get on with the job of designing and building interactive systems using our intuitions, creative genius, experience and knowledge. This is an argument often put forward in certain areas. We do use all the above when we design and build systems and we still get them wrong, often in serious ways that mostly only cost in terns of time or money. Models do not necessarily allow us to avoid the mistakes and pitfalls of design, a model is only an abstraction over detail, a representation of something that we believe is of relevance, interest and/or concern. We have to be creative, ingenius, experienced and knowledgeable in our development and use of models in design. Models do not exist for themselves, and people who engage in developing and using models are not doing that purely for the purpose of developing or perpetuating the use of a particular model. In my view modelling is a way to further understanding, and as our understanding progresses so should the models we use. In this respect I want to consider how my understanding of interaction has progressed and with that my attempts to model aspects of interaction that I am trying to understand so that I might use that understanding to improve interaction design. That means that while task modelling has been and still is important it is not a be all and end all.