I suggest that dissipation, in human settings such as "organisations", occurs in participating in ordinary, everyday conversations. For example, when we contact customers, the practical reason, the economic reason, might be to inform them of the current state of their accounts or to give them information about product specifications. However, this is usually not all we do. In addition we engage in talk characterised by redundancy from the point of view of economic value or business purpose.
Through these redundant communications, we acquire information about contextual variables, such as customer idiosyncrasies or customer intentions. We get information about competitors or about possible technological developments. This information however, is not purposefully sought. At the time, we do not know what to do with it. We did not intend to get it. We did not have any instrumental goal in mind while we were engaged in such ordinary conversations. It is not "knowledge being shared" since we do not use "it" to engage in some interaction. In fact, some times we simply regard it as a waste of precious time, or as politeness towards an unpleasant business acquaintance...
Most of the time, in ordinary conversations, we face some ambiguity and we sometimes have to probe for the meaning of the words pronounced by others. This happens because we engage in conversations using the pattern of talk that is pertinent to our own local interactions and life experiences. While others use different patterns of talk that have been developed in their own local interactions, as we together pattern our experiences of being together in the living present.
It is because each uses a pattern of talk referent to their own life experience that the potential for misunderstanding occurs. I suggest that what is being dissipated in conversation is this misunderstanding as people use their patterns of talk to negotiate the meaning of ambiguous, uncertain and ill-defined current or prospective events. Different patterns of talk interact to produce misunderstanding, that further interaction seeks to amplify as the misunderstanding is dissipated and the temporary stabilisation of new meaning, the dissipative "structure" of meaning, emerges. I want to argue that when the level and quality (potential for misunderstanding) of redundant diversity in a conversation reaches a critical point, usually because different patterns of talk are interacting with each, the potential arises for new patterns of meaning to emerge.