For many knowledge workers a popular unit of exchange when collaborating is an MS Word document, with discussion occurring via tracked edits and comments.

As an approach to collaboration this has drawbacks: even if people use the reviewing features properly it scales poorly. Discussions bleed into emails (which not everyone is copied on), versions fall out of sync, and often someone has to resolve it all manually. Document management helps, but the collaborative aspects remain clunky because the conversation is still tacked on like a collection of Post-It Notes.

However, as an approach to creating an information product it’s great. One might lose version and discussion data, but a standalone information container remains, often in a form that’s easy for newcomers to consume.

So as a unit of information exchange the document is king; but that may be changing.

If you’ve been following the development of applications like Google Wave then you may have noticed the document-centric paradigm being eroded. Words, images and video are all just information assets which pepper the new unit of exchange: the conversation; and I mean conversation not as an abstract concept, but as a thing that can be accessed and manipulated by people and software.

Google Wave makes this shift obvious, but for heavy users of tools like Skype IM and forums this is a change that’s been happening quietly for some time.

Last week, after a long discussion with friends via Skype IM, we realised that since our conversation had not been centred on a document we had two choices for capturing our outcomes: author a memo, or paste the whole chat log into a forum thread. We chose the latter, and I doubt we’re unique – but are we short-sighted or efficient?

A key feature of Google Wave is ‘Playback’, which allows you to re-run an entire conversation (potentially even if you weren’t an original participant). As an approach to collaboration this is superb.

So is the conversation the future king?

Possibly; but if so then it raises questions about how we communicate beyond the original participants in any discussion. Reading the so-called ’scroll-back’ from a conversation may offer rich content, but as an information product it’s vulnerable to bloat, and in the broader context it may scale poorly.

To use a practical example, recall any first day you’ve had in a new job. In the near future will it be a case of reading through a collection of key documents, or scanning an equivalent collection of digital transcripts; and which would you and your colleagues prefer?