Dave Snowden - Small Pieces - Fragmented Narratives @ KMWorld

David Snowden of Cognitive Edge open the first session in the Interactive Thought-Leaders track at KMW07. Title of his session is “The Dynamics of Strategic KM. This track is supposed to enable a set of discussions. He was formerly with IBM. He pioneered the use of Narrative. He focuses on sense-making. “A Leaders Framework for Decision-making” is being published this month in Harvard Business Review.

David’s against the off-site knowledge strategy where the team brings something back. He say: Yet to find any example where this may have worked. My notes are messy and incomplete below. He speaks fast and the concepts are “mind-shifting”. I find some similarities in themes between what Jame Surowiecki shared and what John Peterson of the Arlington Institute was studying or implementing re pursing real time assembly of unstructured information (I’ll find and add the link later).

Today you want rapid real-time assembly of data. Link narrative to content. Then there is the story to give people what it means. The stories provide alternate avenues to content. You do need volume. Collect all the fragments. The name for a story is often more significant thatn the content. What are the key words you would associate with this story. Over 70% of the words were in the content. The content doesn’t contain all of the meaning. So there are some limits.

Story re context. High performing crews … share failure. The assumption of context independence if you want o replicate you have to create the same context. Behavior induced in one context can’t be replicated in another. Example fire crews, airline pilots.

Caller center working with the ladies for a water treatment plant.

• Best performer. had a voice that could get to any middle age male. She could close it within two minutes.
• Worst Performer. took five minutes… but solved customer service queries and deviated from the script.

She wasn’t fired. She had her own little knowledge process. By understanding the context of the conversation she had saved the company a fortune. People can share knowledge at the right level of abstraction. They were able to justify getting the call center employees to drive around with the water engineers every three months.

So Where is KM.
At the end of it lifecycle. It is shrinking in Industry. So sell it to government so you will get another 10 years. KM is about improving the way we make decisions. You don’t manage knowledge for it’s own sake. IT contributed hugely to management science as we have had to take account of people. That’s stressed the system and has sent people back to the natural sciences.

Complex Adaptive Systems. Non-causal system.
How leaders manage under situation of complexity.
from fail safe design to safe-fail experimentation.

Bad sign.. Good Customer vs Bad Customer. Grameen Bank. uses the network to make the primary decisions over who should get a loan. (Surowiecki similarlities)

We are pattern processors rather than information processors. This is a key point, for what’s required from an information system. Human knowledge is fragmented not structured. Given the choice get a case study or do you go for anecdotal fragmented stories from people with knowlege. Narrative reviews of the battlefields reveal more that analysis. We search for fragement and synthesize.

RAHS Software: Been building software on these principles. Originally funded by DARPA. RAHS Software.

• Narrative based fragemented database. Completely flat. raw stories blog etc. Handled by a semi-structed tagging system.
• Collective Expert knowlege caputred in ffragmented form within models. You capture knowledge in the way you assume people will want it.
• Collective intelligence, (rejected Jame’s prediction markets). bringing divesitiy into play in decision-making.
• Get them to look for sensing patterns in the metadata. Solves how you get people to share across data silos. So share meta-data across silos.

How might this work for management+ The ability for the CEO to poll without telling them what the decision is about.

David says, We know scenario planning increases risk in the future. If something happens in between them (the scenarios) it may be too late to see something. So scenarios are now being assembled in real-time. I find this a very provocative statement.

How do you give people direct access to the segments.. small pieces of data.

If you take this science based approach. He shows a graphical representation for mass narrative capture. Capturing 1000’s of stories and showing the maps to show where the key knowledge assets are. Moving away to distributive cognition. For liverpool museum. what impact re different stories. If the director sees a negative trend. Here we have numbers with context.

From Lessons Leaned to Lessons Learning.
Knowledge is about sharing. Think blogging. You must be intimate in order to contribute and get something back. Nobody put new wine into old wineskins. Think knowledge flow and interaction.