Institutional Memory

Roger Schank has several interesting articles posted on his site in the Corporate Memory section, which I decided to dive into recently.

In The Future of Knowledge Management, he says that the main problem with KM systems is that they do not copy how real people think and that unlike a person, a “KM system simply gets slower as a result of more information”. He proposes creating software scripts to organize information, but these must be capable of self-modification. I have not seen any systems that really do this well, yet. Schank concludes:

There is a lot of knowledge in an enterprise that can be used to organize new knowledge that is coming in. People understand new knowledge in terms of what they already know. A smart KM system must know a lot of about an industry and a particular enterprise before it starts up. This is hard but by no means impossible. And it is the future of software – namely software that really knows a great deal about your business.

Until these types of systems are available though, I would encourage individuals to practice personal knowledge management and use enterprise social networks to share within the organization. It may not be as elegant, but I know it can be implemented today, with existing technologies and skills that can be developed by anyone.

Algorithmic search filters that can push things out, based on certain criteria are what Schank calls “Information that Finds You”. Add geo-location and you can get immediate feedback on things around you. These exist, but take time to setup and maintain. In organizations, providing coaching and support on how to optimize our software & hardware tools (our outboard brains) is often lacking. Not only is there a need for a learning concierge but also a basic digital concierge, so that we can use our tools optimally. For instance, even doing an advanced online search query is beyond the grasp of most people on the Net.

Schank also writes about the need for a Reminding Machine, which is based on the premise that knowledge is best communicated just in time.

A reminding machine has thousands of stories from experts in various areas of life telling about important aspects of their lives that have lessons about life in them, the kind of stories you might tell to colleagues or to students … In order to build this machine it is necessary to collect people’s stories and index them according to the goals and plans that a story instantiates.

In his keynote at DARPA in 2010, Schank discusses story telling and KM in great detail. Here are some highlights

  • Stories: should be full of details but short
  • Lecture: people cannot think about what they are thinking and listen to the speaker at the same time
  • Stories, to be effective, must not be too abstract for the person listening. Listeners must be able to absorb the stories.
  • Comprehension means “mapping your stories onto my stories”. It’s difficult to communicate with someone who has different stories.
  • In good stories, we do not give answers.

There are 12 Fundamental Cognitive Processes, according to Schank:

  1. Prediction
  2. Modelling
  3. Experimentation
  4. Evaluation
  5. Diagnosis*
  6. Planning*
  7. Causation
  8. Judgement
  9. Influence
  10. Teamwork
  11. Negotiation
  12. Describing*

* These processes are what Schank calls “The Big Three”.

Several examples of the 12 processes are presented as stories in the second video of the keynote.

For anyone interested in institutional memory, story telling, or knowledge management, all four videos are well worth watching. Roger Schank concludes that the most difficult part in all of this is actually collecting the stories. The best people to collect stories from are those who are able to admit that they mismanaged, botched, or bungled something. This can be a real challenge in organizations that do not discuss failure.