Dave says that four blogs influenced today's talk:
Nicholas Carr on knowledge sharing:
What they say about sexually transmitted diseases seems to apply equally well to data in the Web 2.0 age: You're not just sleeping with your partner; your sleeping with your partner's partner. - Nicholas Carr
Dave likes Nicholas Carr because he challenges conventional thinking and provides great opportunities for responses.
Dave takes issue with Hubert Saint-Onge's perspective on collaboration tools:
... an organization should mandate one tool for collaboration, rather than allowing diversity; but that participation in the use of these tools should be voluntary. - Hubert Saint-Onge
(sorry, missed other two. Took photo and will try to return to this.)
Social computing is more about relationships than categories
The audience did some interesting exercises in categorization and classification, using cows, chickens or grass. (I won't give it away - you nedd to be here) Then Dave railed on perspectives that put people in boxes (e.g. Myers-Briggs, left brain/right brain). Argued for the need to operate relationally.
Corporations haven't yet realized the volatile nature of social computing. Does not think we will see Second Life in a few years; folks are already moving to Facebook and even that will go away to something else. The point is that social computing grows and moves at a whim.
On blogging as a social tool.
Blogging punished ego. (Have you checked your technorati score lately?) Blogging is self-correcting because readers click away. Many companies now using blogging as an active management tool. Example of how managers are using blogging to check the pulse of the organization.
Nature of systems
Order; reductionism & rules, deterministic, observer independent, systems constrains agents
Chaos, independent agents operating without constraint studied through statistics & probability
Complex adaptive systems System constrains agents, but the agents always constrain... Principle of "locality" and constant adaptation. Highly sensitive to starting conditions. System level effects of emergent and non aggregative. (You can't replicate outcomes.)
Corporations and blogging.
Corporations are wasting their time trying to structure and control blogging.
Blogging is meant to be distributed cognition. It's a social tool.
Manage within boundaries, within attractors
Need to create an environment and see how people play. We don't mind if some things do not work as long as we can identify things that do.
We need to manage an ecology (e.g. a social network) to stimulate conversation. Not control the behavior.
Everything is fragmented
The human brain does not make decisions based on rational logical application. (Except people with Autism, who in fact, analyze the patters available to them.) Instead we evolve decisions based on rapid pattern recognition.
Humans are not information processors, we are pattern recognition processors - we recognize patterns based on partial data sets. We make decisions based on incomplete data. That's why people like Google searches - it's fragmented.
Knowledge sharing culture vs. Knowledge sharing ecology?
The issue is not how to create a knowledge sharing culture. It should be how to create a knowledge sharing ecology.
Social computing allows us to do this more easily.
Approaches to metadata
Traditional: professional or author created
(Librarians know more about classification than IT will know in their lives)
Removing ambiguity.
User created: folksonomies, etc,, heavily based on keywords.
Keyword tagging (Del.icio.us & Flickr) work best within boundaries
In effect, these are classification systems "flat name-space" nouns without grammar or agreement.
Browsing not finding. We are moving away from browsing to serendipity. In an uncertain world, we want serendipity, not structure.
Issues:
Scalability - migration from the techies & early adopters.
Individual focus - clustering and village idiots (Funny story on how village idiots assemble into CoPs.)
Semi-structured tagging
...
Not just location and time, its also placement on models, tags can be system of meaning.
Dave recommends 5 types of tags
- Names
- Options
- Key words
- Free text
- Filters
Surprisingly, keywords associated with the content are not found in the content. People use words to add meaning.
Fascinating slides on using visualization tools to see the impact of stories on decision-making.
So how does this relate to Knowledge Management?
Don't roll out CoPs, create portals etc. Etc:: there are no recipes. Much success will happen by accident.
Dave would not do anything. just install blogs and Wikis and freeware. Then get key personnel to get senior people to blog. Offer to spend an hour or two a week to help folks get their daily blog up; amplify and dampen as needed.
Dave's formula for large complex documents:
- One month of free form blogging on topics
- Three people summarize blogs and preload Wiki
- Wiki available for people to amend and change.
- Publish
Start using narrative, linking, and connecting.
Stimulate networks, use natural numbers (5, 15, & 150) (5 is max # of things we can remember; 15 max number of key social relationships, etc.)
Get radical
Ban attachments on email. No reason for attachments if you have a proper repository; use doclinks.
Implement search engines; only allow hot links.
Restrict firewalls
Humans respond to variety We don't expect folks to dress the same to work, why do we expect them to use the same collaboration tools?
Three huristics of KM
- Knowledge can only ever volunteered, not conscripted
- We only know what we know when we need to know it
- We always know more than we can say and will always say more than we can write down
- Dave Snowden
Dave is known for his insightful, direct, and sometimes colorful, approach to calling things as he sees them. This keynote did not disappoint. It was excellent!
Discussion/Comments (0):
Discussion for this entry is now closed.