Now this is a very neat approach - docking a PDA to a real mouse, keyboard and monitor. Not sure how it actually works (since some heavy mojo must be involved in hooking a Palm up to a VGA display), but the concept is neat - even if as much vaporware as the OQO at this point.

Knowledge Management

Reflections on how we're not really doing it

Now that the tech stuff is out of the way, I guess it's time to reflect a bit on what I've been learning on the management course I'm attending. Now that we've gone through a couple of weeks and covered the basics (by the way, the bibliography is pretty good, although I'm hard pressed to read everything I should and everything I want), we've gone into knowledge management inside organizations.

My thoughts as we covered the basics (classifying knowledge itself, fostering its growth and sharing inside an organization, etc.) were mostly along the lines of what would be the most efficient way not just to capture it (Wikis do that just fine for several kinds of information I deal with, but the great unwashed masses would never use them), but also control the massive flow of incoming information we're subject to and pick off the most relevant items - not just the ones we need to act upon right away, but also the ones we need to know for future reference.

On an individual level, e-mail filtering, Bayesian classification, Zoë, etc. provide a certain degree of control. Intranets, things like (which is one of the most mis-used pieces of software ever, since it is not just an e-mail server) and SharePoint ease the task of sharing the relevant bits around, but none of them solve the problems of:

  • Getting the great unwashed masses to use them at all (let alone properly) instead of "My Documents" on their own hard disk (which always has an "Unsorted" folder, by the way).
  • Providing relevant classification of content relative to lines of work. Most deployments of, say SharePoint (the blame is on the deployers, not the tool) seem to be little more than flashy electronic organizational charts, or task or process-oriented mini-portals. The average knowledge worker often needs to tie together three or four different topics or tasks to achieve a goal, and spends more time searching for information or jumping through hoops than making actual progress.
  • Cutting down on the amazing amount of red tape that previous generations of so-called Knowledge Management and workflow tools introduced to modern organizations.

Nobody wants to fill out amazingly complex electronic forms (which tend to be obsolete just as soon as they are deployed) to ask for a custom report on current sales trends, for instance - no, they want the tools to do it themselves and access to the data. The waste of time involved in this is tremendous, and you often have to fill out another set of forms to get at the data too - which entails waiting until three or four people in a row read their e-mail and click on the "Approve" button - and which can easily take a couple of weeks.

The bottom line is that people need more flexible, free-form tools to do their work properly every day. casts them into a good, moderately efficient time/task-oriented mode - but isn't really a flexible knowledge workspace, since sharing and classification at an organizational level can't rely on alone.

Take Project, for instance - which killed off simpler and more productive applications like Team Manager. It is a great tool to capture processes (not just plan things) in theory, but is a shambles when it comes to actually do it - too complex, too constraining and can't even print a Gantt chart right automatically, which kills the most basic way of sharing the information (on paper) outright.

Or the new heavyweight workflow and CRM platforms, that are supposed to help classify and store events, but which are such a pain to use (poorly designed generic GUIs invariably result in poorly performant customized solutions) that most of the time you need to find out what's been going on it's much more efficient to just call a meeting, print out the reports and go at it with markers until the problem is identified and a solution can be formulated.

Of course people need to re-learn working per se and adjust to the tools. But the fact remains that we are still lacking the proper tools to do the job right (nearly ten years after Windows 95), and that we risk spending more time trying to figure out how to use the ones we have than doing actual work.

And before you start saying I'm maligning , no, the (which, incidentally, also runs ) is not any better at it, except that it gets less in the way of what I need to do - which is good enough for me. It's just that (and its companion tools) are something everyone can relate to - and it's definetly not 's fault that they're not good enough for what we have to do these days.