I recently started following Harold Jarche on Twitter. He’s a name I remember from the old days of blogging and personal knowledge management. I think he may have been friends with John Husband.
He’s running an online workshop on personal knowledge management for 500CAD which although it looks interesting is a little rich for my blood right now.
He talks about how it can make you a catalyst and that resonates with me. He also has a diagram with two axes: high & low sharing and high & low sense-making. And that the meta-experience of knowledge management helps you with both. Again I agree.
It reminds me of the Roam approach of capturing the free flow of work and ideas. Connor White-Sullivan, Roams principal author (I’m not sure how many there are in Roam Research) talks about how he drafts just about everything in Roam first (emails, help desk responses, etc) which allows him to, easily, tap into his past experience of such things. In a Roam document you can tag an individual paragraph rather than a whole note (or, in Mentat, Thing).
This is deeply appealing.
But, at the same time, I have a reflex that says I don’t really want all of this knowledge trapped in a text database, no matter how clever. Machines are not good at working with amorphous blocks of text. How do I apply KM algorithmics here?
Again I am reminded of the question of how the “analog” and “digital” worlds of KM can be interfaced. How can I make use of what I put - freely flowing - into a tool like Roam in a tool like Mentat or Tinderbox?
One option is something like the way Lotus Agenda works. That we feed everything going into Roam also into Mentat and allow it to make inferences to extract things into the Mentat prototype database.
This kind of approach could be a best of both worlds also if there was a way to make this bi-directional.