Re: [ba-ohs-talk] Organic Growth of Knowledge
Malcolm Dean wrote: (01)
> ...you support my point that hands-on expertise is required in this
> scenario. The human factor is the point of failure, and any solution which
> ignores this will fail, at least until we have effective artificial
> intelligence. Producing elegant and powerful programming solutions is
> laudable, but only part of the solution. (02)
Have to agree. It's interesting, though, that the "google proposition"
could well have run afoul of the same arguments, only 10 years ago. (03)
Then, the argument would have been, "sorting based on the number
of links makes sense in theory, but it will take a lot of manpower to
construct those links". (04)
Of course, that manpower has been applied -- but not for altruistic
reasons like, "this will make it easier to identify important pages".
Insteasd, there was a value proposition that made it desirable to
add links to one's own pages. The result is a richly interlinked
that makes the google proposition both tenable and useful. (05)
The same results will occur *if* there is a value proposition associated
with categorizing information. If applying some ontological identiers
gives them a higher probability that they will be found (which depends
on having ontology-aware search engines, which is very much the
same chicken and egg problem as" web pages are only valuable if
people have browsers") then ontological identifiers (of *some* kind)
will be applied. (06)
Similarly, if there is some value to applying to an ontological identifier
to someone else's site, as *part* of the link you construct to it (which
requires tools, as well as ontology), then we may wind up with a google-
style classification scheme where *I* think my pearls of wisdom
qualify as "brilliant, insightful commentary" and the rest of the world
outvotes my assessment by placing it in the "hogwash, bull, and flapdoodle"
bin. (07)
Or, to be more serious, what manufacturer A puts on web page as
"specification:techhnical", others may categorize as "marketing:hype".
(That of course raises the possibility for informational warfare, but
ignoring that for the moment, the capacity for google-voted classifications
remains.) (08)