[hotspot_one]

> I seriously doubt there's very much highly relevant old knowledge locked away somewhere.

Interesting take on what "knowledge" means and what makes knowledge valuable.

If I understand "knowledge" as "information directly relevant to a technical problem", then:

- the knowledge which remains relevant to that problem will stay available to practitioners (i.e. the properties of a Gaussian distribution, from Gauss, 1809)

- the knowledge which is no longer relevant to that problem will probably be lost (how to compute the integral of a Gaussian using a slide rule. Slide rules first developed circa 1620, last used circa 1970)

In other words, yes, your point is profoundly true. Knowledge relevant to a specific task stays available, not relevant gets pruned quickly.

My question would be if we want to use that definition of relevant and that understanding of what drives value. i.e. I'm not asking if you are correct, I've just shown that you are correct. My question is if the assumptions/values which make this correct are assumptions/values we are comfortable with. In other words, is is wise?

[shiroiushi]

It's not just technical knowledge either: look at Shakespeare's works. They're centuries old, but there's absolutely no danger of those disappearing. Lots of old stuff is well-preserved and highly duplicated for easy access.