Reply #1 to other comments posted about my proposal; February 1998
---------------------------------------------- Thanks for your comment. I may be so married to IT that I wouldn't know the IT pitfall if it tackled me headlong....
At the same time, there is no knowledge problem I've personally considered that wouldn't be solved by having one system that can coherently address *all* available human knowledge. (This from a one who's spent the last 13 years wondering how he can just know everything.) Ask it a question, it finds the answer, if any human has written that answer, ever. It also provide answers to many issues never considered or written about--like finding patterns in ... ANYTHING, and letting you analyze them, connecting logically from one knowledge aspect to another (since everything known is internally linked by is-a, has-a, and belongs-to relationships).
The comprehensiveness of the concept is the only way I've ever seen to reasonably take everything in the library of Congress, the Web, every historical archive in the world gathering dust, and every other database, and integrate it into one sane, preeminently *usable* repository. And that very usability would engender its use as the starting "thinking point" for knowledge creation. It would be a marvelous way to do "what-if" scenarios, for any logical or physical domain--because all are modeled and can be "handled" and manipulated reasonably by the experimenter.
Think of the questions you could ask about historical periods, or current economics, or anything. Everything is modeled--what we know about patterns of human interaction, period clothing, chemical behaviors, group dynamics, fluid dynamics, thermodynamics, psych, finance, geography, physics, etc. If it's all in a single model we may be able to quantify or approximate subjects that we could only describe or imagine before, simply because of the vastness of detail handled automatically "under the surface" in a single query or 3-d representation. Especially since it has to handle and display the changes in its data over measured time.
Maybe if you helped me better understand what problems you don't see it solving, I could explain better what other things it can do, better than what was in the original post. Let the philosopher lead the engineer.
I accept the risk of seeming or being a gushing, naive geek. But building this would change how we manage our data/info/knowledge, totally.
One bucket, all human knowledge, one organizing principle.