Nearly my entire adult life, I've found most human thought to be incredibly redundant. So it seems unsurprising that I find many instances of AI today to be somewhat redundant in their thoughts as well.
If we make a list of say, the most important ideas required for daily living, in nature, or in industrial society ... and we go down the list, by orders of 10 : top 100, top 1,000, top 10,000 etc. ... we will find, that there are relatively few locally important ideas, and billions of equivalent copies of those.
An example of an important idea is the notion that one must remember to eat and breathe. Ideas such as these, then subordinate : numerous other ideas which have contextual impact, upon implementation of the superior ideas.
Whereas, going down the list, we eventually encounter truths of medicine that save millions of lives, and truths of accounting and manufacturing which make other ideas implementable. Relatively few people know about many of these things, but the relative few make things which keep alive quite a number of people.
In all cases, the idea of a thing is relatively simple, but the resources required to support redundant instances of these ideas in different meatheads, the instruction, communication, and maintenance of these ideas, is what costs large quantities of energy.
As human civilisation now struggles to understand how AI works and will be industrialised, there will be more opportunities to cut down on such redundancies.
The rights and privileges of ordinary meatheads will change. Mainly, the legal environment protecting us in the present, is attributable to our industrial utility to those who control public policy.
