Foundational Epistemology -- part III

What beliefs predict the future well?

1. Existence doesn't make sense as a word without its meaning that the external world exists.
2. The past resembles the future.
3. Induction is the path to (predictive) success, and along with brain-structure, the source of all mind-content.  
• The concept of number, for instance, is gradually learnt, and only by looking at THINGS.
• Verbal (not-mathy) deductive logic can usually safely be assumed (99% success rate) to be non-predictive.  
• Most guesses/hypotheses/conclusions (whichever you call them) are wrong
• Most useful truths are statistical.
• The only reliable path to success is:  Guess, measure, confirm failure, repeat N times, eventually fail less.
4. Other consciousnesses are basically like mine.
• Especially vice-versa.  To predict the self, observe others...introspection sucks.
• The aggregate of other (weighted?informed?) opinions is FAR more likely to be right than mine.
• Weak EMH, alternatively Adaptive Markets Hypothesis.
• Turns out social epistemology predicts incredibly well (compared to many other options) if you're within a couple sigma of average intelligence + wisdom.  
5. Minds makes sense of the world by creating patterns/heuristics from far too little evidence.
• Probably the most useful patterns are Categorization and Math. Categorization may even be built-in to the brain.
• Probably the most important set of categories to life is: Inert, Alive, Intentional, Conscious
• Maybe the 2nd most important set of categories is: Inert Constructed vs. Inert Not-Constructed. 
• Most patterns you see aren't there.
6. You need a LOT of samples before your intuition is worth more than dog poo.
• See: Tycho Brahe
7. Complex Adaptive Systems(The God of the feedback loop) may be the most useful explanatory pattern ever created.  Mechanics is awful good too, but applicable to so much less of the interesting world.
• CAS = Evolution, Economics, Lean.Systems, OODA, Artificial Life, etc.
• The key insight here is that thoughts around purpose are almost always wrongheaded.  Think process, not purpose....and get results that look purposeful.
• Given that humans and human brains evolved, the uniquely human brain is no less a product of evolution than the 20-odd different evolved eyes in the animal kingdom...What reproductive success arms-races did the existence of large brains win, despite the large biological cost of said large brains?  I can even partially answer this one.  It wasn't better prediction or better mammoth-killing.
• Hyper-important issue in evolution.  Inside a species, almost all evolution is intra-group evolution.  How did proto-chimps with bigger brains win the reproduction contest against their stronger, stupider herd-mates? 
8. An awful lot of questions are broken, for dozens of reasons.  Picking good questions may be the hardest of the thinks.
• It is not obvious, for instance, that "Why?" is a coherent question much of the time.  Usually, the question requires clarification....much of the time it assumes an answer.
• Among the best(?!?) bad questions are ones which start of with "What is the meaning/purpose of".  Meaning and purpose are properties of intentional beings...and so meaning and purpose questions are almost always missing a "to whom".  Finish the "to whom" and the question frequently becomes answerable.
9. Hypotheses with no predictive payload do not deserve much prediction brain-time.
• "I believe that elves paint the moonbeams every night"