250 einsteins

Author: MarkBruns

_posts/2023-10-09-theory-of-quantum-information.md

@@ -9,7 +9,11 @@ categories: emergence
 
 From the experimental implementation to complexity classes, from the philosophical justifications for the [Church-Turing thesis](https://en.wikipedia.org/wiki/Church%E2%80%93Turing_thesis) to the nitty-gritty of [Dirac's bra-ket notation](https://en.wikipedia.org/wiki/Bra%E2%80%93ket_notation) and linear algebraic manipulation on [complex vector spaces](https://en.wikipedia.org/wiki/Real_coordinate_space)... we can look at a text like Mike and Ike, we start to understand that knowledge is way beyond a clever idea and depends on far more than one *einstein* ... it's not one guy discovering fire, or the wheel or even something like penicillin or the transistor -- we just are not that smart as individuals, even though we are still stupid enough to file patents that attribute ideas to one individual ... we are getting smarter about knowledge, but we are not that smart yet.
 
-Knowledge itself is an extremely complex thing resting atop an even more complex foundation ... knowledge is way, way, way beyond the simplistic notions underpinning large language models and vectorized words or symbols. There are indeed individual human einsteins ... maybe one in a million  and perhaps a century ago it still made sense to award Nobel prizes to solo einteins working in obscurity ... but not any more. Nowadays, knowledge ***emerges*** from very large populations of *einsteins* arguing with and learning from one another -- but it's not until after the concepts have been is developed by humans ... human knowledge [including the semiotics of how we represent knowledge] does not exist on its own ... 3 and 5 do not know that their sum is 8 or their product is 15; they certainly have no understanding of little rules we use like the commutative property in mathematics ... numerical analysis is terribly important to humans but ONLY to humans [and their numericl machines]... human knowledge is a matter of contexts, but it is a very imperfect, very partial **artifact** of the sum total of all human understanding ... of course, we often say that artificial intelligence is "hallucinating an answer" ... like a free verse rapster riffing on rhymes -- the AI's product is not knowledge and AI functionality or meaningful work product simply cannot extend past the realm of human knowledge, ie after all, exactly WHO is AI trying to be intelligent for?
+Knowledge itself is an extremely complex *thing* resting atop an even more complex foundation ... knowledge is way, way, way above and in addition to the extremely simplistic notions underpinning the simplifying assumptions we need to make large language models and vectorized words or symbols even remotely computable ... it will not matter if computation power grows one million-fold or one billion-fold ... human knowledge will expand faster. 
+
+There are indeed individual human einsteins -- people that bright are not *that* rare ... yes, it's unlikely that we meet an einstein ... since maybe only one einstein in 10,000 finds her/his niche in the world ... but still since it's likely that einstein's IQ put him the the 0.03 percentile, if there are 8 billion humans, that means that there about 2.5 million einsteins alive today ... and maybe 250 of those people are doing OR about to do einstein-level work  ... and COLLECTIVELY humans are getting even smarter than the smartest humans ... but we are not yet smart enough to understand why humility matters.
+
+Perhaps a century ago it still made sense to award Nobel prizes to solo einteins working in obscurity ... but not any more. Nowadays, knowledge ***emerges*** from very large populations of *einsteins* arguing with and learning from one another -- but it's not until after the concepts have been is developed by humans ... human knowledge [including the semiotics of how we represent knowledge] does not exist on its own ... 3 and 5 do not know that their sum is 8 or their product is 15; they certainly have no understanding of little rules we use like the commutative property in mathematics ... numerical analysis is terribly important to humans but ONLY to humans [and their numericl machines]... human knowledge is a matter of contexts, but it is a very imperfect, very partial **artifact** of the sum total of all human understanding ... of course, we often say that artificial intelligence is "hallucinating an answer" ... like a free verse rapster riffing on rhymes -- the AI's product is not knowledge and AI functionality or meaningful work product simply cannot extend past the realm of human knowledge, ie after all, exactly WHO is AI trying to be intelligent for?
 
 It's similar for us ... we need to know our WHY ... need to think about WHO we're doing this for? Frankly, it's mostly for us ... or the guy like us who will pick up where we left off in minutes, days or years from now. We are not that smart, we were certainly not born smart ... but collectively, we GET smarter by working on the process of getting smarter.