NOTES: Hythloday On Penrose

• he's not saying it can't be proved AT ALL, the point is G(R) can't be proved with R but it has to be true if R is true.  So G(R) is true but not provable BY R.  That leads us to what can prove it?  How do we know it?
• "only one technique fails"  I know, but the big mystery was if we could make such a proof WITH THOSE RULES, and the answer is determined.
• elevating human understanding to another realm... dude, I am elevating LIVING understanding if anything, and I don't take myself to be doing even that... I'm seeing an unexplained phenomenon.
• The brain uses Heuristics.
• Does the brain do differential calculus... a turing machine has to do that to simulate the behavior
• Obviously there is a calculus library of some sort built into our motor functions
• Mr Calculator
• Not heuristics, with ESTIMATES and real time correction
• Calculations are occurring by circuit
• a computer doesn't really DO a calculation, the circuit flows in a way that MAPS to mathematics.
• does that look like I'm going to catch it?  look like?  visualizing a solution for the data you have had
• I notice you describe human trial and error a lot like a computer does calculus
• quantum mechanics and the efficiency of photosynthesis
•  why do you feel newtonian mechanics relate to our everyday life?  well, it's just we have a different feeling about this
• true understanding of the high level requires understanding why the sublevel behave that way.
• everything we've discovered at low levels is by noticing the affect at the high level, e.g. glass is transparent
• the chess example, like the computer is smarter?  we understand induction itself.  the stuff you talked about recognizing is something we can't make computers due with standard logic.
• the heuristic approach to AI is not novel.  Humans infer from past behavior better than computers.  It's our wiring.

• it's "natural" and "chemistry".... chemistry is described by quantum mechanics... that's what QM was invented to explain... the atom!
• when you put down the microbial phenomenon, you put down the human one as well, imo
• a very small qm computer can do amazing calculations, solve linear algebraic equations, able to do image recognition and other hard AI techniques much better.  coincidence?
• a good heuristic system NEED qm... quantum computers are good at this pattern/heuristics
• clustering maths, qc is good
• all explained in the realm of the physical?  why do you talk like quantum mechanics isn't physical?  even if it's explainable classically, IT HAS NOT BEEN, so keep looking in that toolkit as much as you want!

• theories exist that cannot be proven... and they are needed to prove R
• one way I heard it is that it proves that in second order logic there are unprovable true and false statements, so it's not complete, and you need second order logic to establish the rules of arithmetic.
• math cannot be undermined... it's useful, we are not justifying math, we are trying to explain why it's so fucking useful
• it's not that there is no method to build them, it's that they have to be built.. they can't build themselves... no sui causus, no A PRIORI and among other things, it means all truth is actually synthetic, and analytics just exists to analyze synthetic truths. 
• That leaves one little problem, math is incredibly reliable and has great fidelity.  
• The way I see it, math/logic is designed to introduce very little distortion during analysis, all the distortion comes importing terms, where one problem is the sort of things you have to ignore to pretend the material objects you are analyzing are the analytic objects you'll be processing.  You have to map these into and out of the logic.
•