what secrets remain in the advantage to the scaling process? and what tradeoffs exist between the complexity of the machine and its ability to solve problems?
these things arent making sense.
im thinking about the difference between a single neuron and a small group, and the difference between a small group of neurons and a coherent brain, and the difference between a coherent brain and a conscious brain? and the difference between a conscious brain (a single human) and a group of consciousnesses? or a group and a group of groups? its like what sort of differences exist between binary and assembly and higher level languages? what sort of differences exist between a single computer and many?
she is the new word for amazing.
there is a relationship between the number of steps that are involved in solving a problem, and the number of steps that a machine takes to solve them... what sorts of differences in running time do you get when you use various different finite state machines? what sorts of advancements might be waiting by a systematic reapplication of this complexity-increasing process?
i need less time to consider 'what if'
i think ill need to learn a lot more computer science before i understand what these questions really mean, and im fairly certain that most of these quesitons are probably stupid or illustrate a lack of understanding on my part. which is why i type them here.
so i think now in order to make A.I., we just need John Koza to build a much larger, much faster machine. and have it use answers from previous problems to help it solve new problems. basically, allow it to store its output, and use any solution from previous output to solve new problems in intermediate steps.
every girl he looked at wasnt her.
wow, suddenly math feels identical to the Chaitin's constant problem stuff relating to complexity and information theory. its as if new mathematical processes are almost just shortcuts, or more compressed previous solutions. so we learn rules about multiplying in order to save ourselves time with repeated adding. previous to the discovery of multiplication (or invention?), adding took much more time.
it wasnt an easy thing to assume, as most would expect.
when i wrote most of that, i was wondering, how could there be no fastest algorithm for multiplication? is there an upper bound for the fastest algorithm, and we can maybe never reach the upper bound?
george foreman: "interesting side note: as a head without a body, i envy the dead"
rich little: "no argument here"
if you could implement Grover's algorithm, itd be really beneficial to AI, because itd enable you to have much larger databases, which are good for memory. though now i realize this is silly, because itd require a fast quantum computer to implement in a dramatically useful way. though maybe im wrong, who knows.
i am the 'me' in 'team'
Scott speaks of 'automating insight' given that P=NP. does John Koza's computer do that? or really does it illustrate that insight is not as special as we previously thought, still weaker than NP?
(the next morning): ah, i suppose Scott stated that insight would be trivial, which is not that same as automate-able; as trivial tends to imply simple, automate-able is more akin to possible (vs impossible).
"im a rich respected doctor, with many surviving patients."
if you cant explain it, then its not science. and if you think youve destroyed some fundamental tenant of science, you damn well better be able to explain what youve done, in detail. if you cant, how can you be so sure youve broken new ground? how can you be so sure that youre incomplete knowledge of the process doesnt hide the answers that fit it back into science?
none of this is in our control.
what we need are bombs that can kill ideas.
ill let you be in my dream, if i can be in yours.
you can always tell if i slept well or not by whether or not my voice is deeper.
suddenly, i am very excited about my job; i could maybe shape it into something i enjoy very much. for some reason (probably mentioned previously), today, right now, i feel 'on top of the world' so to speak; even though im here alone.