looking for the first cregox/random?
here is one of my previous ramblings about what is random?
now i don’t do blog anymore. among many reasons, also because it has no versioning. no true backup.
i do these manifestos, instead. eventually, as usual. chaotically, if you will.
how random is talking about blogging and manifestos in a random manifesto in here? not much, actually…
it’s no coincidence he ends the video talking about “god”.
randomness is basically patterns we can’t recognize. in other words, a sort of intelligence bigger than ours.
because once we become smart enough (and this happens every second both for the individual and the collective) we learn more about new patterns that were unrecognizable before.
with that in mind, let us hail a flying spaghetti monster. or whatever. chaos, perhaps?
on a second thought…
logical true randomness
disclaimer: this is still a huge TK of my understandings. i’ll need to verify some points (and just link to others i’m already certain).
computers are logical machines. they’re the best tool we have to transform pure logic into art. creation.
still, it’s in their nature to be completely deterministic and, thus, predictable. indeed, a software is unable to generate randomness in a practical sense (actually, it wouldn’t be technically impossible to do so). as much as they can look random to us, it’s either because they’re making a complex pattern our brain can’t readily unscramble (but other computers can) or they’re collecting noise from the hardware, in other words, completely outside the logic we program them into (mostly through bugs).
and this very nature can easily lead to a huge mistake: believing computers can’t be unpredictable.
a computer is composed of hardware and software. one never existed without the other. in fact, computers are, and have always been, in practice, at least somewhat unpredictable (given the bugs and hardware malfunction).
true random number generators are no misnomer. they generate real unpredictable numbers, for every and all practical purposes:
- any computer can generate completely random and unpredictable numbers, which not even quantum computers can predict TK.
- flipping a coin, on the other hand, can be 100% predictable TK, even if so many things around it is random. so can dice rolls.
- the hardware used to generate “true” randomness from radiation decay or whatever TK is created based on the belief of what’s random, but by far are not the only ways to do it.
- random.org built a whole business of making random numbers from atmospheric noise because it is truly random.
the misunderstanding here arises because of a few technicalities…
a single home computer device don’t have any of those hardware random generators. but they could still use microphone, camera, keyboard striking, mouse, or finger movement among others to generate the so called true randomness, or randomness that nobody and no machine can predict. but this takes quite some time and effort to generate a single number. and TK thousands of numbers are usually needed to cryptograph operations in the way computing is structured today. reason why this isn’t used.
whatever technique we use today to generate the true number will be time intensive, compared to the speed computers need those numbers for cryptography. thus we use those numbers as seeds. then add the same pseudo number generators algorithms to increment the seed. and that can, indeed, lead to some practical issues in cryptography, reason why it’s argued they’re not truly random. but they are. at least once. because the seeds always are.
a pseudo number generators is like a scrambled puzzle, all pieces faced up. a whole picture is there and we can see it all scrambled, but we can’t humanly see the end result without putting the pieces together (at least enough of them). a computer, however, can put it together in a fraction of a second.
meanwhile true randomness is like a scrambled egg. nobody would ever be able to unscramble it, tell how the egg was exactly or predict how the scrambled egg will look like before scrambling it. and a true number generator can do that. even if it’s eggs and puzzles all mixed together. although we can still unscramble the puzzle.
the trick is that cryptography needs only eggs. puzzles are useless. and in this computer analogy, it’s like scrambling the egg (making the seed) takes a lot of effort. so, instead, we take a picture of a scrambled egg and just scramble the puzzle. by making the puzzle big enough and using different strategies on which algorithms to use, we make them random enough.
yes, quantum computing can break this model, but never the seed. even the fastest most advanced computer would need to simulate the entire conditions around what generated the seed in order to predict the next seed… in practice, it’s impossible.
now, i haven’t even brought up how a software can generate a true random number, in theory… the short version is: ai. just like a single human behaviour can easily be random to us, another intelligent being will, necessarily, be able to be random. the catch 22 here is how to make such an intelligent ai smarter than humans and i would argue that it needs a true random generator to begin with. but this could be a paradox resolved by itself, creating both at the exact same time.
after all, randomness do help math in ways most people (including me) will never comprehend, although to me it’s very clear that’s basically tapping into god. chaos, perhaps? 😁
just something more intelligent than we could ever be. that permeates everything, everywhere, all the time. impossible to tell, since it’s per definition something we can’t recognize. and even if it’s somehow all the same thing, we’ll never know. fits perfectly into the definition of a god, doesn’t it?
in any case, for studying the creation of intelligence, machine learning and all that as already mentioned, it’s a very important and overlooked distinction there. randomness and chaos. a word without a definite consensus on which definition we should commonly use, as evident in the previous link. the suggestion here is to adapt from mythology:
Chaos (Ancient Greek: χάος, romanized: khaos) refers to the void state preceding the creation of the universe or cosmos in the Greek creation myths, or to the initial “gap” created by the original separation of heaven and earth.
it’s true randomness and completely unpredictable. it’s what we need the most when writing neural networks, but most of us keep using just whatever randomness without making the distinction. and in the sum of many operations using plain dumb randomness the ai takes much longer to evolve its intelligence. the connection with chaos.
of course, this is all just my gut feeling, after playing with the basiux. a feeling strong enough to make me keep writing so much about it, years after i abandoned it. even if it may look completely incoherent. just to register it. and hopefully find the right destination in the back of someone’s mind.
thanks for reading! 😁😘