date/2010/10

north eastern stick dancing and pipes

after a short break, let’s continue our series on things. in my previous posts i solved the subject-object problem by claiming the subject is recursively made of objects, and that subjectivity is no more than distorted copies of real information. this was only possible because of my axiomatic opinion that things are real. once i get to minds, subjectivity will become clear by its physicality. you are obviously entitled to make your reality any_real, like _surreal or irreal. doesn’t matter to me, since your thoughts, in this line of thinking, are things too (real things even). this is enough to make a lot of people angry, so i’m moving on.

we had our miniverses, macroscopically different. one with a vocabulary of two (a, b), and one with a vocabulary of six (a, b, c, d, e and f). so now we can define our vocabulary in our own miniverse (the one we live in). this is the quest for the elemental alphabet that we spend so much money on (your bosons give me hadron?). but depending on how deep we want to go, we might not need to know the core letters.

this is what we must ask ourselves every time. where are we drawing the boundary? am i concerned with brownian motion? hydrostatic pressure? rainfall on your wedding? how deep is my zoom in?

let’s try to formalize this question in simple terms in these steps: – what information are you doing work with (how your thing is sufficiently quantifiable, with your chosen distortion or error); – find the simplest alphabet to represent that information, given all the alphabets you have (you can’t use alphabets you don’t have).

an example on how this works. if you are doing simple construction work and want to think (or imagine, or project it), what is your alphabet? let’s take building a simple wall as an example. let’s imagine i don’t want to make them fancy and anti-seismic, rather the old school kind that always kills people at some point. of all the possible letters i could work with (atoms, molecules, chunks of clay, geological formation of rocks, etc), i will choose the simplest amount that allows me to do work within my boundary (this boundary being the one that has my little construction yard inside, and everything else on the outside). my letters are bricks, concrete, the laws of gravity and static forces. why? because to describe my activity (the work i’m doing), this allows me to save and retrieve my project with no loss of information (sometimes the type of brick isn’t important, nor the type of concrete) plus it can be encoded in the smallest chunk of information (occam’s razor). imagine a piece of paper. you could use that piece of paper to write

put bricks alternating on top of each other with concrete between them

or you could use that same piece of paper to write

start a universe make a lot of stars explode and find one that has a good planet have all these things happen to it so we have clay and walking monkeys and concrete and teach them to carry the things and make them build things

in fact, if you only wanted to use core elements, you could go as far as

quark id#1 move to a quark id#2 move to b quark id#3 move to c quark id#4 move to d quark id#5 move to e (…)

in a huge sequence. it’s easy to say that if you choose a deeper alphabet, you might not end up with a simpler explanation. so saying F=ma is simple is a bit naive. it depends on your subject.

so what we’re describing is that activity (work) is what makes us choose our alphabets. for a culture that never deals with earthquakes, a brick and concrete alphabet is enough to do work in their reality (they don’t know earthquakes are real because they haven’t observed them).

again, this is where we specify. every human activity has its own jargon, its own alphabet that is used to accurately do work. that work can be analyzed quantitatively by looking at the thermodynamics of the system. what is generating information, what is destroying it?

so in a couple of lines i jumped from tiny things and miniverses to construction work and made them seem similar. this means that this model is fractal, since it repeats itself in different scales (or zoom factors if you will).

in my trade jargon, you’d call going from lower level to higher level descriptions. obviously everything i write is highly influenced by my trade. again, once we get to minds, it might become clear why. lower level is no more than fewer agents and more data or more descriptions and less models. higher level is the opposite. as you go into more and more abstracted realms, you will have more models and less descriptions.

for example, you have 5 marbles, each in a hue of blue. if you need to telegram this to someone (they call that tweet these days isn’t it?), you could write:

i have a duke blue marble, a federal blue marble, a navy blue marble, a sapphire blue marble and a prussian blue marble

or you could write

i have 5 dark blue marbles

which one do you prefer? one uses more description, the other one uses more modeling. the closest to reality is the descriptive one, since the information has less distortion when compared to the real marble. the smallest one (cheapest one) is the modeling one. they will be equivalent when all your marbles are exactly dark blue (there is no difference in the distortion in both telegrams).

it will always depend on what kind of work you’re doing, and whether you’re a science major or a liberal arts major. literature for example is beautifully descriptive. it can also be a bit exhausting at times. science on the other hand, is beautifully abstractive. it can be a bit unintelligible at times. anywhere in between you can find an expert and a smart ass. that’s how we work, and how we make ourselves feel superior to others.

so is there a more real perspective? the realest perspective is to accurately observe and describe all states of all things at all time. this is impossible. so we can induce from our agents all states of all things at all time. if the distortion between reality and our induction is zero, we have found our theory of everything. this is my version of the formulation of the scientific method.

again, there are some problems that have to do with computation, which i briefly described as “how much you’d have to write on your paper” (we can call that paper the tape on a turing machine), that everyone trying to use models instead of descriptions will have to address. also, due to our light cones, we cannot verify global minimum distortion of our models, only local (even if local is our whole light cone).

so my view of the scientific method, summarized, is the very difficult task of creating a set of agents whose generated thing space has zero distortion relative to the real thing space.

remember the miniverse? our agent 3x(a,b) was our theory of everything. it could generate our miniverse with zero distortion. we could say everything there is to know is known from a model perspective. but we would not know all the arrangements themselves without generating them, which would require, you know it, a lot of paper. in the end, we would be able to telegram our miniverse in a tinier piece of paper, but we would not know the lives of mr. a1 and mr. a2 and how they lived their lives. to do so, we would have to generate their lives from our law. while possible, if you have a lot of letters, it quickly becomes intractable. so we would do:

3x(a,b) = (aaabbb),(ababab),(…),(bbbaaa)

this simple miniverse has only six letters but already 6⁶ ways of arranging things! try writing that in a piece of paper! this is the obvious intractable issue that i fail to have science types understand. but we can see how powerful laws are. ~7 characters (for a mind) can generate 46656 different arrangements of things. it is a compression of 5 orders of magnitude!

math disclaimer: i used the cartesian product of sets (hence 6⁶), not the arrangements. i like to think ababab is not the same word as ababab, because two a_s are not the same _thing, instead they can both be saved and retrieved using the same information (they are redundant). it’s different. but whichever you choose, the compression rate is astounding.

so do i favor models or descriptions? depends on what kind of work i’m doing. so i choose both to signing up to any. but already we can see that reality isn’t black and white, but rather a complex spectrum, like that mutant onion i was talking about, and that it is possible to compress things enough to talk about them, but not to fully represent them (as of today’s knowledge).

now that we’re done with understanding different layers of abstraction and how naturally the scientific method emerges from this way of thinking, we can begin our journey from the objective to the subjective. that will be for the coming posts.

some more gaita from the northeast, including the portuguese harmonic 3 hole flute, and the local explanations of the artifacts

continuing our sequence on things, it should be clear by now that there is a strong ambiguity going on in defining agents versus things. so let’s clarify it. there are no agents, there are things, a lot of them.

let’s take this example to see how a core agent can emerge (the theory of everything of a universe). let’s make a little universe with 6 things, all different, and another, with 6 things but redundant. i don’t want to dabble in “sets of sets” questions, backwards recursion, and so on, it would lead to halting, incompleteness and many other interesting issues, but not practical. as i said in early posts, i take existence as a fact. these problems still exist, but in the realm of ideas, which we are far from approaching. if we live in a real world, then i am imagining real worlds, not hypothetical logical inconsistent worlds, even though they might exist. it doesn’t matter, because i don’t live in one. i can’t tunnel effect through my wall or wave into two people at once. i wish. so let’s begin.

miniverse m: let thing t in miniverse m be in {a,b,c,d,e,f}, arranged in any way;

miniverse n: let thing t in miniverse n be in {a,b,a,b,a,b}, arranged in any way;

on one hand, the possible combinations are the same, because the miniverse doesn’t have a mind and can’t read. so in n, {a,a,a,b,b,b} is just as likely as {a,c,e,b,d,f} (remember letters exist only in our alphabetical minds, for things, they have no clue if they are a, b or c).

but there is a main difference between the two. whereas m needs the 6 letters to be saved and retrieved, n can both require 6 letters or be seen by a mind as an agent acting on things, being this, for example 3 x {a,b}, where “x” means repeat, and “3” means how many repetitions. obviously these two are part of the tools (math) of the mind watching, and not part of the miniverse n. remember n is {a,b,a,b,a,b} just like m is {a,b,c,d,e,f}. neither has a clue of what a letter is. but if a mind exists that knows math, it can compress miniverse n, but not m. note that this compression is no more than a statement of redundacies, or repetition, in things, by observation. the things themselves don’t know and don’t care if they are in m or n, remember they are only letters. but this is only possible because a mind knows that a is the same as a, by observation. for the things themselves, it is impossible to say which of the 3 a_s will be chosen as the central _a. what we’re assuming is that in this miniverse, a single alphabet letter contains all information about the thing. therefore, reading and writing a in sequence over and over is the same as making a universe with genuine a_s from the start. the “3 x {a,b}” expression is an _agent, and it exists as a thing in observer minds, but it doesn’t exist in n.

so this is how i fix the recursion problem. i don’t consider agents real until they themselves can be things. why? well, take this miniverse, there are not enough things to make an alphabet and a fully working mathematics system. so tough the redundancy might be observable from the outside, it isn’t by the things themselves.

so we can say fundamental laws (core agents) only become real once they become things themselves. in a way, an atom can’t appreciate that it’s an atom, but a lot of them combined make a mind capable of understanding that there are only so few atoms around (or letters in my example before).

this demonstrates that in order to fully represent n or m, we can both take a descriptive approach (as i defined the sets), or an algorithmic approach (as i defined with the operator). the two are interchangeable since the set itself is always the same and independent of whoever is observing. this is why reductionism is correct and wrong at the same time. we can use algorithms or laws to describe things just as we can just list them by order, and that gives us compression (saves us time, for what it is uncertain). but this is only possible while the semantics available (the alphabet) is sufficiently rich to save and retrieve all information. that’s why science is working out so well, and why the laws are so good at explaining things (when we explain, we retrieve a copy of the arrangement of things). mathematics is an excellent compressor of data and saves us a lot of time.

but this brings us back to why reductionism is wrong, which is also given by physics itself. even though we can retrieve copies of other things, if they were predicted based purely on agents and not observed directly, it is unwise to say the predicted things are real. they can be real, and that tests how good the agent is. but if we go back to the miniverses, both n and m are real, and they both consist of 6 elements. n isn’t made of 2 elements and 1 agent. it is made of 6 elements, period. and since observation is limited to the observer’s light cone, there is no way of knowing if in a very big miniverse of {a,b}s there isn’t a {c} lurking somewhere.

this speaks only to the limits of observability and agents. it will always be arguable if there is a way to accurately save all the information in something in order to have valid agents, since someone who chooses a spiritual explanation might add non physical things to things, which is legitimate. if we choose that atoms themselves have non physical qualities that need to be saved, which is a legitimate question, then accurately saving and retrieving them is impossible.

the way i approach this, and my simple answer, is that if it can’t be observed, it doesn’t matter. the real for a mind comes from observation. this act of observation can then be refined by improving our means of retrieving information, but is in no way perfect and is highly biased by what we can do as things in our own world. for human minds it gets worse, since the mere act of observing is subjective and depends on previous observations. confirmation bias, change blindness, i could go on. they all remind us that it’s a rocky road to observe without distortion.

so just a quick definition. observing is having the information of a thing fed onto another thing. this is a copy, it can be lossy if information is lost (what i called distorted), or it can be complete if all information is transferred. if redundancies exist, this information can be transferred in terms of both smaller things and agents, or the entire set of things to be copied. usually, when we have restrictions in space or time (our case), we tend to prefer a compressed version of information, rather than the entire data set.

in a way, we can say it is real if it has information to be observed. only real things provide information. and agents become real too, when they exist in minds. but they are a part of them, and don’t exist alone (like in our miniverse). this is where i usually piss off science people. i already lost philosophy people at the sight of sets and curvy brackets. but since this isn’t a popularity test and i’m not trying to convince anyone of anything, i’ll just let this sit as another story.

apropos: pipes made of stuff

this is just a short post not connected to the recent stream about things. if you read this far, it should be easy to understand a common mistake i hear more and more from westerners. more and more people are considering themselves minimalists, saying they only need their computer and/or cell phone. but if you go back to the posts about work, the work required to build a cell phone makes it a bit ridiculous to claim it as a possible minimalism.

even though all we’re holding is a cell phone, to have it we needed a whole chain of production and marketing. we need the mines, the miners, the engineers, the designers, the salesmen, the transport system, the globalized economy, the hardware and software designs, the job to get the money to buy the stuff, which in turn requires an economy to work.

so saying we can just have a cell phone, or any other modern artifact, is the same thing as claiming we just need an entire capitalism globalized market. that is hardly minimalism.

until we can build our own chips and hardware at home, using dust or dirt, there is no way one can be a minimalist. and even if we could build everything ourselves, we would still need the knowledge to build the things, which would come from a society, and probably after a lot of work done by that very same society.

another one i love is how minimalists these days easily dumpster for everything, but still have a laptop and an online connection. the online connection alone is a massive energy sink and requires an immense telecommunications infrastructure. i know when you use your cell phone, you don’t “see” the cables, so it’s like “magic”. but nearby i bet there is at least one cell tower with a broadband connection, and plugged in to the city main electrical grid. there is nothing “independent” about a cell phone except that you can’t see the wire. the same works for everything we do. even if we dumpster everything and use nothing from the main economy, we are still using byproducts of that economy. if capitalism magically “disappeared”, so would dumpstered goods and broadband.

unless the items we use last or made by ourselves (including the mining), it is pretty much useless to dumpster, because the engine is still moving. how long do your items last? we have technology to make electronics (and other items) to last forever (lifetimes). capitalism just doesn’t work that way (see planned obsolescence). the article is highly disputed on wikipedia, but i learned that same technique in project management. it’s just not widely broadcast because it might make people realize what’s really going on.

you make light bulbs. thick filament and glass means it’s more expensive, but that it will last 400 years. you sell a batch to all citizens. you’re out of business. that’s the essence of planned obsolescence, which has been around for a long time (even since henry ford days). there is little extra cost to make a lifetime thing, but there is a big cost to stay in business afterwards.

once we moved from a culture of making things to a culture of making money, lifetime goods are no longer an option.

when was the last time a cell phone lasted you more than 2 years? a tv? a computer? my laptop failed exactly 1 year and 11 months after i bought it. i could call the warranty just for that 1 month, but it’s a bell curve around 2 years, so most fail one month after anyway.

management likes to say this is a conspiracy theory. despite the evidence it isn’t, let’s be a bit naive and just say we’re doing it only for the money. the cheapest components don’t last as long. so the effect of saving money is equivalent to planned obsolescence. electronics break down when the worst part breaks down. some computers can break down because of a fucking cheap capacitor. so even if planned obsolescence wasn’t a management technique (false statement), profit driven management also guarantees planned obsolescence.

so how can we be real minimalists? an easy one would be becoming a hermit. another one is building your own infrastructures and hardware and producing your own goods. this is impossible for most things at this point, since electronics requires a lot of power and rare minerals. it’s easy to see how being a minimalist requires either having nothing or having a massive infrastructure to support you. either way, you’re left with joining society or building your own.

a more practical way of avoiding such traps is reducing consumption, fixing and recycling. but you can’t fix or recycle everything. until you can, you can’t be a minimalist in a western society. you could be if all of a sudden everything you had became self powered and “payed off” the energy necessary to build it. from then on, maybe. but your hardware would be so obsolete you couldn’t plug in to the rest of the world, making it useless. planned obsolescence always wins.

our world is moving fast enough to be unstoppable. i would say jump off, take the speed you got from the ride and start something new, like we’ve been doing with the places we built. but don’t fall for the trap of believing you magically became “sustainable” or a “minimalist”. that requires the entropy math i just did. where are your things coming from?

another disconnected post, and some lightning pipes.

as the web is evolving towards a more audiovisual kind of content, versus text, i think we are beginning to witness our biases online too and the consequent popularity of topics. some examples:

• Jason Clay: How big brands can help save biodiversity, one of the most interesting talks i’ve heard in a long time. unfortunately, he is old, fat, and not particularly attractive or good looking. he is, however, demonstrating a viable, practical way of fixing today’s society and demonstrating results. no ovations, he is a boring speaker;

• Sebastian Seung: I am my connectome despite the interesting ideas, i couldn’t get past the fact that he is incredibly annoying. this made me reject his ideas, even though a wonderful metaphor is presented: the brain as the mountain, the mind as the river, and how they shape each other in different conditions;

• Slavoj Zizek – Examined Life, slavoj zizek has quickly become an internet favorite, mostly because of the showmanship involved. his ideas, many times, are empty and vague, but most of the time, very strong critiques of modern society. i think that part of the enchantment is his lisp, his exotic waving of hands and accent, which adds authenticity to it.

books have a very strong advantage versus multimedia. they don’t test any of our biological biases towards attractiveness, intonation, etc. they don’t allow us to be affected by physical biases. in fact, you can write signing as your opposite gender. it won’t make much of a difference if you’re talented enough.

the internet used to be a bit like this. it was text, a lot of it, that you had to parse through to get meaning. nowadays, we’re slowly moving towards a more audiovisual knowledge transmission system. but this is the very system being exploited by marketing, and a very sub-optimal way of transmitting information.

but this means that in order to be properly listened to online, we are increasingly demanded of more showmanship. it is no longer sufficient to have a good idea. you also have to be a good seller of that idea. many other industries went through this process, it was just a matter of time until it caught up with the internet. now, why should we be listened to anyway? that’s for a later post. i don’t really think we should at all, we should be heard.

gaita, drums and modernized traditional portuguese music

previously i discussed how things can work becoming agents, and how change is always for the worse on a global perspective. but since this is a very sad and gloomy perspective, let’s be a bit naive now and forget about all the stuff that is going to waste and draw a boundary around a system that allows ectropy, like the earth. from now on, all ectropy must be seen as part of a greater, globally increasing, entropy.

a key aspect of information theory is how we can quantify the structure of things when they are arranged in a specific way. but how do we know they are arranged in a specific way, rather than just a random one? this is a common dilemma, and can be exemplified by asking how many grains of sand are there in a pile of sand. this faces us with a simple question. when do things become parts of a bigger thing? when does an alphabet become a novel, and not just any random assortment of letters?

there are many ways of answering this, but since this is my page, i can answer with my own perspective on it. in my view, when agents increase information in a system so that the total information is more than the sum of all the information required for each thing, we’ve just created a new thing, be it an agent or not.

so let’s go back to the sand issue. each grain of sand can be our unit of information, so let’s say we need g bits to adequately describe (save and restore) each grain of sand. a pile of sand would take, therefore, n x g bits. let’s now take an empty hourglass, and say it’s simple enough that its information fits in h bits. our system now has n x g + h bits in total. now let’s say an agent decides to do work on these things. depending on the agent we choose, and depending on whether we want a fully restorable system, we might want to add the agent information too, let’s say a bits. this is optional, since we might just want to use a hourglass, and not make a hourglass. so now our system has n x g + h + a = s bits.

after our agent works on the system (creating entropy outside the system), let’s say he inserted the sand into the hourglass, giving the system extra structure (more information). what is this information? it is the one provided by the new vocabulary available to both the grains of sand and the hourglass, meaning, each one is now an agent acting on the other one, and therefore, since they have a relationship between them, this must be accounted for (remember Kolmogorov complexity?). so the equations that govern sand in an hourglass might be simple, but they no longer fit in our s bits. the whole is more than the sum of the parts, because the relationship between our two things created new agents, let’s call them l, for laws. our total system is s + l.

note that the previous agents already at work are still there. things like gravity and the glass blower that made the hourglass, have been accounted for as requirements for the initial things. but the fact that each grain of sand in the hourglass cannot go anywhere, instead being confined to the hourglass shape, is entirely new to the system. we just witnessed an agent being born! it came from the simple fact that we connected two things and they became agents limiting each other. in this case, it’s more the hourglass than the sand doing the limitation, but nevertheless, the hourglass is not an agent until it has sand inside it.

thanks to structuring, or work, we have increased the information in our system not only by its parts, but also, inadvertently, created more things. this is ectropy, or like we humans like to call it, making things pretty.

now, a hypothetical natural hourglass, fruit only of secret agents (laws of nature without a mind), could be filled with sand just by random chance, thanks to the free energy on earth, coupled with gravity. if this were the case, we would have more information in our system, even though there were no minds to use the hourglass. so information synthesizes itself simply because things jiggle a lot and bump into each other.

this is one of the things we see a lot of on earth. spontaneous order from nothing, and we’ve also seen how easy it is to synthesize information, by merely making things constrain other things. so information increases when there are more restrictions (agents or laws) to a system when it is assembled than the ones that exist for its parts.

this is why you can’t reduce everything to simple equations, because some restrictions emerge from their interaction, that require more information than the original formulation to be saved and retrieved. it should be obvious that these new laws are consequence of earlier “deeper” laws, and anyone with sufficient patience could prove that there is no need for a new thing since the previous thing is enough.

this is superficially true, but it avoids the problem of generating the work between the two states. it’s like saying that between an empty plate and a full plate all we have are molecular interactions. yes, it’s true (more like a deepity). but these molecules are restricted to such an extent (the molecules in the ingredients, the molecules in the maker’s brain, the molecules in the tools, etc), that stating all the restrictions created by the thermodynamic trajectories of all those molecules in time (ectropy, information) would require the same time as the process itself. meaning, you’d have to ask the annoying physicist (they’re always physicists, aren’t they?) to demonstrate with his equations the motion of every single particle of every single molecule he claims is the central thing. or, you can just accept a new thing (or law), that like the hourglass example, governs the macroscopic properties of your system. in this case, which ingredients, which recipe and how good is the cook. much easier than typing the trajectories.

this means that things are created only due to these restrictions that are simpler using a thing than using deeper things. or, the simplest way to explain is usually better. some people call this the Occam’s razor principle. this is only relevant, again, if you want to retrieve and save things (observe, discuss, are subsets of these two simple forms of work). nature itself goes on not giving a shit about how much information it has or doesn’t. it’s its own thing.

it is also where all arguments for and against reductionism amount to. one is saying “all you need is the single one secret agent“, and the other one is saying “you can’t describe everything just using secret agent“. they are both right, and here’s how.

while it is true that there can be a single core secret agent, people call it theory of everything these days (back in the day it was god), it is not possible to save or retrieve the entire work done by this agent in enough time to have a conversation (not even a lifetime). so to allow our tiny monkey brains to discuss big things we either create new laws, or are left with a never ending complexity that is impossible to deal with in useful time (not computable in useful time). just ask any physicist to solve the three body problem. though gravitational equations are beautiful, they grow ugly really quickly. and even though that secret agent might be beautiful, it grows ugly by simple entropy. new laws make it pretty again, not ugly, since laws can emerge from any kind of system, made of things or things of things and so on.

if you can say “sand can’t go through glass”, you save 3 hours of painful molecular dynamics, that would waste your time in discussing whatever idea you were trying to discuss to make yourself feel superior to others. note that you can still dissect that phrase onto simpler things, but doing so would waste time, because the information dealt with is the same! we would be manipulating the same information using first an alphabet of just one thing, and then one with many things on top of things, abstraction. abstraction makes it easy to manipulate massive amounts of information. more things mean a richer vocabulary to store and retrieve information! abstraction is just laziness at work! how terrible!

in a way, it is very similar to the way our cortex works. but that’s for later. for now, more music, we already had enough fun. there are many hard concepts in these paragraphs.

this last one is from a (sadly) deceased portuguese artist that took a lot of old traditional material and mixed it. his entire repertoire is freely available at the tribute website