some ney anban from iran. note the double chanter.
i have a portable chromatic tuner i carry with me all the time. it came with some basic lead batteries, and they eventually ran out. so i modded it for solar energy. an interesting side effect is that since the current is very low, we can actually charge the lead batteries. while some batteries do explode or leak, most can be recharged slowly. this is a risk we can take, since my cells are 75mA only.
materials
any chromatic tuner that runs on batteries
solar cells (i have a bunch of 1V/75mA cells that came from recycled toys, check ebay for that kind of stuff)
a diode that can handle 75mA (almost any diode you find on any scrap electronics from the trash can)
wire, soldering iron, voltmeter and diode meter (if you don’t know which one is the plus and minus)
schematic and pictures
results
it works. you can see the led lighting up in the last pic. looks a bit ugly because the cells aren’t a perfect fit to the tuner’s size, plus i made a board out of newspaper and glue layers to hold it, making it look even worse. that’s about it. considering lead batteries do not have memory, they can be recharged indefinitely. this might sound a bit strange, but as long as they aren’t completely drained, they can be reused over and over again. the main advantage of lithium is that it can be completely drained and still work, plus energy density and weight, but there’s no particular reason not to use these as the chargeable cells in this case. also, the charge voltage is very near the operating voltage (3.3V versus 3V) which means it could work only with sunlight. however, it’s always better to have batteries between the two as they stabilize the voltage.
some dazkarieh this time. i love this song, they have a new album out.
in the previous series we saw how information can be stored and processed by physical entities without any elaborate structure. we also saw how the speed at which it is both processed and stored has increased as life became more complex. today, i will summarize again the whole series from the beginning, including our formulation for minds.
reality R is a set of N discrete things repeated M times
the distribution of these things is non-random (i.e., it has patterns)
this distribution is not static, it can change thanks to free energy, leading to specific distributions that can be exceptionally complex (e.g., minds)
the development of complexity comes simply from random fluctuations and non random possibilities for existence (e.g., anti matter particles still form, but they are annihilated by reality (see 2.), hence the evolutionary trait present in all layers of reality)
these specific arrangements versus the general ones can be quantified using arbitrary boundaries, they are arbitrary because there are no boundaries in reality
minds represent a specific subset of R, whose distribution is non random, which is a subjective representation of the whole reality. this means minds can be conceived depending on the boundary: draw one around a planet, an animal, a cell, an atom, and you can define its observations and its thoughts
complex minds made of neurons represent not only observations themselves, but their structure allows for thoughts to generate hypothetical observations, i.e., by virtue of their hierarchical structure, they can generalize patterns so much that impossible distributions can exist as thoughts. this means that even fantasy is real, because it exists physically as a distribution of things (in this case for example, the electrical signals and wiring of the person with the idea).
note that it is not said whether all of the constituents of reality are observable by us, human beings, i.e., of the Nthings reality is made of, i make no claim on how many, and what type, are they made of. this means that if someone wants to use this model to justify spirituality, they can do so, all they have to do is say there is a “spirit particle”. and though i accept it, as i said before, the quest is for the simplest generative space with minimum distortion versus reality, and the more symbols are added that do not improve fitting, but instead distort it, the worse. but in essence, this is an objective formulation of a subjectivist model. it sounds completely ridiculous, but that’s exactly what it is. it harmonizes both subjectivity and objectivity in a way in which both are complementary, not opposing views.
regarding the minds as generative spaces, i will provide a simplistic formulation on how to look at complex brains and see them at work, and will provide predictions and tests that can be done to verify this model.
as we saw first, we are dealing with a limited and discrete set of symbols to work with. i postulate that brains create multi-dimensional spaces on which they project their sensory signals (sensory signals are not observations, they are interpretations, thoughts done over classification of input). the best way to imagine this is to imagine a two dimensional brain (two neuron brain). one of its neurons can classify red, and another can classify green. if we shine red, the red neuron lights up. if we shine green, the green neuron lights up. but if we shine yellow, both light up. now, each neuron “knows” only one color, but combined they can represent colors that alone they couldn’t represent. this is typical of many dimensional systems. two lines, one dimensional, once put perpendicular to each other can now represent all points in a plane, even though each one of them has only one dimension, together they’ve expanded “each other’s” representation of reality. red could not see yellow without green, and vice versa.
this means, and this is mostly, as i said, a very broad generalization i am making, that a single neuron is a base vector of that multi-dimensional space. so these two color neurons can represent yellow because when combined together they create a base vector for a 2 dimensional space versus two 1 dimensional spaces. obviously this view is a bit influenced by neural networks (which actually estimate an n-dimensional polygon and hyperplanes). but i take a simpler approach: i’m thinking just in terms of axis, projections and expansions.
each neuron is a base vector, and together they create the “brain” (a base matrix). the imaginable space of a “brain” is, therefore, the space expanded from these base vectors. note that it might be that not all neurons are independent of each other, that would be expectable, so this actually overestimates the capacity of this system. the main point is that i’m saying we are dealing with generative complexity, not actual complexity: the ideas that flow are consequence of the possibilities of representation, and this representation is a consequence of a reducible set of base “structures”.
i will develop on this soon, with more concise definitions, and discuss this whole block of ideas in terms of their predictable consequences in understanding the interaction of living things.
a “repasseado”, one of the many portuguese folk dances. this one is particularly fun to dance to. playing is one of the last makers of mirandese pipes, master célio pires.
after the previous definitions, we can move on to bigger structures easily, since the principle is the same. a brain is included in the previous definitions, but it is different in a very important aspect: its feedback loop for thought is done by both reality and its own internal processes: brains (and complex living things) can do work on themselves effectively changing their own information. this has the advantage of not taking hundreds of trials and errors through evolution to produce meaningful responses to environmental pressures. note that DNA can actually act on itself by making proteins that inhibit or change its own structure. overall, this process is more like a spaghetti of loops mixing internal structure and work and external structure and work. the difference i’m focusing on is both how long it takes to cause information to change (the thought ‘clock’), and how much information can be stored (the ‘capacity’). the boundary of “internal” and “external” is artificial, and can be drawn arbitrarily. i tend to prefer a boundary between the part that does work and the part that just provides energy for the work. the first is all things in “motion”, the latter is just the fabric in which they stand.
now, we saw previously that it is not sufficient to account for the information represented, it is also necessary to account for the information reading machine, e.g., knowing a DNA sequence is meaningless without the appropriate cell machinery to process it and an environment that triggers its responses. this additional information should be assumed included in everything i describe as “information”. as i said, it is not sufficient to know the words, it is necessary to know what the words mean. since we generalized the definition of mind and thought, a cell + DNA is actually a small mind with very simple thoughts. its non-random response to the environment signals some sense of understanding of the real world: too hot, too cold, too moist, too dry, stressed, peaceful, all these chemical signals that bathe it are interpreted and processed according to some internal representation of reality. this representation is the chemical structure of the cell itself in its whole: it must include all the parts necessary for the accurate observation and change of reality’s constituents.
the same holds true for a brain of any kind with a big difference. a cell might go through reproduction to change its DNA, but a brain (a brain is a mind made of connected cells we call neurons) can do so on the fly by rewiring itself. a brain, thanks to the capacities of its neurons, can change its observations and its actions easily without necessarily going through the reproductive process (the actions are changed thanks to its relative power over its supporting systems). this allows for it to be a contained self-reorganizing system. but what defines the fitness of this organization is, obviously, reality. the brains that are better at reorganizing themselves to fit an optimal response to reality are favored, and the ones that don’t, aren’t. so it’s plausible to expect that brains are good at observing reality efficiently, otherwise they wouldn’t survive. more on this later.
this representation of information (which we analyzed previously) is quantifiable, like every other before it. but now we have to deal with a self-modifying observation, instead of a passive one (like our planet analogy). a planet has no choice on what information it will store: it is just a consequence of its surroundings. but a mind, through self-modification, can create a subset of information from all the information it gets. both are subjective observers (their observation is a subset of reality), yet the brain can take that subset and change it. this change, sometimes called reflection or learning, can exist in simple and elaborate brains. we like to think that only humans reflect on what they are doing, but i argue that reflection and learning are a consequence of active self-modification.
the reality observed by brains is distorted not only by its location (like the location of a planet), but also by its structural responses and modifications. this means that brains are worse than passive observers at representing reality. this might seem counter intuitive, but as we said before, we may be able to imagine other planets, but this is a long way away from what a dweller of that planet would observe. even a pebble on its surface will know more about the history of the planet than all our telescopes combined. this should be a humbling perspective above all. we must understand that our observations as consequence of our brains are distorted both by being exposed to a limited subset of data and a filtered, biased alteration of the said data. the wrinkles of our skin, the spots and wounds are a much better catalog of what happens to us than our interpretations of what happened to us.
now, there is something very interesting about brains made of neurons, which in my opinion is quite surprising but makes perfect sense. it is impossible to represent reality in a few billion neurons accurately. but, as we saw in our information analysis of abstraction, it is possible to represent increasing quantities of information by creating abstractions. an abstraction is no more than a subset of common information extracted from a large data set. like in my earliest examples, consider that we have the data set {a,b,a,b,a,b}. we can abstract it by saying it is 3 times {a,b}. note that this is not the same as knowing the entire set: it is merely creating an internal observation that has minimal distortion versus the real thing, saving resources, assuming its elements are interchangeable. it is possible, therefore, to represent billions of tons of steel as a mix of a concept of quantity and a concept of steel. note that you won’t be able to point out atom number x in the bulk of steel (therefore your data will not be fully accurate), yet you have a general sense of what it’s made of. this process of pattern recognition and abstraction makes brains naturally scientific. i’m going to get in trouble because of this one i know it. but in this lies a physical flaw in what is abstraction: all the elements of the set abstracted as the same category are considered to be interchangeable and the same, i.e., and iron atom can be swapped by another one with no trouble. while this might turn out to be true for some cases (like chemical elements), it is completely false for other cases (like people). yet our brains are capable of extrapolating data with equal power regardless of the category used. this is not because our brains have elaborate internal representations of both iron atoms and people. it is because the neurons themselves make no distinction between a pattern that is a person and a pattern that is iron. the fact that we can generalize does not come from richness of observation and understanding, instead, comes from a deep lack of understanding of things. this understanding of the differences between things in reality is something that is fine tuned by our interaction with it. to a young child, a puppet is just as alive as a human being: its brain generalized both as living. it is only after much training and learning that these two can be split between a dead puppet and a living being. and what makes the split is the child’s understanding that a puppet can’t move on its own. but the reason why the puppet can’t move on its own comes from the outside world: reality, through the laws of physics, provides that information permanently in every observation. its brain will then, if it’s not deceiving itself, correct its observations to generate a better observation of reality. i hope this is clear because this is essential to understand what will come next.
why is it naturally scientific (actually, i should say naturally philosophic)? let’s go through the process of a mind growing. the mind starts with mind M as it inherited from its ancestors (DNA, cells, etc). this mind M has both observations and actions (as i said before, they are both physical things). if M is faced with a new pattern p and observes and acts on it adequately, then M will prevail. if not, either M corrects itself into a fitter M’ or it will perish. this requires that M is permanently observing reality and testing its internal representation against reality. this is similar to the scientific method: bad ideas are discarded because they don’t work, but whether they work or not is verified by reality. i already provided a definition of the scientific method previously, but now it should become clearer. brains calculate a generative space through abstraction: 3 times {a,b} is a generative space because it can “create” via work a set bigger than itself. abstraction allows us to change the 3 into 4 or 5 and imagine new {a,b} worlds that don’t exist. this is a natural consequence of abstraction. the algorithm used to create the perfect fitting between the abstractions of reality and the observations of reality is the scientific method. its error is the measurement of scientific progress. let’s do a simple example to clarify.
consider miniverse R = {a,b,a,b,a,c}. two competing abstractions exist in minds M1 and M2. M1 = 3 times {a,b} and M2 = 3 times {a}, 2 times {b}, 1 time {c}. which is the one that fits the data best? we generate the extrapolated reality of each, RM1 = {a,b,a,b,a,b} and RM2 = {a,a,a,b,b,c}, and compare them with reality R. we can see that the first mind miscalculated one observation (b instead of c), so it has an error of ~17%. the second calculated all observations correctly. the best choice is, therefore, M2. my formulation of the scientific method is the search for the optimal abstraction whose generated observation space fits reality with 0 error. it is impossible, but an interesting goal. any animal whose understanding of reality is fitter than another will have an advantage over others, and therefore, will thrive. being scientific is no more than being very good at following our survival instinct.
the main point i’m getting at is, however, not how reality as a feedback is key to understanding minds, but actually, how minds themselves represent reality as generative spaces instead of static observations. as we saw above this makes sense if our observation resources are limited to a small quantity of data. it is easier to store the generative space of reality and generate it than to store reality itself. much like my example. but there is a problem. if you accepted my premise that brains store mostly generative spaces instead of real observations, then this means brains can generate realities that do not exist. the existence of brains creates extrapolated realities that do not fit reality. for example, i can imagine an extrapolated crocoduck from my abstractions of living beings, ducks and crocodiles. but this generated observation, when tested, is proven to be false as an existing animal. my abstractions of living beings, ducks and crocodiles are good scientific evidence tested by reality, yet i can generate equally easily completely invalid ones. how can we tell them apart? as i said, for survival reasons, animals with a good sense of “real” should survive better than animals without it. but if survival is out of the picture, then there is no way of stopping fantasy (non-scientific knowledge) from becoming the center actor of a brain’s activity. this hints at the subject of the coming posts. note that even a false idea resides in reality, so it is real only as far as the molecules sustaining its concepts are real. once extrapolated and tested and proved false, it should remain as a hypothetical tale and a proof of the insufficiency of generative spaces.
so far, these seemingly elaborate concepts have been given without any apparent goal. from here on i will begin to build on these concepts and analyze bigger and bigger systems. it’s been very boring, but this sets up the foundation on which i will work, so that ambiguity is removed.
some gaita sanabresa. today this post is shaped like a scientific article well, because it is. if you want to see the results first check the link in the abstract. apparently, CS is not that misogynous after all.
abstract
much has been said about gender bias in hospitality networks, and much of it has been mostly speculation. it has been frequently said that males on these networks exploit them as a way to meet females. therefore, i hereby present a quantitative analysis of a specific case study: the rate of replies to males and females on last minute groups on the website CouchSurfing. the final charts can be seen on this page
methodology
all data collected was done so via publicly available data, much like a search engine would do it. in the cases where users choose to make their data private, or groups choose to require a login, no data is collected.
the data was collected using a data mining technique, using PHP as the main language. this data was then dumped onto a TSV file that contained the timestamp of collection, number of replies, the gender of the person sending the message, and the country and city of the group. group id, message id and user id were also collected for consistency.
data set choice
last minute groups are characterized by being focused on a city or country, so local groups were chosen from various global locations. the software was developed so that more groups can be added easily. in this case, 36 groups were chosen. after mining, 22 groups were left from 18 different countries. data from blocked groups was not gathered since it is not publicly available. 5 pages from each group were collected, totaling 1760 messages.
data analysis
once the TSV file was populated, it was then fed to a SQLite relational database, with a primary key on the postid, so that duplicate posts were not taken into account.
the data was then grouped by gender, gender-country, and gender-country-city tuples, calculating the rates of reply, maximums and averages. the data was then processed into an HTML page that charts all of this information. note that when gender is unknown that can both mean the user did not fill it out or has chosen not to have it publicly available.
for visualization, a bar chart was drawn with the maximum length as the maximum average response rate of all groups by category, allowing for visual comparison of different groups or countries directly by visual inspection.
results
the results can be seen in this html page or embedded below if your browser allows it
discussion
world data
the world data was very leveled, indicating that there is little gender bias in general, though women do tend to have a slightly higher response rate than a general user. this indicates, though not strongly, that being of the female gender favors the response rates expected by a user on this network.
country and city data
the country and city data is perhaps the most interesting part of this study, as very strong contrasts are seen between different groups. for example, India showed 3 times the average response rate to females, versus 1.4 times the average response rate for males, indicating twice the response rate to females versus males. on the other hand, Luxembourg had 3.33 times more replies to males than females (which were 1.81 times the world average), signaling that in Luxembourg it was twice more likely for a male to get a reply than a female. in general, cultural differences are very strong between countries in regards to gender bias: some countries are very gender-sensitive, while others are not significantly sensitive.
conclusions
this demonstrates what is frequently said about gender bias in a global perspective, that females are favored by their gender in finding a place to stay. however, if the data is split regionally, this no longer adds up. the culturally different attitudes towards gender seem to be stronger than what common sense would claim, disproving the hypothesis that females are always favored by their gender. females are favored by their gender in some cases, in others not, so this is sufficient to disprove that there is widespread gender bias on CS. it exists, but it is confined to specific geographical locations, with wildly varying amounts, and not in a very significant way at all.
note that this data is biased by the fact that we only analyze people that couldn’t find a couch and are using last minute groups, which in itself cannot be used to generalize further, though it is already indicative of no significant global gender bias.
sources for replication
i provide all sources and the database collected for download freely, as long as the license of this website is respected. all scripts are prefixed with a shebang line so that they can be used in a shell environment. if your php-cli is located elsewhere, you should change that line. code is not commented, it is too small to be complicated, but if anyone needs assistance just comment below.
[TSV file of the data mined](http://ubuntuone.com/p/raS/) (works with excel/open office)
[data mining PHP script](http://ubuntuone.com/p/raU/) (includes a simple way to add more groups. it randomizes the request interval to avoid firewalls and cause server problems)
[TSV to database script](http://ubuntuone.com/p/raa/) (converts data from one to the other, effectively filling up the database)
[HTML code generator](http://ubuntuone.com/p/rab/) (generates the html graphs from the data present in the database)
[HTML output example](http://ubuntuone.com/p/rac/) (the output of the software for the data in this study)
comment
it is a bit unsettling to me that all this data is made publicly available by CS without any control. this has to do with the default privacy setting: people share everything with the world unless they choose not to. this means that most people that aren’t particularly tech savvy will end up sharing more than they would expect. this is becoming a trend in online websites, share everything by default, which, in my opinion, sets a dangerous precedent. not everyone is like me and is doing this to test out scientific hypotheses. this information can be easily exploited commercially, with some 30 mins of coding like i did.
this study also demonstrates how informatics can be helpful in social sciences, and that with a little bit of coding one can get huge datasets automatically, ready for processing. it took me 30 minutes of coding to set up the mining, left it overnight to crawl the website, and then about 1h to setup the visualization. it should be interesting to see this data on a map, but for the sake of my free time, i’m not going to do it.
just a short notice. this saturday is the FEST-i-GAITA 2011 bagpipe festival in lisbon. there will be workshops and documentaries during the day and concerts at night. me and several other students of the bagpipe school will perform to fill in between sets. the video is of the band of the association i’m part of. the line up includes Remi Decker, Volta e Meia and Roncos do Diabo