Lo(B)! The Universal Library is complete!
This topic contains 148 replies, has 2 voices, and was last updated by jon 1 month, 1 week ago.
December 14, 2015 at 8:54 am #4758
I have been reading all about this for the past few hours and I am flabbergasted. This is really really really amazing. I was thinking that the image version contains (obviously not in order, and as individual frames) a video of every person’s life from every possible angle with every possible choice and interactions with every possible other person at any state of their choices. Mindblowing.December 15, 2015 at 11:09 pm #4910
It’s probably what’s between two mirrors reflecting only each other. Eerie.
Amazing work, and a grip on whatever philosophy is wearing today.
I was wondering whether the same would be possible for music, or sounds in general –
or if I was insensitive and somebody already worked on it.
Indefinite thanks for the site.December 16, 2015 at 11:33 pm #4922
I find the possibility of doing something similar with music to be fascinating, so I tried to run some numbers just to figure out if it’s possible.
Strictly speaking, yes, but it’s not as obvious how to approach the problem (at least to me) as text or images. First off, people can apparently distinguish between about 330,000 different frequencies and intensities of sound. That’s kind of like having to deal with 330,000 different possible letters. But those kind of numbers are apparently manageable, given what Jonathan Basile has accomplished with images. The main issue that makes music different is handling the passage of time. My first thought was just to generate a random frequency every tenth of a second. (That’s a common max rate of a lot of computer-generated music.) It seems entirely plausible that you could generate ‘songs’ like this that last over 3 minutes. The problem is that most music isn’t just one frequency that changes every tenth of a second. Music (and most sound in general) is layered. That’s the reason why you need a band or an orchestra to play a lot of music; one instrument just isn’t enough, no matter how wide of a range of sounds you can make with it.
Another thought I had was that you could maybe try something similar to the recording approach (like how people record sounds on CDs), but then you run into the problem of needing a really high sample rate to be able to generate all the possible sound combinations. (like 44,000 a second instead of just 10)
It solves the “multiple layers” problem, thanks to some sampling theorem that I’m completely unfamiliar with. Unfortunately, picking one of 330,000 possible frequencies/intensities 44,000 times a second gives you less than 4 seconds of sound before you’ve hit the same number of combinations as the image archive here.
But just maybe if you drop the sampling rate to 8,000 (i.e. that of telephones rather than that of CDs), you could generate every possible 20 seconds of sound. Every 20-second telephone call ever sounds worth the effort to me. There’s a decent chance that you could trim out a lot of the different frequencies too with enough experimentation, giving you like 10 more seconds or so. Great, now I’m tempted to waste time experimenting with this.
Thanks for bringing up the idea, it triggered a spree of research on everything from CDs to information theory. I hope someone who knows more about music and sound in general will give their feedback on this concept.December 31, 2015 at 11:33 pm #5906
Amazing work. Could you consider updating the about page now that it shows outdated information on the size of the library? Also do you feel that it in any way goes against the spirit of the project that pages are not stored anywhere and are only generated algorithmically through user request? As you stated on the grains of sand page, it would be impossible to store all of the pages, but it sort of goes against the idea that all possible information is already out there if it doesn’t generate until queried.January 2, 2016 at 12:32 am #5918
Hey explorer – the information on the about page at present is accurate – the library contains every possible page of text – 29^3200 pages. I still hope to expand it to 29^1312000 and to branch out into other character sets.
The question of pre-generation is interesting to me. Even if a full library stored on disk (which would still be compressed) were possible, the user experience would be indistinguishable. I hope the library complicates our notions of presence and absence, existence and nonexistence.January 2, 2016 at 4:05 pm #5922
This is a so cool.
I would love to know which pseudo random generator you use and how do you reverse it to make the search working? you speak about “a successful formula combining modular arithmetic and bit-shifting operations, and the result is the library you see today.” in this page but what is the formula ?January 8, 2016 at 1:44 pm #5962
this site is surely one of the greatest things on the web as on the world. I came here just yesterday via Vsauce and spent the whole evening exploring and learning about this thing, reading the theory and Borges’ story and so on. I think I’ve even dreamed of this last night, something, I’m experiencing not quite often. This site makes a difference. It’s a breakthrough on several grounds and so hard to grasp.
I can clearly remember myself as a kid asking my mother if there would ever come an “end” to storytelling, if one day in the future every possible story with every possible setting and outcome would have been told and there would be nothing left to tell. She answered no. She claimed as long as there were humans they’d always come up with something new, something never thought or said or written before. There would be no end.
Regarding the Library of Babel, one could say she wasn’t right. Every story one could possibly tell can be found in the library. So there is nothing to say, that isn’t already contained in some of its volumes.
But on the other hand at the moment the algorithm is limited to 3200 characters per page. So a story longer than that can’t be found on one page, but probably on several pages. Also the algorithm could be used to run with a higher number of characters on each page, growing the library potentially bigger.
But in any case the total sum of characters written in the library is finite. So eventually an author putting veeeeery much effort and time in it could write a story with more characters, one more would be sufficient, and by that would create something, that cant be found in the library. Even if computers were technically able to run the algorithm with a trillion characters on each page, theoretically somebody still could come up with a story containing more of them. (One would have to grow quite old to achive it, but who knows how future technology might prolong mans lifetime.)
And even if the algorithm is set to run with an infinite amount of characters, a dedicated author could also start to write a neverending story, a story that is passed over from writer to writer over generations. If they keep up good working, the library could never ever contain all of it. If the writers can possibly maintain to write at the same speed the library generates, there would be a neverending status quo.
So this would prove my mother was right!
Is it correct thinking or am I going nuts?January 10, 2016 at 10:20 am #5974
I have information regarding how to link the library to the present and thus the past and future using another overlaying 4 dimensional model which consists of the Flower of Life and megladons cube. Contact me if you are interested in taking this into a whole new level.January 10, 2016 at 11:53 am #5975
What an excellent site. Where can I find the page that explains the mathematics and stuff behind this algorithm. Until you post it online eventually. 😉January 19, 2016 at 12:40 pm #6040
I found the page where Pi starts! Now I only have to read the following pages to beat the number of digits record!
https://libraryofbabel.info/bookmark.cgi?where_pi_startsMarch 15, 2016 at 12:36 am #6391
I must not have the iq to grasp what on earth im supposed to do with this site or even the image search. I have not deviled deep yet but on the surface its confusing.April 25, 2016 at 2:19 am #6788
I have been looking for this site for years. I was so desperate that I even wondered if I could be able, with my abysmally inadequate mathematical and programming tools, to write up something even remotely similar. Your site, with the option to randomly find any existing text within the Library, is even better than anything I could even dream of, let alone realize.
I will make a tiny material contribution to your spectacular endeavour by donating a very small amount of money – more is sadly beyond my current possibilities.
I would also like to share my (hopeless) ideas and my approach to the building of a Universal Library. My approach was very similiar to what has been proposed by Eugene Koontz in this thread (post #828) and to the suggestions of others.
Just last night, before finding your jewel, I wasted some precious sleep hours trying to calculate if I could have a three dimensional system of 64 bits coordinates large enough to contain my own version of the Library as I thought it: The Abridged Library of Babel.
What I thought of was to generate combinations of words rather than characters. Language would have to be English, because of its ubiquity and of its simpleness (you ad an -s, an -ed, and you got all the inflections for most words; doing the same with, say, Italian, would be much worse). Using a dictionary of 10<sup>5</sup> words and names, I wanted to create all the possible thousand-words combinations, and that should sum up to 10<sup>5000</sup> articles (hope I’m not horribly wrong). Of course, just sequences of dictionary indexes would be created; that would mean that by just switching the dictionary you would get an altogether different Abridged Library!
This virtual Universe would be far more modest and dim than those envisioned by Borges and brilliantly realized by you, but you could have more chances of coming by something meaningful-like. And you could unleash programmable bot-librarians searching for apparently sound language constructions (that would be far, far more likely to occur than in a full, character based Library) and reporting the articles’ coordinates for flesh librarians to examine.
What do you think?
Anyway, congratulations and thanks for The Library!May 2, 2016 at 1:29 am #6837
I believe someone made a site like the one you’re describing. Take a look here and see what you think: http://libraryofwords.info/May 3, 2016 at 11:03 pm #6852
Again, thank you. Pepe’s Library of Words realizes exactly what I had in mind. I’ll have a look at his code to see if I can figure out how it works.May 24, 2016 at 12:05 am #7023
You are a genius Jonathan Basile, this is a magnific tribute to Borges. Congratulations!