Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Library of Words (libraryofwords.info)
68 points by chishaku on Oct 10, 2015 | hide | past | favorite | 25 comments



Projects like this are so fascinating to me. I saw an art piece once that consisted of a TV that rotated through all possible pixel combinations. Think about that for a second: at some point in time, it will display any possible image anyone could ever make or dream of (in that given format). Any person's face, any movie-frame, any written sentence, etc. So fascinating, the vastness of it all.


Most likely most of those images are just random noise, and nothing intelligible. It would take longer than the heat death of the universe for that TV to display all possible images.

Let's examine an extreme underestimate to demonstrate this.

Say the TV is monochrome with an 8x8 resolution, and it can change it's image every millisecond (1000 fps). It would then take 2^(8*8)/1000/3600/24/365 = 584,942,417 years for that TV to generate all possible 8x8 monochromatic images.

Then when you consider that the TV is likely running at something more like 640x480 resolution with 24-bit color at 60 fps, you can see that it will take an astronomically larger amount of time for the actual TV to display all possible images. The time would be larger still if the TV were running at 1920x1080 resolution (or 4k!).

Further more, a large majority of the sample space is just unintelligible noise, and the vast minority of the space is actual recognizable images. So likely, the art project will just be displaying noise until the end of the human race.

I don't think this takes anything away from the artistic merit of the work, it just changes the message a little. We humans are capable of creating a machine that can display every image, but we will never be able to see them all. It is an interesting contrast between the creation of a generator and the execution of it. I don't think this is what the artist initially intended with their work, but it is interesting to ponder nonetheless.


Possibly "Every Icon" by John Simon Jr., 1997

http://www.radicalart.info/AlgorithmicArt/grid/every/EveryIc...


I am glad you liked it! There's already some image search engines, as posted above, including this one: https://babelia.libraryofbabel.info/


Is it possible to know when in time its pixel values would form certain images?



These infinity monkeys wrote shakespeare:

preconizer arrythmically to be or not to be that is the question whether it is nobler in the mind to suffer the slings and arrows of outrageous fortune or to take arms against a sea of troubles spearfishes stifles nonswearer belabouring apperceptionism

http://goo.gl/j6XHOR


Interesting continuation to one of the most famous pieces by shakespeare.


Yeah! right? But if I understand your project correctly, it also has every other possible continuation to that stanza in it somewhere?


Correct! A (quite long) game would be to find one more interesting than his version


Shouldn't there be an "s" on the end of "Jorge Luis Borge[s]"?


Actually, "Joseluí Borge" is a very common pronunciation in some suburbs of Buenos Aires, for example the city council of Quilmes (http://www.lanacion.com.ar/422798-impiden-que-una-calle-llev...).


So not true. The Peronista councilpeople were saying "José Luis Borges" (as opposed to "Jorge Luis Borges") to mockingly express their opposition to the measure of honoring him. From the article you cite:

"[...] agregó Morelli, molesto porque varios de los ediles que expusieron en la sesión para rechazar el proyecto se referían al escritor como 'José Luis Borges'."

Please don't spread misinformation in HN.


Of course, that's pronunciation rather than spelling.


True, will correct soon, thanks!


I've always looked forward to this being applied to music. Imagine searching for all possible generations of notes composed on pages.


But what would you expect that "search" to give you? If you search this library for "to be or not to be", it will give you a page containing the phrase "to be or not to be" preceded and followed by random gibberish.

You could similarly build an analogous music library that you could search for the opening motif of Beethoven's 5th. But the notes surrounding it would also just be random noise; you wouldn't find anything interesting in them.

You don't need a "search" to generate random notes, you can have that more directly by writing a program to do it.


You might enjoy Chris Ford's talk "Kolmogorov Music" from this year's Strange Loop conference: https://www.youtube.com/watch?v=Qg3XOfioapI

I recommend the whole thing, but he starts talking about infinitely generated music at about 26:30 into the talk, and plays a sample at 31:30

And his source for the talk is at https://github.com/ctford/kolmogorov-music


Reminds me of this: http://www.directory.io/


Great find!


Pretty cool!


"Given the size of the vocabulary used in this library," - really? Is 320 words that big? Is a list of these 320 words available?


Found your comment :D

[removed link]

Near the bottom.

EDIT: it seems the direct link does not work… Try entering this in the browse page (http://libraryofwords.info/browse.html) :

    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


Seems to work to me. Check at the end of that page and I see "given the size of the vocabulary used in this library really is three hundred and twenty words that big is a list of these three hundred and twenty words available". If you are referring to the whole page being in bold, that happens because you are looking up for that page exactly. Use the "string only" search if you want a location string for a particular sentence, to preserve the bold around that sentence. The issue with that is that it will give you a different page (although still containing that string). E.g:

nLagn9BPkh5lAc7tCbWQNjkt271OF9JHi0NITtGLUPx9iJ1HxUgUq67kEoANXkJ8oDw3Fn79yNEs1sVqGm6LZGjl3P6YS


It's 320 words per page.

The English dictionary the site uses contains 354939 words.




Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: