@Rem Hydron: See this thread on the points you have raised!
@all the others: It would be an interesting thing to set up an RSS feed which sends out every 10 millionth picture. I would definitely subscribe to that! Would you...?
"We have done the impossible and that makes us mighty." - Malcolm ReynoldsSort of reminds me of mezzacotta's billions of years of comics.
[1] This facsimile operated in part by synAC.No, why would I be interested in an RSS feed of random pixel noise? For every picture of Harry Potter having sex with Paris Hilton there's be a trillion pictures of garbled ugly noise. You'd be more likely to find a winning lottery ticket in the street than to see something interesting.
edited 3rd Aug '11 1:13:06 AM by SlightlyEvilDoctor
Point that somewhere else, or I'll reengage the harmonic tachyon modulator.Hm... 1.7% in 8.5 years? If my math is right, that's about 0.21% a year, so if they keep that pace, they'll be done by around 2503. If they could quadruple the speed they generate the images, we could see it mostly completed in our lifetime.
Still the whole "shut down for child pornography" risk is there, I suppose, which is unfortunate, but depending on how they generate the images, won't be an issue for quite awhile. Especially if they do it by, say, starting with a white image, then make 250,000 variations of it with a single black pixel everywhere it could be, moving through each of the 256 possible colors in that manner, then doing the same with two black pixels and so on and so forth.
They lost me. Forgot me. Made you from parts of me. If you're the One, my father's son, what am I supposed to be?I sent an email to the project's owners with a few questions. As soon as they answer and there is something of interest, I will share it with you, guys.
I don't get something about the child pr0n thing: isn't that only an issue if there were... you know, actual children involved...?
"We have done the impossible and that makes us mighty." - Malcolm ReynoldsWell, opinions differ - drop by the 'paedophilia laws' thread to talk about that.
Be not afraid...@ Slightly Evil Doctor: I bet all of the pictures could be interesting. If nothing else, in a psychedelic fun kind of way...
"We have done the impossible and that makes us mighty." - Malcolm ReynoldsWe'll probably have quantum computers long before 2500; as I understand it, proof of concept already exists. A quantum computer could not only finish the project in a short period of time, but it could also search the database based on any number of chosen algorithms.
One way to do this, perhaps, would be to run different search algorithms which would each apply a tag of sorts to matching images. For example, an algorithm to recognize faces, one for numbers and letters, and perhaps others designed to recognize a given type of thing. At the very least, this would strip away much of the noise. By using some manner of elimination process, one would be able to slowly narrow it down. Once narrowed down enough, the database of remaining images could be skimmed through manually by the curious adventurer. Even better, it could be placed on the internet and users could pinpoint images of interest and add tags.
Where were you when I laid the earth’s foundation? Tell me, if you understand. Who marked off its dimensions? Surely you know! ~ GODYep, that's what I thought too.
Moreover, I thought of a project that could "pretend" that we already have the entire library: let's have a website with a simple textbox where users could put strings of text. Every string specifies an exact image (with the help of some algorhytms if the text is too short, not unlike the "shells" in Minecraft), which would be generated on-the-spot.
Of course it would be downright impossible to find text strings that generate comprehensible images, but hey, I would test my luck by pasting the text of this post and then all my blogposts and then...
I have always wanted to see, how my blogposts look like.
"We have done the impossible and that makes us mighty." - Malcolm ReynoldsEven the fastest quantum search algorithms scale something like O(sqrt(n)). Considering the n is larger than the amount of Planck volumes in the universe, you still won't get anywhere in any reasonable time.
edited 3rd Aug '11 4:42:00 AM by Yej
Da Rules excuse all the inaccuracy in the world. Listen to them, not me.They have funny figures. As far as I can tell, the actual number of possible images, given that the 256 colors are fixed*, is 2 to the power of 2,000,000. (it's equivalent to a (500*500*8)-bit number, because we are talking about every possible permutation.)
- whether this is the case was not clarified in the article.
(yes, I think the possibility of this project completing EVER is quite small. something like 1/(aleph 9 infinity) ;)
'Don't beg for anything, do it yourself, or else you won't get anything.'Is there a chance that my "text to image" solution works (description in my previous post)? I'm no programmer, so I'm not aware of the pitfalls, if there are any...
edited 4th Aug '11 12:12:21 AM by Sati1984
"We have done the impossible and that makes us mighty." - Malcolm ReynoldsYeah, it'd be pretty easy (probably less than an hour) to make a program that generates a random image given a bit of text, such that the same text always returns the same image, and that pretty much any image is possible (ensuring all possible images can be generated is a bit tougher and require ssome good understanding of pseudorandom generators etc.).
99.999% of pictures will be garbled meaningless noise though.
Point that somewhere else, or I'll reengage the harmonic tachyon modulator.I have found a binary-to-text converter. 99,99% I randomly enter the binary field converts to meaningless garbled text. But a meaningless garbled picture would be more interesting...
"We have done the impossible and that makes us mighty." - Malcolm Reynoldsreally◊?
Point that somewhere else, or I'll reengage the harmonic tachyon modulator.This looks like a stereogram
"We have done the impossible and that makes us mighty." - Malcolm ReynoldsAnyone else has an opinion on this maybe? (bump)
"We have done the impossible and that makes us mighty." - Malcolm ReynoldsConsidering:
and this post
Now let's say a computer can generate 10000 per nanosecond (10 trillion per second, well beyond what current technology is capable of, but still.) and there are 10 billion computers in the world (another massive overestimate, ) every single one of which is capable of this absurd power (even netbooks.) Let's say that there's a distributed-computing effort that uses every single one. Now we have the project done in 3.1*10^602036 seconds, or 2.3*10^602019 times the age of the universe.
Yeah, how did they get to 1.7% in such a short time, considering those two posts I quoted?
something
Yeah, it's probably much easier to find anything squicky with Google than it would with the "Picture Library of Babel". Making an algorithm for searching for a specific kind of porn in those pictures is pretty much equivalent to making an algorithm for generating porn pictures.
Point that somewhere else, or I'll reengage the harmonic tachyon modulator.