jafiwam's Journal: Million Monkeys

My boss has a program called "Million Monkeys" that uses the Windows UI to pound through an arbitrarily long brute-force sequence in some program installation routines... basically a generalized brute force key generator that can be pointed at a "key" field and run for several days to get a valid key.

So there's often discussions about the classic Shakespeare question; if a million monkeys sat in front of a million typewriters would they eventually produce the complete works of Shakespeare?

It's a fun question.

Here's some point's of view;

Yes. Well, if you assume the example of the keygen program fits for larger field sizes, then absolutely yes. True random, or sequential tries would eventually produce the desired results, as long as you have an arbitrarily long length of time to get there.

Also, interestingly enough there's a mathematical proof out there that says any specific string of arbitrarily long length exists in the numbers after the decimal points in Pi. These words, they are in there. deCSS code, it's in there. Britney's next song lyrics (before she has written it... if she writes her own songs...) is in there. So is Shakespeare. All of it. [I'll try to find a link.]

No. Shakespeare's words on typewriters are not random, monkeys (in real life) are likely to produce non-random keystrokes as well. These two non-random data sets do not intersect. Monkeys simply won't type "King Lear" (the phrase, not the play) because the word "Lear" requires a shift-L stroke, then the e, a, and r strokes in a certain order. Monkey keystrokes will be grouped in monkey-sized feet or fist groups. (Though the monkey's I've met probably would try to pry the keys up and chew on them.) So even though the randomness can be arbitrarily long, the data-set produced is always not the same data set as english words.

Push the "no" idea a little further, and the Internet can be used as a test bed to prove the point.

To begin, I start with a quote; "We used to think that a million monkeys in front of a million keyboards would eventually produce the complete works of Shakespeare, now, thanks to the Internet, we know this is not true." - P. Lutus.

Internet users are going to produce by far strings of characters that are FAR less random than that of monkeys. Yet, without deliberately attempting to, they haven't produced the complete works of Shakespeare.

Ok. On to the point. This quote from tbannist This quote concerning the question of AOLers producing the works of Shakespeare is incredibly funny (even though it doesn't subscribe to my same point of view).

"No, the key here is monkeys are a random string generator. You can't replace a random string generator with a stupid string generator and expect to get the same result."