/sci/ - Science & Math

So i just read about the infinite monkey theorem.

Hamlet by Shakespeare has around 130.000 letters and ignoring punctuation and upper-case, getting 1 letter right is a 1 out of 26 chance. The probability of getting the entire text right on first try is one in 3.4x10^183.946.

The chance is fuckballs small ofc but i was wondering, would x number of monkeys be enough to type out the text correctly first try if x was Grahams Number?

>>6988723
>>6988748
Streaming games has its limitations in responsiveness. it's fine for civilization or some slowpoke strategy games, but for shooters and other action based games, every delay hurts. Also bandwidth is a serious problem even if it's build out. I'd like to ask who would clutter those connections with streaming.. but then there's netflix - oh well.
And the whole distributed computation stuff is about overcoming limitations.
Cant ramp up the clock rate? Slap in another processor and deal with figuring out how to use both in an efficient manner (this is where doubling what you have rarely halves your computation time)
Single computer can't take the heat? Slap a network and some more computers on it.
Smartphone can't compute compute the weather forecast because it's energy supply would run dry and you'd be faster watching the sky? Install a weather app.

Also, on topic.. i don't think any AI could surpass it's creator unless it has some kind of evolutionary approach. And even then, how do you find the 'successful' evolutions?
My bet would be, given enough computational resources, an AI whose sole superiority would be access to the vast human knowledge and the generalization of knowledge. Since humans have to rely on communication all progress is slowed because you have to 'synchronize' each and every researcher (like reading papers and understand them in the same way the others do).

TL;DR: Infinite monkeys hopping on a typewriter and creating shakespears works.. but who will proofread and tell them when they're done?