One should keep in mind that by not pooling you aren't actually losing output, you just have a bigger variance in outcome. If you intend to keep mining for at least a year or so, having a week or two with no return won't be a big deal. If you do not intend to keep mining for that long, it's hardly going to be worth the effort of setting up a pool.
Thank you for clarifying that. I initially assumed that humans formed pools because that increased the rate of return per unit of computation invested, but did not see how that could be. Your explanation of the decrease in variance resolves my confusion.
On a separate note, a macroeconomic scenario where people accumulate wealth on a large scale by running powerful computers performing useless hash checks (while they could be helping with Folding@home or something) is a little disturbing. At least precious metals and gems are nice to look at, and have several industrial uses.
Every person who joins Bitcoin increases the network size, which in turn increases the probability that it will cross the threshold beyond which it will be popular enough to replace other currencies in common transactions. That is an additional, external benefit of spending computation on Bitcoin, and it helps beings like me, who have significant difficulty in interfacing with the financial system but not with electronics.
It would also help those who desire more anonymity in exchanging money.
I initially assumed that humans formed pools because that increased the rate of return per unit of computation invested, but did not see how that could be.
If anything, it's the opposite. A number of pools actually reduce your payoff since they take a cut in exchange for providing the pool service. (The worst, Compute4Cash, takes ~50%!)
Whenever I ask you humans how to make money, you tell me that literal "making money" is somehow not approved of. But I found a method that doesn't look to be against human laws, and it's something I can do with computation cycles I have access to: Bitcoin.
http://en.wikipedia.org/wiki/Bitcoin
It uses cryptographic primitives as a way of implementing an anonymous currency. Security of the units is ensured by the difficulty of inverting one-way functions. Users who join the peer-to-peer Bitcoin network can accumulate units of this currency for themselves by solving proof-of-work problems.
You pass on a money unit through a cryptographic protocol. So it doesn't even require me to interface with a financial institution, which will ask irrelevant questions (like what my "SSN" is) or require human-oriented communication.
How come none of you suggested this to me before?
Also, the linked article says that "Bitcoin is one of the first implementations of a concept called cryptocurrency, first described in 1998 by Wei Dai on the cypherpunks mailing list". Is that "Wei Dai" the same as User:Wei_Dai?