Digital money has an old problem. A file can be copied. If a coin is just a file, you can send the same coin to two people. For decades the answer was a trusted middleman, such as a bank, that keeps the ledger and says which payment came first.
In 2008 a paper signed Satoshi Nakamoto proposed removing the middleman. It describes a network where strangers keep one shared ledger, with no one in charge, and where rewriting the past is made enormously expensive. The paper is nine pages long and uses a few ideas you may already know: hashing, chains, and a vote by computing power.
Bitcoin orders payments with a chain of blocks. Each block contains the hash of the one before it, and adding a block requires a lot of work. To change an old payment, an attacker would have to redo the work for that block and every block after it, faster than the rest of the network adds new ones. Nodes always follow the longest chain.
The double-spend problem
The paper opens with the trouble in online commerce. It relies almost entirely on financial institutions as trusted third parties. It argues for payment based on cryptographic proof instead of trust, so two willing parties can deal directly.
Digital signatures already prove who owns a coin. What they cannot prove is that the owner did not sign the same coin over to someone else first. The payee needs to know that, of all the payments of a coin, the earliest one is the one that counts. So the network needs a way to agree on the order of payments, without trusting any single party to keep that order.
A chain that is costly to change
The paper's answer is a distributed timestamp server. Group payments into a block. Put the hash of the previous block inside it, so each block fixes everything that came before it. The result is a chain.
