At the moment, most of your income and expenditure probably goes through your bank (or, perhaps, banks). Here in New Zealand, only a handful of banks dominate the market, and I think the processing of EFTPOS and credit card transactions involves even more market concentration.
This centralization of money transfers has several downsides. Fewer players means less competition, greater risk of fraud by employees of the big companies, and obvious targets for malicious hackers. So how could we decentralize money transfers?
One idea came to my attention several years ago, by the name Ripple. I’ll try to explain it.
Imagine Alice wants Bob to do some work for her for $10. She doesn’t have the money on her, but Bob knows Alice and trusts that she’ll get the money to him when he needs it. So he provides the service and accepts an IOU from Alice. Alice owes Bob $10.
Then, Bob wants to buy something from Charles for $30. Charles doesn’t know Bob, so he won’t accept an IOU from him. But Charles knows and trusts Alice, so they come to an arrangement: Alice gives Charles an IOU for $30; Alice owed Bob $10, so Bob cancels that debt and gives gives Alice an IOU for $20. Alice still owes a net $10, so she’s no better or worse off from the deal; Bob now owes $20, instead of being owed $10, since he paid $30 indirectly to Charles, who is owed $30.
This can be extended further. Suppose Denise wants Bob to do some work for $15. Neither Bob nor Alice really know Denise, so they don’t trust her, but Charles does. So Denise gives Charles an IOU for $15, Charles reduces Alice’s debt from $30 to $15, and Alice reduces Bob’s debt from $20 to $5. Alice still owes a net $10 and Charles is still owed $30, so they come out even; Bob now owes only $5 instead of $20, since he was indirectly paid $15 by Denise, who now owes $15.
So money transfers can take place without actual money changing hands, and without anyone having to trust faceless, centralized banks or credit card companies; everyone only has to trust their friends (and can decide for themselves how much they trust each friend).
The problem is that in practice, such chains can be difficult to arrange. In order to pay Bob with an IOU, Denise would need to ask all her friends if they know Bob; when she finds out they don’t, she’d need to ask all her friends to ask their friends if they know Bob. Luckily for Denise, the quest ended here, but you can see how it might easily escalate into something quite complicated.
But supposing that instead of Denise asking her friends if they knew Bob, she and her friends, and her friends’ friends, and Bob, all had a computer program to do the work for them. Denise’s computer asks her friends’ computers if they can complete the transaction; their computers ask their friends’ computers if they can complete the transaction, and so on, until Bob’s computer is asked if it can complete the transaction, at which point, the reply is sent back along the chain, and the other computers are told to stop the search. It could all happen very quickly and automatically.
That, as I understand it, was the idea behind Ripple. But that idea has been around for a while, and so has the internet, but it hasn’t really taken off yet. Why would you run a Ripple-like thing on your computer if none of your friends did? It would just be a waste of time, since you’d have no connections through which to make money transfers.
So my idea is this: Someone should build a plugin for a social network that implements a Ripple-like protocol. People already have connections with their friends via social networks, so half their work is already done; they would just need to set credit limits for their friends, and let the software do its thing.
Now, I’m not talking about Facebook here. Money transfer via Facebook would be about as centralized as money transfer via Visa, Mastercard, or Paypal. For this to result in a truly decentralized money transfer network, it would need to be built on a decentralized social network like Friendica or Diaspora (which interoperate, by the way, so this needn’t be an either/or thing).
It looks like Ripple has recently seen some renewed activity, and I hope they’re successful. (I hope their success is truly decentralized, too; if everyone (or everyone non-geeky) has to sign up through their site, then their site would be a single point of failure.) But I can’t help thinking that they might succeed faster if they build plugins for decentralized social networks, as well as perhaps standalone Ripple software.