I’ve been having a lot of trouble with banks lately. Either they don’t trust who I am, or I don’t trust who they are (and what they’re doing with my money). For instance, I went into my local bank to withdraw some cash. They proceeded to ask me for identification and I told them I had none. They informed me that the transaction could not be completed without ID because of identity-theft. I explained that I had no ID because I did not want it to be stolen and one cannot steal something I don’t have. I then asked them for ID. They explained that all they had was a namebadge. I informed them that the transaction could not be completed without ID because of identity-theft. I promptly left, leaving a very perplexed bank official in my wake.
One of the easiest ways to protect against problems in computer science is to modularise a program. Thus one faulty or deliberately malicous component does not have the ability to interfere with another component. Java achieves this using type-safety and the object-oriented methodology, however it can (and has) been done in C using the memory protection provided by the process abstraction (think
In computer science we can easily seperate concerns. I’d like to apply this to banking. I’d like to be able to organise my cash and credit into separate contexts so that we can lower the authentication barrier so that I don’t need a DNA test any time I want to withdraw Â£20. In fact I think that the imposition of technology into banking has caused this problem that they nearly need biometrics to verify who you are. The old system of actually knowing your bank staff is impossible these days given the employee churn they seem to have.
So I could ssparate my cash and credit accounts. I could have an electricity bill account, which I credit with small amounts any time I need to pay a bill. This would mean that if an electronic interloper or incompetant bank official were to
cat /dev/random > AidansElectricAccount I would only loose a small amount of money. I’d be spreading the risk and be able to use strong authentication on the account that receives my wages, but weaker authentication on accounts that contain little.
What would be really cool is if I could decouple my identification from the account. Given a bearer document or code or such similar there would be no need to store my height, eye-colour and mothers maiden name or other personal data. Thus I get more privacy. I can buy my monthly copy of ILoveNakedPeopleInAnimalCostumes Journal using the credit in my anonymous account.
So with the rising cost of identity-theft and the lowering cost of adminstration of multiple accounts will this ever happen? I doubt it. The banks would loose control over us, their minions, and the government might loose a little cash to money launderers. That’s not considering that any smart money launderer is already getting away with their crime or simply becomes an Irish politician and opens an off-shore account.
For once I’d like to see something that both protects my monetary security and protects my privacy. Otherwise I’ll have to start my own bank
But think of the boost this might give to web commerce if you had a web payments account which you only credited with the minimum amount of cash that you needed for the particular transaction. There would be little fear in handing online retailers Â£100 or so as that would be the most you could loose. It would suck to loose Â£100, but it’s a lot less than the possible one months wages that you could loose from the one account you’re probably dealing out of now.
In fact you could buy something from an online retailer by simply transferring ownership of the account to them. It can be done as identity is decoupled from the account. Just like cash!