Earlier this week I commented on End-to-end arguments about Bitcoin.
Later this week we saw this same issue very clearly in the testimony at the New York financial regulator’s hearings (eg. see NYT report). These hearings are chewing over how to make Bitcoin fit into the existing financial system.
For those of you not familiar with the term, “end-to-end arguments” in engineering refer to a a very general issue in system design. I generally refer people to an excellent old paper,
Saltzer, M., D. P. Reed, and D. Clark, End to end arguments in system design. ACM Transactions on Computing Systems, 2 (4):277-288, 1984. ((A similar paper is available online.)
The essential point is that it is critical to trace communication or data movement (the same thing, of course) all the way from one end to the other. In the case of privacy, this means that you need to establish a private channel from the sender (human) to receiver (human). The paper makes clear that is it not sufficient to have parts of that transaction private. For example, encrypting the mailbox on a server does not make the email private, unless the channels to and from the server are encrypted. In fact, encrypting on the server is nearly useless as far as privacy is concerned. Look at the paper for a very clear explanation.
This principle must be applied to Bitcoin and other online transaction systems. In particular, there are various claims about the protocol and systems, such as “secure”, “transparent”, and so on. These claims must be analyzed in an end-to-end framework.
The hearings in Manhattan aired various claims and critiques about Bitcoin from different points of view. One thing that came out of the hearings, aside from the fact that some people are more interested in “free” and others in “lawful” commerce, is how end-to-end analysis is critical for thinking about these issues.
The clearest case is in the claims of “transparencey”. Bitcoin is supposed to be both anonymous and transparent (which seems contradictory on the face of it). The “transparency” refers to the fact that all transactions are known, and can be traced. This is an essential property of Bitcoin and similar systems, and without it, Bitcoin would not work.
Proponents argue that this also makes criminal behavior unlikely, because everything can be known and proven by law enforcement.
But law enforcement witnesses counter that the open ledger can trace only the specific computer address (and anonymous public key), not to the person or organization acting. From the point of view of investigators, this is not transparent at all, and is a boon to criminals, tax evaders, and terrorists (these are not equivalent, but all share the desire to hide money from governments). (See earlier post about hard core financial warfare.)
My view is that both of these points of view are correct, though they are using different definitions of “transparency”.
Within the Bitcoin system, everything is, indeed, completely transparent. It is a beautifly elegant system, and pretty much impossible to cheat–from wallet to wallet. It would not be wise to try ti hack Bitcoin, or to try to cheat people within the system. You will be caught, probably before any gain.
From the point of view of financial regulation, though, the “system” extends from sender to receiver, person to person. Bitcoin has no information about the people. By its very design, Bitcoin ends at the “wallet”. Anyone can send BC to anyone, and there is no way for anyone to know who is who from the Bitcoin system.
So, to make Bitcoin “transparent” is the same way that conventional banking is transparent will require mechanisms outside Bitcoin, which will be the same mechanisms used in conventional banking, because goal and problems are the same: to prevent anonymous financial flows.
Or, as some people would see it, Bitcoin is going to have to follow the same rules as everybody else. Sorry to tell you that, kids.
It is very, very important to understand the “end-to-end argument” here. Not only are their different points of view, “transparency” of the Bitcoin protocol and system is essentially irrelevant to the requirements for “transparency” from sender to receiver. It is not sufficient, and not even necessary. (Read the paper I cited.)
I note that at least some grownups are beginning to talk sense. For example, Fred Ehrsam of Coinbase is quoted in the NYT to say “Regulation could be a good thing.” I agree. No way is Bitcoin going to operate outside the international banking system. Not. Going. To. Happen.
So let’s get things worked out reasonably, and get Bitcoin inside the tent.
Robert McGrath's Blog 