On 11/06/11 7:42 PM, Eugen Leitl wrote:
On Sat, Jun 11, 2011 at 02:16:55AM -0000, John Levine wrote:
In article<021ccba9-9203-4896-8412-481b94595...@cs.columbia.edu>  you write:

I wouldn't call bitcoins digital cash.  They're more like digital
tulip bulbs, or bearer shares of theglobe.com.

Whatever they are, it's a self limiting problem since the bubble will
burst soon enough.

Unlike fiat currencies, algorithms assert limit of total volume.
And the mint and transaction infrastructure is decentral, so there's
no single point of control. These both are very useful properties.

Cryptographically and computer-science-wise, very elegant. But that doesn't mean they are useful properties to any particular application.

We expect money to be a store of value, among a few other things. BitCoin has nothing in it that speaks to that goal, that I can see [0]. This anti-property would however make it more ideal for a bubble [1].

Quite how to fix that and retain the decentralised control aspect, I'm unsure. The essence of a contract is that someone delivers something to someone else; without that first party (which speaks to centralisation at some level) it is hard.

Can we have a community promise that maintains in a distributed contract? And delivers a store of value? I'm not sure. It's certainly an interesting thought experiment, I think I can see how to make it work, but not in the current circumstance.

I don't expect Bitcoin to be it, but it is definitely a predecessor
to a number of such currencies which will become practical both
for machines and people.

That could be! At an architectural level, I think that community is far too obsessed with the "big bad government" scenario when it comes to money issuance [2]. And looking for ways around this, when money is a far more complex subject than just one attacker.

In this sense, DigiCash and BitCoin are cypherpunk bubbles [3]. When we run out of geeks to be fascinated with the cryptographic elegance, the currency will collapse.


[0] It may be that people see decentralisation, and the algorithmic control over volumes as being that store of value. But this is to confuse the primary requirement (store of value) with the secondary governance techniques used by some issuers (i.e., governments) who have a particular flaw of inflation to deal with. IOW, BitCoin's design might be more appropriate if it were run by a government, perversely.

[1] E.g., tulip bulbs were not a store of value (I don't think they could last over the season, they had to be grown that season) whereas it is extremely rare for housing "bubbles" to collapse more than half their peak. And normally it's like 10-20% of their peak, which is not really a bubble, just a correction from a bit too much exhuberance.

[2] David Chaum made the same mistake when he replicated the hand-to-hand payments transaction in blinding digitial coins. His obsession with preserving the privacy feature implemented in the h2h method led to a very elegant cryptographic breakthrough, but took him too far away from the meaning of money and the consumer's real needs in making trade & transactions. A privacy thing that isn't used as money isn't much use. For another small example, the geeks amongst us insisted on client programs to manage the money, but the world transacts these days on browser access (Which by the bye requires some form of client-server architecture). Which can be seen with Paypal; oldtimers will recall it was originally a client program on Palm Pilots doing hand-to-hand transfers.

[3] and, the various other ne'redowellers in there.
cryptography mailing list

Reply via email to