[cap-talk] Bit gold

[Nick Szabo]

I would be quite remiss to fail to mention here RPOW, Hal Finney's actual 
working version of bit gold (as compared to my mere sketch).  RPOW uses 
a different solution for the vulnerability-to-third-party (V23P)
(formerly known as "trusted third party") problem -- a single server 
on a tamper-evident card running audited code, the state of which 
can be remotely attested.  Not only is RPOW a great idea, 
but the idea of remotely attesting servers instead of clients is a 
huge, huge idea applicable to a wide variety of V23P problems.  Remote 
attestation of servers is much more likely to be a success than remote 
attestation of clients.  See
	http://rpow.net/

> On that very different topic of money:
> > Of course any such mechanism of value exchange must map
> > into other means of exchange - like dollars.
> 
> Bit gold need not be "mapped" to any other kind of money, although
> I'm sure if it became widespread exchanges would pop up.  The value
> of gold and other unforgeably costly things, unlike fiat or 
> credit instruments (including all current and envisioned
> Internet payment systems besides bit gold), exists without the 
> need for trust in (reliance upon upright behavior of) human 
> institutions.  Trust-free, unforgeably costly money has far
> more history behind it than, say, dollars -- see 
> 	http://szabo.best.vwh.net/shell.html.
> 
> Bit gold would come very close to this ideal (i.e., it would
> achieve trust-free money up to Byzantine agreement in the titles 
> directory and timestamping service, and its reliance on certain 
> cryptographic assumptions that may be shaky).

Alan's suggestion helps to clarify this sentence -- the goal here
is to achieve vulnerability-free money, not to achieve money
that can't be trusted!

Nick Szabo