Re:It Looks Like Nobel Economics Laureates Don't Like Bitcoin
by popinman322 on Thursday November 30, 2017 @11:18AM
(#15816914)
You can get a little safer than escrow via multisignature wallets; you need M-of-N signatures before bitcoins can be sent. You can tie the extra signatures to machine-interpretable signals like the state of a tracked package (when it hits delivered the funds are automatically released, etc.).
Still some trust involved, but the third party can't pocket your money directly.
Re:It Looks Like Nobel Economics Laureates Don't Like Bitcoin
by theoh on Thursday November 30, 2017 @02:13PM
(#15818592)
Presumably you mean "second party", because the concept here is that there is no middleman. EDIT: to clarify, I'm sure there are cryptographic solutions which prevent an escrow agent from doing anything beyond releasing the funds. A system of escrow that involved total trust in the third party escrow service would be an extremely basic proposition. I'd have to research it but it seems like a protocol probably exists; maybe part of the techniques to which you referred. Still, both parties in the transaction count on the escrow service, which is the trust I am talking about.
The example system you mention, which involves having faith in (for example) a courier's delivery notification service, seems to me underline the ubiquitous implicit relationships of trust we implicitly rely on currently, generally being of the form "individual trusts institution/infrastructure".
It hardly needs saying that cypherpunks don't like that unreciprocated trust, or even deny its importance to their own security.
For the important transactions you mention, presumably an escrow service is needed. This reintroduces the requirement for trust in a third party.