[jstanley]

How could a cryptocurrency be backed by something?

If I need to "trust" you that I can get <something> in exchange for my cryptocurrency, then that's not much of a cryptocurrency is it? It may as well be a MySQL database controlled by you, as that's what it amounts to anyway.

[asgioiobuio]

I've been toying with the idea of a computation-backed currency. Any computer can obviously pay another computer in computer time. This means that any node in the system can issue currency backed by computer time. This would require trust, but it would be trust between neighbors rather than trust in one central exchange.

The problem is that that's not really a currency. It's zillions of IOUs. I don't know how to make transactions between two nodes that don't trust each other or how to come up with one number for someone's net worth.

[jgautsch]

Some economists would argue that "zillions of IOUs" is basically the precursor to currency/money. Then you just need a consistent unit, and that ability to trade/transact (which is a solved problem).

> Graeber proposes that money as a unit of account was invented the moment when the unquantifiable obligation "I owe you one" transformed into the quantifiable notion of "I owe you one unit of something". In this view, money emerged first as credit and only later acquired the functions of a medium of exchange and a store of value. [1]

[1] https://en.wikipedia.org/wiki/History_of_money

[jgautsch]

The difference between a cryptocurrency backed by something and a MySQL database is that in the former case, underlying value is indeed centralized, but in the latter, both underlying value and the exchange of stake in that value is centralized.

Like chips in a casino can be exchanged with the house for cash, but I don't need to tell the house that I'm giving some chips to my friend.

[jstanley]

Fair point, but that's then no different to Tether.