[devx]

I wonder if someone could build a new digital currency that starts let's say at 1 trillion units, and is expanded by 2 percent per year (that 2 percent being split so a slight increase happens every day, so it doesn't happen all at once, but after a year, the increase rounds up to 2 percent).

The problem would be seeding it and distributing it early on. But it could still be mined much like Bitcoin, with the difference being that the total amount will never be limited. Also, it should be easier to mine them, since the (non-finite) total (1 trillion) would be a lot larger than the 21 limit for Bitcoins. Someone should also do the math for this new currency's adoption rate, so all of the 1 trillion units (adjusted with the 2 percent inflation rate) get distributed in 3-5 decades.

Basically we don't want a few thousand people to get that whole trillion within 10 years. I'd also prefer if anonymity would be built it (look to Zerocoin for that). Perhaps it shouldn't be anonymous by default, but you should be able to "switch it on"). It should also be as easy to use and as secure as possible (the clients), so normal people can use it safely on all of their devices.

[vbuterin]

Freicoin takes the inflationary cryptocurrency idea:

http://freico.in/

PPcoin is also slightly inflationary, although it offers you an interest rate higher than the inflation rate if you participate in the validation process (basically like mining except you "mine" with your ppcoins rather than with your cpu cycles):

http://ppcoin.org

Personally, I don't mind the inflation/deflation argument too much; one thing I would like to see, however, is a coin with a semi-centralized distribution model - an organization bound by contract to give X units to every person in the world. Ripple could have been it, but they went with keeping the coins to themselves instead.

[icebraining]

an organization bound by contract

Bound by contract to whom? Who would enforce that contract, and in what jurisdiction?

[brazzy]

The problem is that for a real currency you ideally want inflation to match the growth of the economy. Fixed inflation rates (including 0) are decidedly non-optimal.

[vbuterin]

There are arguments in favor of the fixed supply model. The usual economic one is this:

When you earn and stockpile currency units, what you are doing is producing more than you consume, so in real terms you are allowing others to temporarily consume more than they produce. This essentially means that you are investing the difference into society as a whole. Thus, it makes sense that if the economy grows by 2x, your currency units should grow 2x in value, as that's the way any other investment works.