[jrimbault]

I was thinking with some friends recently (new year's eve) : considering a bitcoin model with a fixed finite amount of currency, won't every coin be lost at some point due to storage failure/lost keys/etc ? Statistically ? And rather sooner than later, if my thinking is right ? Like the birthday problem ?

There is a maximum of 21x10^6 bitcoins, imagining a 0.01 chance of losing 1 bitcoin/day ?

[kim0]

Correct, that's why Monero applies a tail emission that offers less than 1% constant inflation, closely modeling actual gold.

Also it's untested if miners will continue mining after Bitcoin inflation completely stops

[Geee]

Yes, but every bitcoin is divided into 100000000 satoshis, and it's possible to add even smaller units in the future.

[jrimbault]

Who and/or what decides by what mechanisms and when satoshis would be divided into smaller units ?

Doesn't that make it virtually valueless by definition ?

[Geee]

No, it doesn't change the value. It's just more decimals. A hard fork would be needed to make the change in the protocol.

Similarly a bank can use whatever amount of decimals they wish to store their dollar amounts, it doesn't create new money. You can also divide gold into infinitesimal amounts.

You can already send millisatoshis on the Lightning network, which is rounded to a nearest satoshi when it's settled on the blockchain.

[duncan-donuts]

As the scarcity increase would they become more valuable or less valuable?

[zhoujianfu]

Typically in economics less supply increases price.