Lost Keys Are Not a Reason to Print More Bitcoin
An old debate got a new twist today. Eli Ben-Sasson, CEO of StarkWare, posted on X that capping Bitcoin's supply at 21 million does not actually make sense.
His main argument is that private keys get lost over time, and if you look at an infinite time horizon, every key will eventually be lost.
Other arguments came up in the discussion too, but I had heard all of those before. This one, about the infinite time horizon and lost private keys, was new to me.
At first glance, it might sound logical. But in my view, Ben-Sasson is conflating the number of bitcoins with the number of private keys. New addresses appear on the Bitcoin blockchain in huge numbers in every block: thousands of new addresses per hour. Each address has its own private key. That means wallets are generating thousands of new private keys every hour, and that number has nothing to do with Bitcoin's issuance schedule.
Sure, the number of possible private keys is technically finite too. But at the current rate of key generation, it will not run out on any timescale that matters in practice. Even at a rate of 1012 keys per second, it would take roughly 3.7 × 1057 years to exhaust the keyspace. Humanity will be long gone before that happens.
What's more, even if Bitcoin's issuance cap were raised, it would not change the private-key limit at all. So the problem Ben-Sasson raised is not actually solved this way. And every other argument in the thread has already been hashed out before.
That said, if anyone genuinely wants to raise the issuance cap, they are free to fork Bitcoin and do it. Maybe that is exactly what Ben-Sasson is hinting at. Though I suspect such a fork would not fare much better than StarkWare's STRK, which is currently available for BTC swaps on rabbit.io at a rate of 0.00000048 BTC per STRK.
