RPOW — Reusable Proof Of Work

BUT THEY ARE REUSABLE.

In 2004, Hal Finney introduced RPOW, an experiment in reusable proof-of-work tokens.

In the dark workshop, Hal Finney holds up a bright new token. A stamped envelope waits in the input tray, while an older token sits inside the machine. The warm light falls across his face and the secure hardware. This is an imagined way to show a change in what proof of work could do: a proof used to make something that could be passed on.

On 15 August 2004, Finney invited the cypherpunks mailing list to test his new server at rpow.net. It accepted a Hashcash proof-of-work token and returned an RSA-signed token he called a Reusable Proof of Work, or RPOW. A recipient could later exchange that RPOW for a fresh one of equal value. Each individual token could be spent only once, yet the chain of exchanges let the value move from person to person. As Finney put it, unlike Hashcash, “they are reusable.”

The promise depended on trust in the machine doing the exchange. Finney ran the server on an IBM 4758 secure cryptographic coprocessor. Its signed attestation was designed to let a sufficiently motivated user check the software configuration against the published source code. That was the safeguard behind his claim that the server would issue a new token only when it received an equal-value proof. RPOW was an experiment with a trusted server, not Bitcoin or a decentralised blockchain.

Finney called the system an early beta and warned users that a server reload might invalidate tokens already made. The empty output tray in our scene catches that uncertainty alongside the excitement. In the foreground, Eric Hughes's earlier manifesto and the Hashcash envelope connect cards 002 and 003 to this one. Their presence is HISTROVE's visual storytelling, not a claim that these objects sat on Finney's actual desk.