Hal Finney was a programmer and cypherpunk: an advocate of privacy protected by cryptography. He worked on PGP, created the RPOW experiment and helped test early Bitcoin. He received a test payment of 10 BTC from Satoshi.
Imagine sending a letter that intermediaries cannot read along the way. Encryption helps provide that protection online: it transforms a message into something readable only with the appropriate key. Practical possibilities like this interested Hal Finney long before Bitcoin. In his 2013 recollection, he described working with Phil Zimmermann on PGP, software for protecting communications. He also ran an anonymous remailer, a service that forwards messages to help conceal the sender. He belonged to the cypherpunks, who sought to support privacy through usable software. A natural question followed: could similar tools protect digital payments too? [1]
On 15 August 2004, Finney introduced RPOW, short for Reusable Proof of Work. It built on Hashcash: a computer searches for a result that requires computational effort but is easy for others to verify. Such proof of work can, for example, make sending bulk spam more expensive. Finney’s server accepted these proofs and issued tokens signed using RSA, a cryptographic method for verifying the issuer’s signature. Think of a token as a digital receipt representing work done, rather than a physical coin. It was not BTC and did not use Bitcoin’s rules for agreeing on payment history. This was a different experiment that actually ran. [2]
You can send an ordinary file to Anna and keep a copy. That is a problem for money: a recipient cannot safely accept value if its previous owner can still spend it. RPOW addressed this through sequential exchanges. A submitted token was used once, and the server issued a new token of equal value; it was supposed to reject the old one on a second attempt. When Anna passed the value on, another exchange followed. What travelled repeatedly was the value represented by the proof of work, not permission to spend the same token simultaneously in several places. That sequence is what “reusable” meant. [2]
Could the operator secretly issue extra tokens? RPOW used an IBM 4758 security card to address that question. The card could provide a signed statement identifying its running software, a process called remote attestation. A user could compare the attested program with the published source code. The aim was to reduce reliance on the operator’s word. Trust did not disappear: the design depended on the hardware working correctly and on verification of its origin, including IBM’s role. The operator could also disconnect the server. A verifiable server was therefore not a decentralized blockchain with a network of participants checking its history. The archive now marks the project as terminated. [3] [5]
When Satoshi released the first Bitcoin software, Finney tried it. In his 2013 recollection, he described receiving 10 BTC from Satoshi as a test transaction. This was neither a purchase nor a block reward: it tested whether one participant could transfer existing coins to another. That distinction matters for beginners. Mining helps record payments in a shared history and carries a block reward; an ordinary transaction changes who can subsequently spend particular funds. Finney was therefore more than a reader of a proposal. He was the person at the other end of a working test. [1]
Finney did not describe those first days as a finished, flawless product. He exchanged messages with Satoshi, reporting bugs that Satoshi fixed. That is an important part of development: someone runs the program in real conditions, describes a problem and helps verify a correction. His account even includes a mundane detail: he eventually stopped mining because his computer ran hot and the fan bothered him. It makes early Bitcoin feel less like a predetermined legend and more like software people were learning to use. A tester contributes effort and expertise; that does not give them a permanent right to decide for every user. [1]
On 17 March 2013, Finney introduced another experiment, bcflick, inspired by suggestions from Mike Hearn. Its question was straightforward: could malware be prevented from emptying a wallet in one go? The design used Flicker to run protected code briefly and a TPM, a computer security chip, to protect sensitive data. Transaction signing was to pass through a module enforcing a daily limit. A limit would constrain an attacker, not automatically prevent every theft. The experiment required compatible hardware and secure initialization, and allowed an authorized user to exceed the limit while operating in a safe mode. It is a historical protection design, not a current wallet recommendation. [4]
Bitcoin and me, his autobiographical recollection from 2013, also reveals another part of his life. Finney described how ALS progressively limited his movement and how he operated his computer through eye tracking. Programming remained meaningful to him even though it had become much slower. His story is therefore also about independence and being able to continue working. The account should also be read as personal testimony: his recollection of correspondence does not establish who was behind the name Satoshi Nakamoto. His own achievements need no such answer. They remain documented in programs, announcements and collaboration that helped turn cryptographic ideas into practice. [1]
For the clearest picture, read this entry together with Financialization, Adam Back, Reusable Proofs of Work, Satoshi Nakamoto. The reverse links also lead from Satoshi Nakamoto, Hashcash, Bit Gold, Reusable Proofs of Work.