b-money, Bit Gold and Hashcash: the almost-bitcoins that each held one piece
Hashcash (1997), b-money (1998) and Bit Gold (1998, published 2005) each solved part of the digital money puzzle: costly work, a shared ledger, coins made by computing. None put all the pieces together. Bitcoin did, in 2008.
In every gold rush, the first prospectors don't arrive together. One finds a creek with colour in it. Another works out how to pan. A third draws a rough map of the hills. None of them has the whole picture, and most of them go home.
Digital money had three prospectors like that in the late 1990s. Adam Back made computers pay for things with work. Wei Dai described a ledger kept by everyone. Nick Szabo described coins made from computing, chained one after another.
Each held one piece of Bitcoin. None of them built it. That took another ten years and somebody who put the pieces on one table.
What they are
- Hashcash is a way to make a computer prove it did some work before sending something. Adam Back first proposed it in May 1997 to slow down spam (Back, 2002).
- b-money is a short text by Wei Dai, dated 1998, describing money run by anonymous participants who each keep the accounts (Dai, 1998).
- Bit Gold is Nick Szabo's design for "unforgeably costly bits" created online. He says he came up with it in 1998 and described it in public in 2005 (Szabo, 2011).
How each piece worked
Hashcash: the pickaxe. Back's 2002 paper says hashcash was proposed "to throttle systematic abuse of un-metered internet resources such as email". The idea: before your email is accepted, your computer must keep guessing until it finds a hash that starts with a set number of zero bits (Back calls this finding "partial hash collisions"). There is no shortcut, only guessing. Cheap for one email, expensive for a million spam emails. Back calls the result "a proof-of-work".
b-money: the shared ledger. Dai's text opens with a nod to the cypherpunks: "I am fascinated by Tim May's crypto-anarchy." Then the key line: "Anyone can create money by broadcasting the solution to a previously unsolved computational problem." In his first version, "every participant maintains a (seperate) database of how much money belongs to each pseudonym". The spelling is his. The idea is Bitcoin's ledger, ten years early. In a second version, the accounts are kept by a subset of participants he calls "servers".
Bit Gold: the chain. Szabo starts from the same complaint as everyone else: money that depends on trusting a third party. His fix: a computer solves a puzzle, and "the resulting string of bits is the proof of work". Then: "The last-created string of bit gold provides the challenge bits for the next-created string." Each piece of gold points to the one before it. That's the shape of a blockchain.
The step-by-step trail
- May 1997. Back proposes hashcash (Back, 2002).
- 1998. Dai publishes b-money. Szabo works out Bit Gold "on the same private mailing list where Dai was coming up with b-money" (Szabo, 2011).
- 1 August 2002. Back publishes the full hashcash paper.
- December 2005. Szabo posts Bit Gold on his blog (Szabo, 2005).
- October 2008. The Bitcoin white paper cites b-money as reference 1 and Hashcash as reference 6 (Nakamoto, 2008). Bit Gold isn't in the list of eight references.
What was missing
The almost-bitcoins each had a hole.
- Hashcash made work costly, but the work wasn't money. You burned it on one email and it was gone.
- b-money had a shared ledger, but no clear way for strangers to agree on which version was right when two people disagreed.
- Bit Gold made scarce bits, but Szabo still leaned on outside timestamp services, ideally several "so that no particular timestamp service need be substantially relied on".
Bitcoin's trick was to combine them: hashcash-style work decides who adds the next page, everyone keeps the ledger, each page points to the last, and the longest chain of work wins. No timestamp service, no server, no bank.
Why it matters
- Bitcoin wasn't a lightning bolt. It was assembly. Almost every part had been tried in public before 2008.
- The credit is shared. When people call one person "the inventor of crypto", they skip a decade of prospectors.
- The miners' pickaxe came from spam. Proof of work, the thing Bitcoin miners burn electricity on, started life as a way to make junk email expensive.
Where people get it wrong
- "Satoshi copied Bit Gold." The white paper doesn't cite it. Bitcoin shares ideas with it, but the parts that make it work together are new.
- "b-money was launched." It was a text, not a running system. Nobody ever paid for anything in b-money.
- "Hashcash was a currency." It was an anti-spam stamp. The paper's title is "A Denial of Service Counter-Measure".
- "These people were trying to get rich." They wrote proposals on mailing lists and blogs. There was nothing to sell.
The sting: ask which piece is new
Every few months a new coin claims to be "the next Bitcoin". Ask the 1998 question: which missing piece does it add? If the answer is a new logo and a presale, it isn't a piece. It's a souvenir map of a valley that was mapped in 1998.
Frequently asked questions
What is Hashcash?
Hashcash is a proof-of-work system Adam Back proposed in May 1997 to make spam expensive. A computer has to do a small amount of guessing work before a message is accepted. Bitcoin mining uses the same idea, and the white paper cites Back's 2002 paper.
What is b-money?
b-money is a 1998 proposal by Wei Dai for digital money run by anonymous participants. Money is created by solving computational problems, and in the first version every participant keeps a copy of the accounts. It was never built, but it is reference 1 in the Bitcoin white paper.
What is Bit Gold?
Bit Gold is Nick Szabo's design for scarce digital bits created through proof of work, with each new string built on the one before. Szabo says he came up with it in 1998 and posted it publicly in December 2005. It was never launched.
Did Bitcoin copy these ideas?
Bitcoin combined ideas that existed before it, and the white paper cites b-money and Hashcash. What was new was the way it put them together, so that strangers could agree on one ledger without any central server or timestamp service.
Why did none of them become money?
Hashcash wasn't designed as money. b-money and Bit Gold were proposals that were never released as working software, and both left open how strangers agree on a single record. Bitcoin, released in January 2009, was the first to run.

