DISCOVERING DIGITAL GOLD

Guest Written by Aaron Van Wirdum

Nick Szabo was one of the Cypherpunks who worked at David Chaum’s Digicash startup. But as he laboured away in their Amsterdam office, he found that Ecash had a fundamental flaw: it was centralised. Employees could potentially manipulate the Ecash system by secretly issuing more of the electronic cash notes.

This revelation made Szabo realise that internet protocols really shouldn’t rely on trusted third parties for their security. “Trusted Third Parties are security holes,” he’d explain in one of his popular blog posts.

About a year after Adam Back had proposed Hashcash, Szabo used its proof of work function to design an electronic cash system that, crucially, did not rely on any single person or company to operate: Bit Gold.

As with Hashcash, Bit Gold units were created through hashing, where only some hashes would be considered valid. But instead of emails, users would have to hash a publicly known “candidate string” – basically just a number.

When someone found a valid hash, they would announce this to other Bit Gold users. Interestingly, they would then get to “own” this hash: it would be attributed to their public key.

If they later decided to spend the hash, they would publish a message announcing the new owner of the hash (or rather, their public key), and sign this message using their own private key.

There were, however, two enormous challenges Bit Gold faced before it could be used as money.

The first of these was inflation. Over time, as computers become more powerful, it would be easier and easier to find valid hashes. New Bit Gold hashes could flood the system, eventually rendering the electronic cash worthless. To solve this, Szabo envisioned an elaborate system of special exchanges and new types of banks that would trade and collect Bit Gold hashes and use them as backing for a digital currency system of their own. A viable solution, perhaps — but also one that complicated the overall design.

The other huge challenge was keeping track of who (i.e. which public keys) owned which Bit Gold hashes. For this, Szabo envisioned a special registry. But Trusted Third Parties are security holes, Szabo knew, so he envisioned this registry having to be maintained by a group of users rather than one person or a company.

In the case of a conflict — for example because a single private key signed two or more conflicting transactions (a “double spend”) — this group of users would use a sophisticated voting scheme to determine the true new owner.

Unfortunately, Szabo did not detail who or what these users would be, or what would stop them from colluding and cheating the system. Or worse, what would stop a single person from pretending to be several of these users? 

In the end, the Bit Gold design was not quite equipped to solve these kinds of problems and, perhaps for this reason, was never actually implemented. Szabo’s proposal did, however, represent another giant leap in the concept of electronic cash.