I wrote this in June 2016 but wanted to link to it while writing a piece today.
Three weeks [2019 edit: and 2 years] ago I sat on a panel with the CEO of a company trying to modernize trade finance. When asked to talk about the benefits of blockchain technology, he asserted that blockchain has solved trust.
Blockchain technology hasn’t solved trust. It can’t. Nothing can “solve” trust. Trust is an attitude towards a proposition or entity. Motivations for trusting another party come from a variety of evidence, from faith in the enforcement power of a legal system to the look in someone else’s eyes. Trust is a function of one’s model of the world and no software product can satisfy all possible models. Indeed, some people are destined to harbor inconsolable paranoia.
Practically, trust is a spectrum. For example, I rely on my credit card to finance many of my daily expenses but I also verify that no false transactions were registered when I receive my monthly statement. Most people talk about trust the same way they would about probabilities, with degrees of belief qualified by certain conditions and evidence.
Bitcoin provides very robust methods for ensuring the integrity of its ledger which satisfy most reasonable people. Before Bitcoin, people had to resort to centralized solutions to create a ledger everyone would trust in a distributed solution. Bitcoin maintains a network which is difficult to tamper or cheat through a proof-of-work consensus mechanism and public asset registry, though it also has vulnerabilities with respect to aligning mining incentives to owners of Bitcoin.
Distributed ledgers (and specifically blockchains) facilitate a framework where all parties privy to the ledger can evaluate the provenance of the information in it to determine if there has been tampering. This feature makes them more appealing to trading partners than having trade data managed by a third party.
In theory, a blockchain or distributed ledger prevents the more obvious types of data tampering by validating information submitted to the ledger with a unified protocol and replicating all valid data across many nodes. It falls short, however, when ensuring the integrity of features which fall outside of the ledger’s jurisdiction. Distributed ledgers cannot prevent collusion, for example. So, while I think distributed ledgers are an excellent tool in enhancing the integrity of trade agreements, we should not fool ourselves into thinking that “trust” can be listed in the software spec of any blockchain solution.
