Terms like: “trustless network”; “getting rid of trusted intermediaries” and “trust replaced with mathematics” crop up all the time. Trust comes up so often and is clearly a fundamental aspect of why people think blockchain technology is important, but what does it really mean? Is it true that blockchains are “trustless”? Can code really replace such an abstract quality? And if so, why is that important?

“Old-Fashioned” Trust

When I set up an account on say, Coinbase or Localbitcoins and trade in my USD or GBP, I’m still trusting that I’ll be receiving some bitcoin — there’s no mechanism woven into the fabric of the universe that guarantees I will. And every day, I’m trusting that whomever owns the Coinbase server doesn’t suddenly decide what’s mine is theirs. Clearly the need for trust still exists in the stack — exchanges like Coinbase and Localbitcoins are just third party on/off-ramps for the underlying Bitcoin blockchain. So here, the “old-fashioned” need for trust remains intact (for now).

“Trustless” Refers to the Underlying Blockchain Protocol, Not the Rest of the Stack.

When people say “trustless”, they’re referring to the actual blockchain itself, not the exchanges, nor any other part of the value chain. Committing transactions to the blockchain and keeping track of who owns what doesn’t need a central authority. Therefore there’s no focal point of reliance — or “trust” — in the same way that you would rely on your bank to maintain your balance and execute transactions, or trust that Mr Madoff really did invest your money where he said he would. By giving everyone a copy of the transaction ledger, we no longer need a central authority to govern and administer the state of the world, and therefore the system is transparent and trustless. Bingo!

Not Quite…

That still doesn’t explain why the system is “trustless”. All we’ve done is replace the need to trust a central authority with the need to trust a distributed “black box”, and for lots of people the former is probably more reassuring! We need to dig a little further. The blockchain comprises a “chain” of “blocks”, hence the name. Put very simply, each block contains information about which transactions have occurred in the last 10 minutes or so, and is tacked on to the end of all the previous blocks holding information about transactions in previous 10-or-so minute periods, all the way back to the genesis block in 2009. In the old world, a central authority (that you must trust) would tally all these together and execute on them. In the world of blockchain, the community comes to agreement.

But Without a Central Authority, How Does Anyone Know What’s What?

You use a consensus mechanism: a set of rules that governs how the community agrees on which transactions took place. Bitcoin uses a Proof-of-Work consensus mechanism, but lots of others exist. The point is to ensure that “the truth” is established, and nobody is able to write anything to the ledger that is untrue. Feel free to explore the technicalities elsewhere, but some detail is needed to complete the story. Lots of computers in the ecosystem (aka miners) are working to establish which transactions to include in the next block, and to do this they’re expending a famously vast amount of computing — and therefore costly electrical power to solve a problem linked to the transactions. Crucially, these problems are hard to solve, but easy to verify.

As soon as one party claims to have solved the problem, others will seek to verify their achievement. When enough verifications are achieved, the next block of transactions is committed to the blockchain and the miner who solved the problem will receive some bitcoin as a reward. If anyone attempts to write false transactions to the blockchain, they will not receive verifications from other honest members of the community. The huge cost of solving the problem in the first place means there is no incentive for one bad actor (or small group of actors) to mess around, because they won’t get the verifications they need to complete the fraudulent act.

Trust… Sort Of

It is possible to imagine a situation where a group of colluders get together: some write bad transactions and others deliberately verify them (commonly known as a 51% attack), but in order to succeed a large proportion of the mining community needs to be in on the game. By transacting on the bitcoin network, I’m still trusting that this isn’t happening — that large swathes of the community are not acting nefariously in concert — but the requirement for trust is no longer centered upon an individual actor. It’s a lot easier for a single bank manager to embezzle money than it is to coordinate 51% of the bank’s customers to commit fraud together, and blockchain technology essentially denies the possibility of the former.

Towards a Definition

By “trustless” we could say that we mean the consensus mechanism is designed such that there is no incentive, nor possibility, for any individual or reasonably-sized group to behave badly. So far, it’s working well. For years, there’s been a multi-billion dollar prize (i.e. Bitcoin’s market cap) for anyone who can break this system and yet nobody has done it, despite the massive financial incentive. Sure, we’ve all heard of the hacks that happen periodically, but these are a problem that plagues exchanges rather than the underlying protocol. Lots of smart people are working on decentralised exchanges to deal with problems of security and hacking, but until they’re solved, exchanges remain a security hole.

Why Does Trust Matter?

By spreading trust over a much larger surface area, blockchains (or more generally distributed ledger protocols) allow us to create incentive structures that could lead to very different ways for us to organize ourselves. Our ability to organize is arguably the most powerful human trait, and changes therein have been among the most powerful forces in human history.