Just because it’s on a blockchain it doesn’t mean it’s true
This short article attempts to explain what people mean when they are talking about blockchains being a “single source of truth”. In classic Chinese Whispers style, the narrative has become confused about what is meant by “truth”.
This is currently relevant to discussions in the insurance industry where blockchain enthusiasts may be eager to promote blockchains as a solution to the problem of verifying if something has happened or not.
Here, I permanently recorded on Bitcoin’s blockchain a non-truth about the world.
Graffiti box
When you make a bitcoin transaction, you have the opportunity to type in a short amount of text in a field called OP_RETURN. This gets submitted with the transaction and is stored on bitcoin’s blockchain when the transaction is included in a block.
This is similar to the free-text field in a banking payment where you can type a short note like an invoice number or some initials.
If it’s on a blockchain…
Here’s an example of a transaction where OP_RETURN has been used (hat tip to Eternity Wall):
Notice that at the bottom there is a permanent record on bitcoin’s blockchain with the comment:
a_lewis says 1+1=3 true story
This is “on the blockchain” which some will have you believe makes it “true”.
However it’s not true on two counts:
- One plus one does not equal three
- I (a_lewis) didn’t actually say this
So what is true? What was validated?
Well, inclusion into a valid block means that the transaction was valid – ie it passed some technical requirements (the size of the data was below a maximum threshold, the signature was valid, etc) and some business requirements (bitcoins weren’t created out of thin air, etc). That so much is true.
The block creator (self-reported as CKPool Kano) validated the transaction and included it in block number 412,248. Later, 5,500 or so nodes (according to Bitnodes at time of writing) all agreed that this transaction, and the text within it, was valid and occurred. Each full node has recorded this on their copies of the blockchain.
However the validation done on the content was limited to some technical checks (like message length) and clearly not the logic of the comment (1+1=3) or if the the event actually happened in real life (it didn’t).
When people talk about a single source of truth, they should really be talking about a single, mutually agreed, version of record, but being careful that this is not over-sold as “truth” or “fact”.
The old short piece of text in OP_RETURN!
How does this relate to something like virtual notary, where the explicit goal is to record some ‘truth’ even to be used as proof in legal situations?
I think that the idea of hashing a document and storing the hash on a timestamped blockchain serves as a proof of existence of the document at that point in time, which may be useful eg to prove that I didn’t write last year’s entry in my diary yesterday. However hashing docs doesn’t preclude hashing every card in a pack then producing the “correct” one for climax of the trick.
Separately, notaries serve to affirm that a copy of a document is a true copy, I don’t think they generally certify that the original is genuine, but I could be wrong. If the doc or the hash is cryptographically sighed by the issuer, then that would add value as to how genuine the document is, ie proof of provenance as opposed to just proof of existence.
So, does this mean that blockchains are useless for transactions where we want to to proof that the content is something is true or valid? Suppose I want to use a blockchain for registry of a first idea, invention or patent, how do we know that I was really the first with this invention or idea and that my idea is valid ? Blockchains only confirm that a transaction was made by a certain person at a certain point in time. Correct?
Yes my understanding is that is correct. It’s a place to put a timestamped piece of data, and it doesn’t intrinsically mean that it was the first time it was created. Blockchains also won’t tell you if your idea is valid.