CoinGeek Conversations

Stuart Haber and Scott Stornetta: How our timestamping mechanism was used in Bitcoin

October 27, 2021 CoinGeek Season 6 Episode 6
CoinGeek Conversations
Stuart Haber and Scott Stornetta: How our timestamping mechanism was used in Bitcoin
Show Notes

Working at Bellcore Labs in the 1990s, scientists Stuart Haber and Scott Stornetta were allowed to pick their own research projects. For Scott, a problem worth solving was the authentication of digital documents: how could you be sure the version you were looking at had not been altered from its original? 

The two worked as a team and were on the point of proving to themselves that the question of trust made the problem insoluble: there always needed to be some independent person or body to verify authenticity – but what if they were also part of a collusion?  

But then, Scott says, he had a brainwave. If you needed to keep adding extra trusted parties to vouch for the honesty of the existing players, then, logically, the list would expand infinitely until the whole world was required. And that still wouldn’t be enough.

Scott’s insight was, as he explains it, that “I realised that if you turn that upside down and created a system of interlinked documents with essentially everyone as a witness, then you had, in fact, solved the problem.”

As Stuart points out, time is an agreed social construct. So, using time means that, in effect, the world would be able to validate your document as having existed in a particular form at a particular time. 

On this week’s CoinGeek Conversations, Scott and Stuart talk about how they turned their theory into working code and how that became a key part of the Bitcoin White Paper.

Their original paper, How to Timestamp a Digital Document, was published in 1990. It showed how the various elements of a timestamp verification system would work. In practical terms, they used hash functions to generate the additional files. Stuart describes a one way hash function as “a way of …taking the fingerprint of a file”. And that was the basis of the system that Bitcoin uses to this day:

“Now we call it a blockchain, but we built a chain of blocks. And in order to achieve worldwide agreement …on the registration requests that our service had received once a week, we would compute a fingerprint, a small string of numbers and letters …that efficiently summarised the entire week of requests that had preceded it, and we published that.”

The two eventually started their own company, Surety, which operates to this day. And the Sunday edition of the New York Times still publishes, as an ad the company pays for, the hash that acts as the ‘fingerprint’ for all of the last week’s transactions. 

Scott says he and Stuart welcomed Satoshi Nakamoto using their timestamping work in Bitcoin: “it was terrific in my mind that on top of a foundational layer, Satoshi had this terrific innovation.”

An important part of Satoshi’s development of the idea was to limit it to money, Scott says. They had had wider ambitions, but a more focussed approach was more productive: “we expected all the world's financial records would be done with this system. But in addition, all the world's photographs, all the world's movies. So we weren't short on ambition, but I think there was great insight to limiting it to a very specific use case.” 

Stuart says he and Scott enjoyed being invited to the CoinGeek conference in New York, where they met Dr Craig Wright for the first time: “Scott and I both both met Craig …and we've enjoyed our conversation, great conversation with him, both on stage and off stage, about all sorts of topics - not all of them involving hash functions.”