The backstory
Looking back, I recall two moments that were pivotal in my early understanding of transactional proof of work and what it meant for blockchains, both born out of an economist’s frustration with trying to fit a round peg into a square hole.
The first was at a New York City diner, having breakfast with a friend a few blocks from Pfizer’s headquarters. We had both been at a “Blockchain in Healthcare” conference the week before and had left with some questions that no one had wanted to answer and a lot of peer pressure to come up with an idea that would suit an ICO. This was at the height of the boom. I had, for months, been struggling to find a viable technology solution for outcomes-based pharma contracting for Pfizer and other clients, and had the gut feeling that the problem needed a distributed database solution, but I couldn’t shoehorn in a token.
“I can’t figure out how to make it work,” I remember saying to my friend. “A token has to either remain at around the same value—or appreciate in value—to make these ICO methods work. But no drug and no contract has a consistent value over time. Ideally, you would want a token to represent each individual contract, but that can’t fit with the ICO model. And it’s hard enough to get the Basel office and the US office of these multi-national pharmas to work with both Swiss francs and dollars, let alone introducing a fluctuating-price token that would have to be used to initiate or manage a contract, so we already know that’s a total no-go.” We were stumped, but we were also right.
The second moment was coming out of a two-day off-site meeting I had been at with Bristol Myers Squibb leaders and physicians from a national research network. We had a several-hour roundtable on value and market access, with all of the participants adamant that one of the paths out of an endless spiral of higher drug prices would be innovative contracts that could link price to clinical outcome, thus incentivizing breakthroughs such as the immuno-oncology drugs developed by BMS and disincentivizing “me too” drugs or those that minimally move the needle. It was a late-season snowy day outside the Denver airport Westin, and I phoned a friend who was good at talking through technical things and who had been keeping tabs on this project. Coming out of the BMS meeting, I felt a renewed urgency to find a solution.
“We have to find a way to gate the mining function based on what’s happening in the real world,” I said. “The network should only allow a block to be created, for example, when the data signals that the contract requirements have been met. But I can’t figure out how to do this with Ethereum or anything else, and most of the developers I’ve talked to think I’m completely out of my mind. I don’t think they work with many economists.”
The breakthrough came very shortly after I was introduced to Arka through a friend of a friend. I had been complaining in a “Ladies Who Tech” social media group about my dilemma, and one suggested that I phone “this guy” she knew. I cold-called Arka from the Whole Foods parking lot in Mill Valley, he was in Emeryville, and thirty minutes later we were having coffee and comparing our notebooks.
In trying to solve a different problem for a different industry, he had also stumbled upon the building blocks of transactional proof of work. We discussed our analogies for this innovation; one of my favorite ones as a sailor of old, leaky, wooden boats was the bilge float switch—once a certain requirement is met and the sensor recognizes it, a circuit is tripped, and the pump cycles on. I envisioned a system that could work similarly—once the real world signals confirm that certain criteria have been met, a switch flips, and the events that compose the qualifying criteria are packaged up and confirmed into a block.
What it does & why it works
The blockchain trailblazers were focused on cryptocurrency applications for good reason. Money is a fascinating and valid use case for publicly-competitive, cryptographic, computational proof of work, and it should be no surprise that Bitcoin, Ethereum, and others have captured the imagination of so many innovators across so many sectors. Blockchain technologies are incredibly powerful.
When we start to deal with inherently valuable data, however, we can unmoor ourselves from external sources of value, whether dollars or Bitcoins, and consider how that data itself can signal valuable outcomes in the real world. This is the scenario we find with healthcare data; the answers to our questions about whether or not the valuable outcome—better health—has been achieved are there, perhaps scattered across a multitude of data sources, but if it’s possible to link them together in a system for distributed, shared answers, the result can be determined, digitally packaged up, and priced.
To solve the outcomes-based/value-based contracting problem, we developed the Miraculum Platform as described in this (updated since the original 2018 publication) white paper. Miraculum uses the Lydion Engine and its method of Transactional Proof of Work for many reasons, but primary among them is the inherent and provable value of healthcare data, as well as the need for a wide variety of data assets to be generated and exchanged by the platform, some of which are expected to appreciate, others which are expected to lose value, and some which have fixed or predicted value but should vary in their speed of generation based on real-world events. Miraculum, we believe, vastly expands the feasibility of many of the blockchain healthcare applications envisioned over the past handful of years, and also meets the legal and regulatory requirements specific to the healthcare industry. (And for those who underestimate the legal and regulatory requirements in this industry, it was either a Roche or Novartis friend who once joked to me that “this company is really just a very large law firm that also happens to manufacture a wide variety of medicines.”)
We hope you learn something from this paper, and we ardently encourage your feedback.