✌️
The future belongs to those who show up. This an open offer to come work with with me disguised as a “provoking by design” set of predictions. Try not to skip ahead. If you know me already this is probably familiar sounding in some ways.
Over the last ten years of startups and software, some of my beliefs about the future have wildly exceeded even my most optimistic expectations. One of those beliefs is in decentralization and the replacement of paper bureaucracies with protocols of permission.
That process is very clearly playing out in finance and banking. Cryptographic proofs of title, identity, and mechanism design are on their way to totally replacing an incumbent industry wholesale. The outside world seems to be recognizing this.
I think those same fundamental forces are going to play out in medicine and the life sciences. That prediction is made complicated by another one. It seems inevitable that the future also has whole new kinds of access control zones and biosecurity measures (real and theatrical alike). A shift that will reshape our relationship to states. Probably first and furtherest in the densest urban spaces. Many globally influential states already issue cryptographic bio-certificates for access control purposes.
That same future though will offer new kinds of asymmetric benefits to individuals — imagine what you could do with a complete medical history of yourself. A case history from birth.
I’ve been thinking about this longer than I even realized at first. Back in 2013, I nearly stumbled upon the same intuition. I’m trying to be impartial but I think you know which part of this future I’m rooting for.
The boundary between traditional kinds of self-surveillance in personal health like sleep schedule tracking and institutional health records is going to collapse. Patient access and the ability to own your medical history will radically change the delivery of healthcare (on a long enough timeline — the only one worth working on).
The start of the health record tracks closely to the start of the institutional genesis of the hospital. Case histories predate this significantly — by thousands of years. But structured health record keeping pretty much tracks with the existence of the “structured” hospital. The earliest examples in America are from 1808 when institutional hospitals were charitable poorhouses:
Development of the clinical record in America occurred first in major teaching hospitals … the hospital was a charitable institution for the poor staffed by volunteer physicians because little of benefit was available there that the patient with means could not obtain at home. In 1808, the New York Hospital began copying selected case reports from physician notebooks into bound medical and surgical volumes for preservation in the library.1
Similar developments played out elsewhere - the development of structured health records in the UK tracks the development of NHS.
There are ways in which I think this twin “institutionalization” continues in terms of new kinds of access control associated with COVID crisis management thinking. I think we should avoid arguments for inevitability and turn to arguments for opportunity though. Inevitability is a kind of a laziness, but also traps us in local optima. We show up for a future where individual rights of access and control of health data matter too.
One practical way of making those things “matter” is by creating and experimenting at the intersection of startups, developer tools, and decentralization/web 3. That’s what we do every day at Automate Medical.
Some obvious opportunities from our perspective:
Decentralized Health IDs + Auth0/Plaid for Health. OAuth for patient data is just getting started. New W3C standards like DIDs create new possibilities for direct patient ownership of, and permissioned access to, health data.
Health Records for Web 3. DHIDs + FHIR + JOSE could lead to completely user controlled medical histories. “Healthchains”. An integrated, “self-custodied” case history of your body.
IPFS for healthcare data. A CDN for FHIR data sets. Can we use IPFS Cluster to establish public repositories of FHIR Bundles for data where transparent, public disclosure is a feature? Can we distribute “algorithmic” medical data like CQL in unpermissioned ways? Even just this morning I saw I found Xirva (arxiv on IPFS).
Sounds interesting? There’s an offer for you.
The Offer
Maybe this all seems kind of whacky. It probably should. At least a little. As I like to remind people, every actually good idea isn’t going to sound particularly believable. Part of it being an actually good idea is that it’s uncommonly believed. Most of our sense of believability is based on assumption of stability — not where things could change.
Andrew, Thomas, and I launched Automate Medical to work on healthcare problems. We build dev tools that make launching in digital health easy. We’re looking for people who want to show up to the future and we’re hiring an especially important role: dev #1. Come join us.
Company formed at start of the year by ex-YC founder group of 3, raised $2M at the start of 2021 — cumulatively, we’ve raised nearly $20M in our careers so far
You will be dev #1 and help us launch Plaid for health (we can show you a working demo)
You will alter the trajectory of our R&D work with rapid prototype development and asynchronous (remote) work — we will pay you well
We are looking for someone who is:
Authentic and opinionated — with a story to tell
Thinking about the future a lot these days
An ex-founder or a future founder
Deeply technical (software++)
Bonus: if you know what FHIR, HL7 V2, X12 are, tell us when you apply.
Our current stack includes:
We don’t expect this particular set of tools to be the only things we ever use. You’re probably a polyglot anyways. You can see some of our open source work publicly on GitHub and at docs.sero.run.
Apply here or by emailing jobs@automatemedical.com
Don’t forget to Subscribe!
Gillum, R. F. (2013). From Papyrus to the Electronic Tablet: A Brief History of the Clinical Medical Record with Lessons for the Digital Age. The American Journal of Medicine, 126(10), 853–857. doi:10.1016/j.amjmed.2013.03.024