Over the past week, I watched a stock I have never owned climb on the back of a narrative I have spent a decade trying to decode: artificial intelligence is coming for medicine, and the market has decided that Doximity—the physician network long dismissed as "LinkedIn for doctors"—is one of its purest bets.
I do not trade healthcare equities. But the pattern stopped me cold. Here was a company whose own analysts concede it is not a traditional medical product company, being re-priced overnight as if it had invented the stethoscope. The same happened to tokens in DeFi Summer, and to JPEGs in 2021. The market does not buy products. It buys stories about networks. And networks, as I learned through five years of DAO governance work, are exactly where the soul gets lost.
For context: Doximity operates the largest professional medical network in the United States, connecting the vast majority of physicians through telehealth, faxing, and collaboration tools. It grew quietly, protected by professional inertia and regulatory moats. Then the generative AI wave arrived, and investors began hunting for established platforms that could bolt on AI features without rebuilding distribution. Doximity fit: installed clinicians, structured workflow data, and a model that needs no software evangelism—just doctors staying logged in.
The source material I am working from is strikingly thin: a disclaimer admitting the original research contained just a title, an abstract, and a source attribution, with conclusions inferred from public industry knowledge. I found this oddly refreshing. It is an honest confession that much of what we call research in both healthcare technology and crypto is narrative construction over a thimble of data. That does not make the phenomenon less real. It makes it more dangerous.
Here is what the sparse material actually tells us, stripped of hype: Doximity is not a medical product company. It is a network company. That distinction is the key to understanding both its AI upside and its governance fragility—and it maps almost perfectly onto how I have learned to evaluate blockchain protocols.
A product company owns its value creation. A network company merely coordinates participants who create value for one another. Physicians join because other physicians are there; advertisers pay because physicians are captive; AI potential rests on the structured interactions already flowing across the rails. Doximity is a curator of attention and trust, not a manufacturer of therapies.
This is why I abandoned product-based evaluation frameworks in crypto. When I audited MakerDAO governance proposals in 2020, I kept reaching for traditional metrics—total value locked, revenue, utilization. I kept missing the point. The protocol was not a bank making loans; it was a coordination layer for people who disagree about money. I once framed this as economic empathy: the quiet architecture letting divergent interests coexist without violence. Its real asset was not its smart contracts but its legitimacy. The same is true of Doximity, and the same is true of every network that survives a hype cycle.
The governance gap. In my experience drafting governance structures for DAOs, I have learned that networks fail not when they lose users but when they lose the ability to arbitrate values. Doximity's AI push raises a question its shareholders are not asking: who decides what happens to the patient data flowing through its AI features? The company has commercial incentives to monetize that data. Physicians have professional incentives to protect patient autonomy. When those incentives collide, someone must govern the collision.
Crypto believed code would solve this. It did not. The Tornado Cash sanctions taught me that writing code can be treated as a crime, turning every smart contract into a potential liability. But the deeper lesson is symmetrical: a network's rules are only as durable as its community's willingness to renegotiate them. Doximity's physicians never voted on the AI roadmap. Neither did most token holders in the protocols I studied. Governance is the missing layer—in medical networks, in social networks, and on-chain. I wrote once about the quiet collapse of equity in code; this is what I meant.
The curation lens. What fascinates me about the source material is what it omits. There is no mention of clinical outcomes, no discussion of whether these AI features actually improve patient care. The entire assessment is positional: where does Doximity sit, and what can be inferred? This is how we spoke about NFTs in 2021, treating provenance as a feature rather than a commitment. I spent three months manually verifying artistic intent for a small archive DAO that year, and I learned that the difference between a durable digital artifact and a derivative clone is exactly this: someone must curate meaning. The market rarely does. I see the same omission in Doximity's AI coverage.
The contrarian angle: the medical AI boom may be less about intelligence than about entrenching centralized data monopolies as patients begin demanding data sovereignty. Doximity's rally is a warning to crypto builders, not a validation of network models.
I have spent months translating legal jargon into philosophical commitments for municipal data projects. The hardest opponents are not regulators. They are incumbents who offer convenience in exchange for autonomy. Doximity's AI feels like convenience—until the data exhaust becomes a liability. The insight I keep circling: the most profitable networks are not the best governed; they are the most conveniently captured. If blockchain cannot offer an alternative model for medical and professional data governance, then its promise was always rhetorical. Decentralization is, finally, an emotional security question, not just a technical one. The playbook is the same: convenience first, governance later, and by the time users ask who holds the keys, the answer is an acquirer.
I do not know whether Doximity's AI features will work clinically. I know the question the market is not asking: who holds the keys to the network's soul? Curating the soul in a world of derivative clones means asking governance questions before they become existential. The protocol that builds legitimacy through transparent, empathetic data governance will outlast the hype. The one that merely rides the AI narrative will be replaced by the next story. The network is the product. Governance is the soul.