My favorite three words in science? “I don’t know.” Not as defeat — as a trailhead.
I’ve been sick for years. The doctors have names for some of what’s happening in my body, but not all of it. The honest answer to “what causes your disease?” is still we don’t know. And after the fourth specialist, the sixth round of bloodwork, the third time you hear “everything looks normal” when nothing feels normal — you learn something about the space between questions and answers.
That space is where I live now. Professionally, personally, and scientifically.
The Diagnostic Odyssey Is a Data Problem
When a patient presents with symptoms that don’t fit clean diagnostic categories, the system does what systems do: it routes you to the closest match and moves on. But rare diseases barely fit those categories at all. “Rare” means what it says — each one affects a small number of people — and small numbers are the whole problem: fewer specialists who have ever seen your pattern, fewer studies, fewer well-worn paths to a name.
Part of the gap is medicine that doesn’t exist yet — questions no one has answered for anyone. But a maddening share of it is infrastructure. Sometimes the data does exist somewhere: a paper in a journal no one reads, a case study at an institution you’ll never visit, a pattern in a dataset that no one has connected yet. The problem is that these fragments don’t talk to each other. And the patient — the person least equipped to coordinate a research program — becomes the de facto project manager of their own diagnosis.
Not Knowing as a Practice
Feynman said it plainly in a 1981 BBC interview: he could live with doubt and uncertainty and not knowing. I used to think that was about intellectual humility. Now I think it’s about endurance. You have to be willing to hold a question for a long time — years, sometimes — without collapsing it into a convenient answer.
That’s what rare disease patients do every day. We hold questions that the healthcare system has given up on. And the best scientists I’ve worked with do the same thing: they sit with not-knowing long enough for the real pattern to emerge.
Why It Matters for What I Build
Lattice Protocol exists because of this gap. Federated research infrastructure — compute that crosses institutional boundaries, architectures designed so knowledge can move while the data stays protected, models that could learn from far more patient encounters than any one clinic ever sees — is the engineering response I’m building to the diagnostic odyssey.
I don’t know what causes my disease. I may never know. But I can build systems that make the next patient’s path shorter. That’s not optimism — it’s engineering. And it starts with being honest about what we don’t know yet.