Strong opinions, loosely held

I come back to this topic often, only because it's so critical.

There is a particular kind of meeting that looks like rigor and is actually decay. Everyone is well prepared. The slides are clean. Every position is qualified, every counterargument acknowledged, every decision deferred until more data arrives. The room exhales after ninety minutes of careful nothing. I have sat in versions of this meeting in hospitals, academic seminars, and in the rooms where startups die slowly. The disease is the same: nobody is willing to be wrong.

Science, in its honest moments, models the alternative. A hypothesis is not a belief. It is a bet placed under a light bright enough to be embarrassing. The experimentalist proposes; the world disposes. Popper understood that the value of an idea lies not in its plausibility but in its willingness to be killed by evidence. The point of stating something clearly is to let reality refute it.

This is the discipline a founder must import without the laboratory's patience. Startups do not die from bad ideas. They die from ambiguity, from the slow accumulation of half decisions, from the polite agreement that we should keep exploring. A committee can look thoughtful while practicing paralysis. Optionality, treated as strategy rather than as a temporary state, is the refusal to commit dressed in business school clothing.

The right posture is closer to medicine than to philosophy. You see a patient, you form a working diagnosis, you treat. You hold the diagnosis loosely enough that new evidence can dislodge it, and tightly enough to do something this afternoon. The clinician who waits for certainty is not careful. He is negligent. Certainty is a luxury the suffering body does not extend.

Strong opinions become dangerous the moment they fuse with identity. The hypothesis stops being a hypothesis and becomes a piece of the self. Evidence then feels like an attack rather than a correction. You stop listening. You read every data point through the filter of what you have already decided. You punish dissent because dissent feels like betrayal. This is how founders build the wrong product for eighteen months, how investors stay long after the thesis has rotted, how doctors prescribe treatments the literature has quietly buried. The opinion may have been correct on the day it was formed. The refusal to revise is what is fatal.

The opposite failure is as common. A leader with no view forces every debate into politics. Without a hypothesis to test, conversations become negotiations over who carries the most authority in the room. Vagueness creates power vacuums, and power vacuums fill with personality rather than truth. Teams do not want a founder who has memorised every argument and chosen none. They also do not want one who cannot bear to be wrong. They want someone who will look at the evidence, place the bet, and call it off when the bet stops paying.

There is a longer view that helps. Whatever we believe today with conviction, a reader a thousand years from now will find quaint, mistaken in ways we cannot yet see, perhaps morally embarrassing. The history of medicine is a museum of confident error. Bloodletting was not stupidity. It was the best available theory, held by careful men, for a very long time. The lesson is not that conviction is foolish. It is that conviction is provisional by nature, and the people who do real work act with full force inside that provisionality.

To hold an opinion loosely is not to hold it weakly. It is to hold it the way a scientist holds a result he intends to publish and a competitor intends to break. You stand behind it because you have done the work. You let it go when something better arrives, without grief and without performance. I've noticed the best founders and clinicians I know have this rhythm. They commit, they listen, they revise, and they refuse to mistake any of those motions for who they are.