Knowledge is incomplete until you act

I have come back often to a simple claim: the purpose of knowledge is action, not the accumulation of knowledge for its own sake.

When we seek knowledge, we should also know what knowledge is. For me it begins with knowing that we know, and knowing that we do not know. There is a useful public distinction between known knowns, known unknowns, and unknown unknowns. Whatever one thinks of the politics around the moment that popularised it, the structure is honest. Our rush to complete our picture of a situation often creates a false sense of understanding. I have found this in myself repeatedly: things I thought I understood had large gaps once I looked more carefully, or once I tried to do something with them.

That leads to a second point that matters more in practice. Knowledge stays incomplete until we take action. Before we act, we only have a partial grasp of the context. Forums and social media are full of people who discuss problems as if they already understand both the problem and the solution, and who offer direction to others with confidence. I have done the same.

The constraint I want to keep on myself is different: I should not treat my understanding of a problem and its solution as clear until I have actually tried to implement something. Implementation is where the unknown unknowns appear. Until then, humility is not a moral fashion. It is accuracy.

I remain open to refining even this view. There may be domains where pure study is the right next step. For the work I care about, talk without action is a soft substitute for understanding.

Drawn from writing on X as devendrasm, including a Manthan note on knowledge and action (Dec 2024).