Conviction Without Adaptation Is Just Stubbornness

When I first became a founder, I thought conviction was one of the most important traits I could develop.

Startups reward people who are willing to believe something before everyone else does. You have to recruit people before the company feels inevitable, sell customers before the product feels complete, and keep moving when there is much more evidence for why something might fail than why it might work.

So I learned to value conviction.

Then I learned its failure mode.

At Underdog, we built infrastructure for digital assets on Solana. We found real customers, generated revenue, won a major ecosystem hackathon, and had enough traction to believe we were onto something.

Then the market changed.

Crypto entered a much colder period. Customer priorities shifted. Technologies that had seemed urgent became less urgent. New standards changed what developers needed. Things that had worked stopped working as well.

My instinct was initially to push harder.

That instinct is useful sometimes. A difficult market can make a good idea look temporarily bad, and founders who change direction every time something gets uncomfortable never build anything durable.

But there is another possibility: reality is giving you new information.

The hard part is knowing which situation you're in.

I don't think there is a framework that resolves that perfectly. What changed for me was learning to separate conviction in the mission from conviction in the current implementation.

I still believed software should make complicated financial and technical systems easier for people to use.

I became much less attached to the idea that our current product was the final expression of that mission.

That distinction gave us room to experiment.

We tried new products. Some went nowhere. We paid attention to what users were doing instead of what we wished they were doing. We watched changes in AI closely and started asking whether natural-language interfaces could remove even more complexity than the APIs we had spent years building.

Eventually that experimentation became Griffain.

The company changed from developer infrastructure into an agentic consumer product. The technology changed. The user changed. The interface changed. Even the way I spent my time changed.

The underlying belief did not: powerful technology becomes more valuable when more people can actually use it.

That period changed my definition of leadership.

I used to think a leader's job was to have the answer and convince everyone else to follow it.

Now I think the job is harder.

A leader has to create enough clarity for a team to move even when the answer is incomplete. They have to distinguish temporary pain from structural change. They have to admit when a previous decision is no longer right without making the team feel that every decision is arbitrary. And they have to absorb uncertainty without pretending uncertainty doesn't exist.

Most importantly, they have to change their mind without losing their direction.

There is a version of conviction that is really ego: I said this would work, therefore I need it to work.

There is a version of adaptation that is really fear: this is hard, therefore we should do something else.

Good leadership lives somewhere between them.

I still value conviction. Building anything new requires it.

But I no longer see changing your mind as evidence that the original conviction was weak.

Sometimes changing your mind is the most serious expression of conviction: caring enough about the outcome to stop defending the path that is no longer getting you there.