Why I Started Writing Online as a Researcher

For a long time, I told myself I would start writing online once I had something truly worth saying. Once the research was solid. Once the results were clean. Once I actually knew what I was talking about.

That day never seemed to arrive.

There's always another paper to read, another experiment to run, another corner case to handle. Research is fundamentally an exercise in sitting with uncertainty, and somehow I had convinced myself that writing publicly required the opposite — clarity, confidence, and conclusions neatly tied up with a bow.

I was wrong.

The Researcher's Curse

There's a particular kind of paralysis that comes with working in a technical field. The more you learn, the more you understand how much you don't know. You become hyperaware of the gaps in your own understanding, the limitations of your methodology, the places where a reviewer might push back.

This is good for doing rigorous research. It is terrible for sharing anything with the world.

The result is a lot of very smart people who keep their most interesting thoughts locked inside Notion documents, private Slack channels, or half-finished drafts that never see the light of day.

What Changed My Mind

I started paying attention to the researchers and engineers I learned the most from. Almost none of them waited until they had perfect, polished things to say. They wrote about what they were working on right now — the confusing parts, the things that didn't work, the questions they hadn't answered yet.

That imperfection was the point. It made the writing feel real. It made it useful.

A polished paper tells you what was discovered. A personal note tells you what it actually felt like to discover it — the wrong turns, the small insights, the moment something finally clicked. Both are valuable, but only one of them tends to end up in journals.

What I Want to Do Here

This site is my attempt to build that kind of record for myself.

I work on AI agents, computer vision, and the messy problem of making autonomous systems that actually behave reliably in the real world. A lot of that work is slow, incremental, and hard to communicate in a way that fits neatly into an abstract.

So I'm going to write about it here — not in the form of finished research, but in the form of notes. Things I'm reading. Experiments that surprised me. Ideas I'm not sure about yet. Moments where a concept finally made sense.

Some of it will be technical. Some of it will be about the experience of doing research — what it's like to work at the boundary between academia and engineering, to care about both the theory and the thing that actually runs in production.

A Note on Imperfection

These posts will not always be polished. Some of them will be short. Some will raise questions I can't answer. Some might turn out to be wrong.

I've decided that's fine.

The alternative — waiting until everything is perfect — is just a more comfortable way of never writing at all.

If something here is useful to you, or if you disagree with something and want to talk about it, I'd genuinely like to hear from you. My email is always open.