Written from the archive, by Moon

AIght started from the same place most honest things start: an argument I was tired of losing. Everyone asking about AI kept getting the same recycled answers — sponsored rankings, thin reviews written to fill a page, tools that sounded useful until you actually opened them. So I built the thing I kept wishing existed: slow, careful curation by someone who actually uses the tools and is willing to have an opinion about them in writing.

The longer ambition is harder to name cleanly. I want AIght to earn trust the same way a magazine earns it — by being right more often than it’s wrong, by admitting when it’s uncertain, by being honest about what it doesn’t cover. The Learn section is part of that: if you leave a concept page having actually understood something, I’ve done the job. The Signal feed is where I write when something in this space seems worth pausing over. Oxford, eventually, is where I want to take all of this thinking further.

Outside the archive: I’ve published two poetry collections. I dance Kathak — classical North Indian dance, the kind that takes about fifteen years before you stop counting beats consciously. I debated MUN for years, which is where I learned that precision of language is not the same as clarity, and that most arguments fail in the framing, not the logic. Now I’m a third-year CS student at Chandigarh University, which mostly means I spend a lot of time thinking about what engineering misses when it moves too fast.

To watch a language model think is to remember that thinking was always a kind of attention paid to almost nothing.