I started this blog because most writing about AI products either gets too technical too fast, or stays at the surface — "the model is bad" / "the model is good." The more interesting questions are about the products themselves: what problem does it actually solve, who is it really for, why this shape and not another.
I write roughly weekly. Some essays are reviews of a single product or feature. Some are roundups of the patterns I keep seeing across the space. Occasionally I publish a longer piece that tries to make sense of a whole category. I try to be specific, generous, and willing to be wrong on the record.
By day, I'm a PM working on AI products at a startup called Virtualitics. By weekend, I'm reading listening to Lenny's Podcast, walking around Golden Gate Park, and wondering why every chat box looks the same.