There’s a difference between using AI and being used by it.

First real research project. Aarav started thinking the answer was “kids are picky.” Eight interviews in, the actual pattern showed up: temperature, not taste. He learned the hard part of interviewing is not asking, it is shutting up long enough for the second answer to arrive. The notebook is the artifact. The synthesis is the proof.

Aarav sat at the same junction four afternoons in a row with a pen and a stopwatch. He counted near-misses, measured kerb heights, timed signal cycles. The proposal is not a fantasy redesign. It is a kerb extension and a fourteen-second pedestrian phase, defended with his own count data. He learned that the city is a thing you can study, not a thing that happens to you.

Looks like a kid game. Reads like strategy. Aarav noticed his brother held +2 cards back instead of playing them early and started asking why that worked. He sketched the decision tree, tested three rule changes across ten rounds with his cousins, and found one that actually shifted the win rate. The deliverable is not the new rules. It is the habit of asking what makes a system tilt.

The first poster failed. Aarav had drawn a sorting chart with seven categories. The watchman told him nobody read past the first row. The third poster had eight words and a photo. Compliance tripled. Real marketing is not what you want to say. It is what the person walking past at 7am will actually register. Aarav now reads every ad on the back of an auto with that filter.

This is not a dashboard. Aarav specifically refused to build a dashboard. The tool answers one question for one neighborhood at one time of day, which is the only question he actually wanted answered. He learned that good engineering is choosing what to leave out. He also learned, the hard way, that the IMD radar feed updates on a thirty-minute lag and his first version was confidently wrong for two weeks.

First draft was a fan letter. Second draft was a fan letter with statistics. The third draft has a paragraph titled “the case against my own argument,” and that paragraph is the reason the essay is now good. Aarav learned that you do not win an argument by ignoring the other side. You win it by stating the other side better than the people on it can. He still thinks Virat is back. He can now defend it.

Two months. Four cafés. Sixteen recorded interviews. Aarav learned that Kayani, Britannia, Yazdani, and Sassanian are not nostalgic backdrops. They are running businesses with rent problems and grandchildren who do not want the trade. The guide takes a position. It argues these places are infrastructure, not heritage, and that the difference matters. It is the kind of work an adult would shake his hand for.