Nicole Junkerman on India's AI Moment

Why the word moment fits

A moment, in the sense meant here, is not a prediction. It is a description of the present tense: a great many people in one place working on the same set of problems at the same time, with enough momentum that the work has started to shape itself. India is having one of those with artificial intelligence, and the interesting parts of it are unglamorous. Nicole Junkerman reads it the way this notebook reads most things, from the craft upward rather than the headline down.

Languages first

The single most distinctive constraint is language. A tool intended to be useful across India has to handle a long list of languages and scripts, spoken and written by people who frequently move between two or three of them inside a single conversation. That is genuinely difficult, and it is difficult in a productive way: work that survives that much variety tends to be more careful and more forgiving than work built for one tidy language. It also quietly benefits everybody else, because the habits it teaches travel.

People, in numbers

The second thing is simply how many people are involved. Engineering and computing are studied there on an enormous scale, in institutions of every size, and a great deal of the learning happens outside them as well. What comes of that is breadth rather than a house style: many approaches attempted at once, a short distance between an idea and somebody prepared to try it, and a large number of people who have read the same problem from different angles.

The test of an ordinary day

The third is the standard by which things get judged. The conditions that matter are the everyday ones: an unreliable connection, a shared handset, a queue with no patience in it, a shopkeeper with thirty seconds to spare. Anything that earns a place under those conditions has been argued with by a very large number of people, and the argument leaves a mark. That kind of usefulness is undramatic and tends to outlast the excitement around it.

• Many languages in one pocket, which forces a tool to be patient about how people actually write.
• Engineers in great numbers, which makes the distance from idea to attempt unusually short.
• A standard set by crowded mornings rather than by quiet rooms.

What a notebook takes from it

None of this needs a forecast attached to it, and none of it is a comparison with anywhere else. It is a reminder that a lot of careful work goes on out of view, and that the conditions a thing is made under leave their fingerprints on it. The travel pages make the same point about places, and the reading pages make it about books. An older note on curiosity and the promise of this subject is the companion piece, and yesterday's note on attention is the other half of the week.