Nicole Junkerman
Notes
Nicole Junkerman on India's AI Story, From a Distance
Reading about a country's technology from several thousand miles away has obvious limits and one small advantage: the headlines fade and the ordinary parts stay.
Reading about a country's technology from several thousand miles away has obvious limits and one small advantage: the headlines fade and the ordinary parts stay.
Reading from a long way off
Most of what reaches a reader in Europe about artificial intelligence in India arrives as a number. The numbers are large and, on their own, not very informative. What is more interesting, and much harder to get at from this distance, is the texture of it: who is doing the work, what they are doing it for, and which problems are considered worth a year of somebody's life.
Nicole Junkerman reads about it the way she reads about anywhere she has not lived, which is slowly and with a low tolerance for confident summaries. A country of that size does not have one story about anything, and the sensible thing to do with a distant subject is to look for the shape rather than the verdict.
Scale as a design constraint
The thing that seems to shape the work most is scale, though not in the way the word is usually meant. Here it reads as a constraint rather than a boast. Anything intended to be useful across the country has to survive dozens of languages and several scripts, an enormous range of handsets, connections that come and go, and people who have no interest whatsoever in the tool itself and every interest in the thing it is meant to help with.
Constraints of that kind tend to produce plainer work. Something that has to run on a modest phone, in a language its designer may not speak, for somebody who will give it one attempt before giving up, cannot afford to be elaborate. It has to be obvious. That is a harder brief than it sounds and a better one than most.
The craft rather than the headline
The part that is genuinely admirable, and the part that travels least well into a headline, is the sheer number of people doing careful technical work: engineers, teachers, students, and a great many universities turning out graduates who go straight to the practical end of it. A few things stand out even from here.
- The problems chosen are often mundane, which is usually a sign of seriousness rather than a shortage of ambition.
- Language is treated as central rather than as something to be added later, because it has to be.
- Teaching and making happen in the same rooms, and often in the same week.
- The work is spread widely rather than gathered into a single city.
What travels back
The reason any of this is worth a reader's evening has nothing to do with comparison. Ranking countries against one another tells you very little about whether a piece of software is any good, and it is a poor use of an hour. The useful observation runs the other way. Something designed for a modest phone, a second language and a person in a hurry will generally turn out better for everybody than something designed for ideal conditions. Constraints travel.
That is the note Nicole Junkerman came away with, and it is a small one, which is appropriate for a view from this far off. Reading about a place is not the same as knowing it, and the city notes in this journal are a standing reminder of how much a single week on the ground changes the reading. What distance does allow is a look at the shape of the thing without the noise, and the shape looks like patience, craft, and a very large number of people quietly getting on with it.
Explore more of the journal: AI in Everyday Life and Reading Room, or return to the Nicole Junkerman home page.