Learning AI

Learning about AI today means two things at once, and it is worth keeping them apart: on one hand, understanding how these systems work —what a gradient is, what a transformer does, what it means to train a model—; on the other, understanding what is at stake when they are deployed, who decides, with whose data, over which territories, and at what cost.

The chapter covers both, in four stretches. It opens with the foundations —what the word “AI” names, where the idea comes from, why it worked now, what kind of knowledge it produces—, because these are the questions you answer anyway, whether or not you ask them. Then the map of free, annotated resources for getting started, then its extension for anyone who wants to go further, and finally the critical counterpart, read from Latin America and the Global South.