Why training trainers is the only way to a million Pakistani learners

Three ways to see it

1. 1

   The National AI Policy 2025 commits Pakistan to one million AI-literate citizens by the end of the decade. The arithmetic of how that happens is not optional. One trainer running cohorts of twenty-five learners, two cohorts a quarter, eight cohorts a year, reaches two hundred learners annually. To hit one million in five years we need ten thousand active trainers. Pakistan today has perhaps two hundred people who could credibly run such a cohort. Every other path, MOOCs, YouTube courses, broadcast television, has been tried in literacy work for forty years. They reach awareness, not competence. Competence needs a human in the room.

2. 2

   Way one to think about the bottleneck: trainers are the rate-limiting reagent. If you have unlimited curriculum, unlimited compute, and unlimited learner demand, but only two hundred trainers, your annual output is forty thousand learners. Curriculum and compute are now cheap, demand is loud. The single chemical missing is the trainer. Training-of-trainers is therefore not a side-stream activity. It is the main stream. Everything else, content, LMS, certification, exists to support the trainer in the room.

3. 3

   Way two: every trainer you certify is a multiplier, not an addition. A new learner adds one to the count. A new trainer multiplies the count by two hundred per year for the rest of their working life. The economics of any literacy mission, whether it was Anand Kumar's Super-30 in Patna or the Adult Functional Literacy Programme in Punjab, hinge on the multiplier. The decision to spend a thousand hours teaching one trainer instead of teaching twenty learners directly is not romantic. It is the only honest answer to a target with a denominator of one million.

The why-tree

Try this with Claude

AI-edge prompt: 'I am a senior trainer at NIM Karachi planning Pakistan's AI literacy roll-out. Given a target of one million learners by 2030, calculate how many trainers I need at three throughput assumptions: 100, 200, and 400 learners per trainer per year. For each assumption, recommend the recruitment funnel size assuming a 30% certification pass rate and a 70% three-year retention rate. Output as a markdown table with footnotes on the assumptions.'