After AI, there is no average: Sidu Ponnappa on judgement and work
A provocative organizational argument that AI amplifies the operator's judgement rather than lifting everyone to the same level, shifting the scarce skill from producing work to specifying, checking, and improving how agents do it.
Logged at IST: 2026-10-14 19:16 IST
What it is: Sidu Ponnappa's essay expanding on his Zerodha Subtext conversation about AI-native firms and what he calls the disappearing middle tier of talent.
Gist: Ponnappa argues that AI does not automatically raise the median worker to the former level of the best. Instead, agents amplify the judgement, standards, and operating discipline of whoever directs them: strong operators can produce far more, while weak judgement can generate low-quality work at much greater speed. As agents compress a month of execution into a day, human review and decision-making stamina become constraints; people may keep using agents after their capacity to judge the output has faded.
The essay's proposed shift is from doing the work by instinct to making good work reproducible: define workflows, standards, instructions, and checks so agents can produce reliable results with less review. That also challenges organizations built around a broad "good enough" middle, including hiring, performance reviews, pay, and centrally defined standards. This is Ponnappa's argument and forecast, not an established universal outcome; the essay itself frames it as a view of the situation in October 2026.
Why it matters: It reframes AI adoption as a question of judgement and organizational design, not just tool access or productivity. The practical test is whether teams can make quality explicit and build systems that sustain it without exhausting human reviewers.
Sources read: https://sidu.in/essays/after-ai-there-is-no-average.html