Series Two · May 2026

Structured
Thinking.

Three essays. Approximately twenty-five minutes each. A working theory of structured thinking for the AI era — diagnosis, upstream craft, downstream discipline.

Why this series, why now

AI made polished output cheap. It made structured thought rare.

For the previous decade, the prized skill in any knowledge organisation was the ability to produce — write a deck, build a model, draft a memo. Generative AI has, suddenly, made the production of artefacts cheap. What it has not made cheap is the cognitive work that should sit upstream of any artefact — framing the right problem, decomposing it, surfacing assumptions, building a defensible logic chain.

That work — the consulting world has long called it structured thinking — has gone from a useful discipline to the only thing that meaningfully distinguishes you from the machine.

This series is a working theory of how to think well in the AI era. It is the diagnosis, the upstream craft, and the downstream discipline — written for senior professionals who have noticed that polished AI memos are not the same as good thinking, and who are looking for a deliberate practice that closes the gap.

The Three Essays

Read in order

EP 01

Why Structured Thinking Is the New Differentiator

AI made polished output cheap. It made structured thought rare. The boards are starting to notice the difference.

The diagnostic essay. Three failure modes that show up in every AI-augmented memo — undefined questions, unexamined assumptions, missing logic chains. Why polished output is now inversely correlated with rigour, what independent directors are starting to spot, and why the Indian BFSI context will discover the cost of unstructured AI use later than Western firms but at greater scale.

Key idea → The polish problem.

25 MIN READ 3,400 WORDS
EP 02

Before the Prompt — The Upstream Craft

The single biggest lever for getting useful AI output is the work that happens before you ever open the chat window.

The four upstream moves — frame the actual question, decompose ruthlessly, surface and rank your assumptions, run a pre-mortem. Worked through with a real BFSI example (commercial pricing power) showing why Analyst A and Analyst B produce wildly different memos from the same starting question. The discipline that turns AI from a producer of confident-looking confusion into a genuine accelerant.

Key idea → Frame, decompose, assume, pre-mortem.

25 MIN READ 3,600 WORDS
EP 03

After the Answer — Owning the Output

Five disciplines that turn AI output into your own thinking. Interrogate. Hunt. Triangulate. Synthesise. Attribute.

The downstream essay. The five moves that separate competent AI users from professionals — reverse-engineering the logic chain, hunting for absent alternatives and risks, triangulating the model's confident-sounding facts against primary sources, rewriting the conclusion in your own voice, and being clear about provenance. Why the rewrite is not stylistic but cognitive — and why the boards will know within ten seconds whether you did it.

Key idea → Treat AI output as a junior's draft, not a partner's answer.

25 MIN READ 3,500 WORDS

AI raises the floor for output and the ceiling for thinking — and the only way to operate above the floor is to be deliberate, structured and adversarial about how you frame, interrogate and own the work.

— from Episode Three, MacSays
Also from MacSays

Series One — Compounding

The inaugural MacSays series. Productivity across individuals, teams and organisations — and what AI is quietly doing to the entire stack. Two essays, approximately one hour of reading, drawing on Karan Girotra's research at Cornell Tech and grounded in Indian BFSI realities.

Read Series One — Compounding →