
Prof. Piyush Sinha
Faculty, IIM Ahmedabad
“AI is my servant, not my master. The only thing we have to learn is how to swim with sharks, how to sleep with the enemy.”
Session recap
In a Day-2 morning workshop, "The Art and Science of Case Teaching", Prof. Piyush Sinha did something most workshops on this subject avoid — he taught a case live before debriefing on the teaching technique. Introduced by Dr. Sushma Vishnani and felicitated by Sri Shreevats Jaipuria and Dr. Prabhat Pankaj, Prof. Sinha (currently at MICA, soon to be Dean & Principal; formerly IIM Ahmedabad and IIM Calcutta) used the Planet Health case as his canvas and the Domain of Retailing note as scaffolding for ninety minutes of demonstration, followed by thirty minutes of methodology. The opening was deliberately provocative. AI, he said, has stripped faculty of every traditional source of power — content, concepts, examples, analysis, PPTs. "What's left for us as teachers? Power distance is finished." His answer: the only place faculty still hold power is inside the classroom, and the only skill AI cannot replicate is probing. The session became a live demonstration of that probing — calling on faculty one after another, refusing every "fake CP" answer, pushing each respondent through five rounds of why and how until the underlying assumption — not the conclusion — became visible. The case itself was a vehicle. The real curriculum was the meta — how to choose between inductive and deductive sequencing, when to use a case vs. a case study, how to pitch the class at the top 30%, why "confusing the class first" is a feature not a bug ("only when you churn the milk does first curd, then poison, then nectar come out — the whole class discussion is a Manthan process"), and how to time-box a case discussion when the case itself can be taught from marketing, finance, HR, distribution, macroeconomic, or legal angles. He closed with two practical gifts: a complete strategic framework for retail businesses (markets → opportunity → shopper → value proposition → merchandise → pricing → location → format → category → positioning) and a single sentence to beat AI in the classroom — "5Y and 1H. AI will get three of the five whys wrong. That's where we still have power." Two frameworks he left with faculty are worth keeping at hand. First, a Pedagogy Matrix of Teachability × Learnability: high/high → inductive case with a good debrief; low teachability + high learnability → pure deductive case method; high teachability + low learnability → lecture; low/low → avoid. Inductive (concept first, case as illustration) is easy to teach but produces shallower learning; deductive (case first, concept emerges) is harder to teach but goes deeper. Choose by concept type, audience maturity and time available. Second, three modes of decision-making rooted in biology — database/linear (mind), intuitive (heart), and gut (literal stomach, "butterflies"). Teach students to recognise which mode they are in, and to use data as scaffolding for the creative leap, not as a substitute for taking it.
Editorial summary compiled by the FDP team — not a verbatim transcript. Spotted an inaccuracy? Let us know.
Prof. Piyush Sinha has been honoured with a tree at Trees for Tigers®, Sundarbans National Park, West Bengal, planted by Jaipuria Institute of Management as part of the 14th Annual FDP, 2026. Certificate No. 5533756, dated 10 May 2026. The plantation is geotagged and trackable via Grow-Trees.com — "these trees will provide flowers, fruits, fodder and fuel to living creatures and improve water catchment areas."
A tree planted in his name
A Bountiful Tree at Trees for Tigers, Sundarbans National Park
As a token of gratitude for Prof. Piyush Sinha’s presence at the 14th Annual Faculty Development Programme, Jaipuria Institute of Management has planted a tree in his honour. This tree will provide flowers, fruits, fodder and fuel to living creatures and help improve water catchment areas in the Sundarbans tiger habitat.
- Planted on
- 10 May 2026
- Certificate №
- 5533756
Key takeaways for faculty
- 1
Pitch the class at the top 30%
Pitching to the middle bores the top, who'll spread word that the class isn't challenging. Pitching to the bottom makes you a social reformer. The top pulls the middle up; the bottom learns by survival pressure.
- 2
You're not a teacher — you're a performer in a theatre
Teaching is a performing art. Once a student has read the case and AI has given them a solution, your only remaining job is to confuse them, then probe them, then bring them to a concept they couldn't have arrived at alone.
- 3
Confuse first. Clarify last.
Class discussion is the Manthan — churn the milk and first comes curd, then poison, then nectar. The unstructured-looking middle of a case discussion is doing work; resist the urge to bring structure too early.
- 4
Decide your concept before you walk in
A case can be taught from marketing, finance, HR, distribution, legal, or macroeconomic angles. Pick three concepts you want to land. Anything else is improvisation, not pedagogy.
- 5
Case ≠ case study
A case (no solution) is for forcing decision-making. A case study (with solution) is for illustrating a concept. Use them differently. Most faculty conflate them and lose both edges.
- 6
5Y and 1H is your AI-proofing tool
AI will get the what right. It will often get the how right. But in three out of five whys it doesn't know — because it has no lived experience. Probe the why and the how. That's where faculty judgment still outranks the engine.
- 7
Don't try to finish the case — exit it
A case cannot be finished, only exited. Time-box the discussion (e.g., 20 min open, 20 min probe, 15 min concept, 5 min summary) and exit when you've landed the three concepts you came to land.
Speaking at One Jaipuria FDP
- Live case demonstration — Planet Health taught front-to-back, in real time, with the audience as the class
- Power distance in the AI-age classroom — what faculty have lost, and the one skill (probing) AI cannot replicate
- Inductive vs. deductive sequencing — when to put the concept first and when to let it emerge from the case
- Case vs. case study — forcing decisions vs. illustrating concepts, and why conflating them dulls both
- Pitching to the top 30% and the "confuse first, clarify last" rhythm of a Manthan-style discussion
- AI-proofing a case discussion with the 5Y and 1H probe
- Strategic framework for retail businesses: markets → opportunity → shopper → value proposition → merchandise → pricing → location → format → category → positioning; the three retail levers (space · manpower · merchandise); and the "switching business" framing for store-loyalty
Q&A captured
Q. When we teach a case, should we force the manager to make a decision? Many startups would have done better if they hadn't made certain decisions.
Data is used for building scaffolding — but how big a leap you take from that scaffolding is up to you. The decision is emotional in the end. But every business question must answer: where will the money come from? Show me the revenue model. Even NGOs are now being asked to show ROI.
Q. If we continuously probe one student, won't others stop participating for fear of being probed?
Your point is valid. But fear is also pedagogy — if there is no risk, no one is listening. The trick is that the whole class is watching a drama unfold — and they're learning from someone else's discomfort as much as their own. The probing is the lesson, not the punishment.
Q. How would your method apply to a finance case — say, valuing the shares of a company?
The question becomes — how much money do you want to raise? At what price? On what multiple? Even a finance case is a decision in a context, not a calculation. The DCF is scaffolding. The leap is what stake you give and at what valuation. Same craft.
Q. Is the assumption of ₹1,000 per sq.ft. per day in the Planet Health case realistic?
Realism is irrelevant. It's a personal threshold the protagonist has set — like a faculty member who refuses to teach if feedback is below 3.5. It's not a market truth; it's a business choice. Teach students to spot that — those are the levers that actually drive decisions.

