Roehrig's claim is that AI is not the next wave in the sequence that built Indian services — offshoring, then digital transformation — but a change of order, because those earlier shifts applied technology inside existing ways of working while this one changes how knowledge work itself gets done. His image is the factory owner with a perfectly good water wheel who is offered electricity: you can see sparks, safety hazards and retraining, or you can see that the factory no longer has to sit next to a moving river. That makes boardroom caution reasonable — a healthcare payer or a big bank cannot roll the dice — but it makes waiting fatal. The biggest mistake a C-suite can make right now, he argues, is to tap out for three or five years; the alternative is bounded experiments on the processes that are already too slow, too expensive and built on technology from the 1980s, run with humans deliberately kept in the loop. The bank platform Ascendion modernised was 40-year-old technology, rebuilt in half the time by people who knew how to use AI. For India the argument lands on the GCC, which he pitches as an actual offering rather than a slogan — 'GCC to the power of AI', stacking AI arbitrage on top of wage and skills arbitrage for roughly half the cost and half the time. And where AI stops is the last mile: it moves weak writing toward the mean, which is fine for most marketing content, but the line that makes a story pop still comes from human hands.
Worth your time if you are
CIOs whose board has quietly decided to wait three years
GCC heads being asked what they are doing about AI
B2B marketers watching an LLM draft their copy
Operators in regulated workflows — payers, providers, banks
Writers arguing about where the machine stops
