The System Anti-Gambit When paradigms break, the biggest threat to an investor or operator is their own most sensible instinct. How to tell a real structural move from a panic reflex? When India liberalised in 1991, it made the sensible bet. Decades of Nehruvian investment had left the country with graduates, so India sold their brainpower to the world through IT and knowledge services: hire a young graduate for a couple of thousand dollars a year, give him a month of standardised training on a campus in Mysore or Chennai, bill his time to a Fortune 500 company, and keep a twenty to twenty-five per cent operating margin. It worked for three decades and earned three or four hundred billion dollars of foreign exchange. It was a system gambit, and it paid off for a generation. AI is close to purpose-built to take it apart. When Saurabh Mukherjea came on my podcast to discuss his book Breakpoint, this was the case he kept returning to. The Fortune 500 client no longer cares how many people you deploy. It names a fixed price for a CRM system and leaves the method to you, and the method that wins now is two senior people and AI rather than a floor of graduates billed by the hour. Smaller Indian IT firms have already started to shut down. The move that built the industry is now the move dismantling it. In The System Gambit I call the winning move a system gambit and its shadow the system anti-gambit. The system anti-gambit is what you do when the ground shifts and you answer by doing more of what used to work. It is the most rational-feeling response available, and that is what makes it dangerous. Subscribe for free to receive new posts The party most exposed to the system anti-gambit is the one currently winning. Sit at the top of a paradigm and your entire apparatus, your training, your hiring, your metrics, your instinct for what good looks like, is tuned to that paradigm, so when it shifts your immune response is to intensify the very thing that put you on top. Germany won the World Cup in 2014 by doing the classical things well. Find the children early. Drill them. Make them fit and disciplined. Then Germany kept doing exactly that while France, Spain and Morocco moved to fluid roles, fluid formations, and, in Morocco’s case, a platform model that pulled world-class players from its diaspora toward one clear mission. Watching Germany play France today, Saurabh said, is like watching teams from different planets. The 2014 title was the setup for the decline that followed. India produces eight to nine million graduates a year, close to the population of Germany every decade. Graduate unemployment runs between thirty and forty per cent, while the rate for non-graduates is around three. The degree now predicts unemployment. The old script, study hard, earn the degree, work hard, land the job, arrive at a middle-class life, has inverted. And the sensible-sounding instruction, send more young people to university, is the system anti-gambit at national scale. Arguing against university feels wrong. That feeling is the trap. AI sharpens all of this through one property. The transforming technologies of the past forced a firm to change how it operated before it could use them at all, and that friction is exactly what made them worth capturing. AI runs the other way. It bolts onto whatever you already do, which means the gains from bolting it on are available to anyone with a credit card and an API key, and anything available to everyone earns nothing that lasts. This is why the most common reaction I meet, the senior executive who asks me how to learn prompting, is the system anti-gambit in a new form. It carries an industrial-era instinct, learn a defined new skill, tick the box, feel safe, into a situation where the field itself has changed. Sangeet Paul Choudary makes the point in Reshuffle: the game has changed underneath you, and learning to dribble faster is no help once the court has become a rink. If the standardised skill is losing its value, where the value flows next depends on what a system actually does. In the book I separate four modes of controlling any system, and they stack into a ladder. The lowest is regulation, the thermostat that holds a variable inside a band. Above it is coordination, the traffic light that keeps the flows from colliding without telling anyone where to go. Above that is orchestration, the conductor who plays no instrument yet makes a written score sound like music. At the top is governance, the captain of a ship who, with a storm ahead and no way to know the outcome, decides whether to stop, turn, reroute, or sail straight through. AI is already good at the thermostat, and it is climbing into the traffic light and the scored parts of the orchestra. What it cannot do is the captain’s judgement under genuine uncertainty. So value drains out of the lower rungs and pools at governance. That ladder explains the value migrations Saurabh tracks as an investor. Index funds turned asset management, the discretionary stock-picking that Saurabh and I both do for a living, into something close to a regulated utility, and the market answered by paying far higher multiples for wealth managers, whose real job is the ungovernable part, sitting with a frightened client and talking them out of selling at the bottom. The same logic is draining broadcast and print into YouTube and OTT, where what survives is the customised and the emotionally specific rather than the templated segment. Anything that is cut, paste, and repeat is being repriced towards zero, while anything that turns on reading a particular person in a particular moment is gaining. Saurabh reframes it as three fluidities: whether a person’s skills are fluid or fixed, whether their role is fluid or fixed, and whether the organisation around them is rigid or fluid. The industrial revolution, from Adam Smith’s pin factory to Ford’s line at River Rouge, made all three as static as it could, deliberately. The coming years will hurt mainly because most firms are still built that way while the world has started paying for the opposite. None of this tells you which specific move to make, and it should not, because there is no formula that guarantees the result. But you can test a candidate move before the market passes its verdict. There are three checks. The first check is whether the move builds a self-improving loop. If each turn of the wheel makes the next turn better, you own something that compounds. If it does not, you have swapped an old asset for a newer one, and a newer asset is not a position. The second check is path dependence. A real structural position cannot be reached by a richer competitor writing a cheque, because that competitor would have to run the loop from the start rather than buy their way to your thousandth iteration. This is what makes Bajaj Finance so hard to copy. Bajaj has collected granular data on Indian borrowers at the point of sale, down to whether a doctor is a cardiologist or a dentist, trained at AIIMS or an ordinary college, practising in an elite Mumbai suburb or the boondocks, and that data lets it price risk for doctors better than anyone and hold the largest medical loan book in the country. If Saurabh and I raised money tomorrow to rebuild that dataset, we could not, because the signal was gathered over years at the core bottleneck by an agent standing at the checkout. That is a moat. A skill anyone can buy with an API key never becomes one. The third check is the one people skip, and the one I rate highest. I call it management-logic antagonism. It means doing the new thing in a way that runs against the logic of the old system, so the incumbent cannot copy you without dismantling itself. Ukraine is the best illustration. On paper, a country attacked by a neighbour with ten times its economy and the largest nuclear arsenal in the world should fold in days. Ukraine survived by refusing to fight the industrial war Russia had prepared for, the war of feeding more recruits into the line, and inverting it. Cheap drones improved through fast learning loops. A cost asymmetry in which a few hundred dollars destroys a tank. Strikes aimed at the binding constraint, the oil refineries deep inside Russia that fund the state, rather than at the trench. For Russia to answer in kind, it would have to abandon a doctrine it has run on since Stalingrad, and institutions do not abandon the thing that once made them win. That reluctance is the antagonism, and the antagonism is the moat. This essay is based on my new book The System Gambit: For an investor, the three checks are how you improve the odds without pretending to certainty. You are hunting for operators who have rethought their own system gambit again and again as the ground moved, and who keep their skills, their roles, and their structures fluid on purpose. Bajaj Finance rotates its people through departments so that nobody’s thinking calcifies, and that habit tells you more about the next decade than any single product. The question Saurabh frames is worth taking into every strategy meeting: is this the France of corporate life, reinventing after every cycle, or the Germany, superb at a game the world has stopped playing? The discipline is unglamorous. See the system for what it is rather than what the last decade trained you to see. Notice the reflex to double down before you act on it. Think two or three levels below the surface story. Durable value compounds in the places money cannot buy its way into and judgement cannot be standardised out of. And the one thing standing between an operator and those places is, almost always, their own most sensible instinct. Paperback: https://www.amazon.com/System-Gambit-Leverage-Unlock-Compounding/dp/3982897211/ Kindle: https://www.amazon.com/System-Gambit-Leverage-Unlock-Compounding-ebook/dp/B0GY8J23SK/ Hardcover: https://www.amazon.com/System-Gambit-Leverage-Unlock-Compounding/dp/398289722X Sub