Founding Batch

January 2027 is priced at founding rates. Fees rise 30% from Batch 2

Apply at Founding Rates
Zrooth
Menu
← All Insights

Whitepaper · The Shift

Applied Knowledge Over Degrees: The Case for Learning Every Subject

17 November 2026 · 14 min read

Applied knowledge is the ability to take what you have learnt, in any subject, and use it to solve a real problem. This paper argues that the broad, every-subject schooling many Indians grew up with is a sound foundation for it, that the foundation only pays off when it is applied, and that in an age of AI a degree on its own proves very little.

Executive Summary

Students in India have long been required to study a wide spread of subjects, whether or not they enjoy them, and a common complaint follows them into adult life: why learn algebra, geometry or economics if your career never uses them? This paper takes the opposite view, with three arguments and a set of limits.

  • Breadth helps people choose. Meeting every subject early is how most people discover which ones they want to pursue in depth.
  • Breadth is what building a business demands. Economic research on founders suggests that people who start firms tend to be competent across many areas rather than excellent in one.
  • AI raises the value of breadth. AI can now carry out much of the specialist work, but only a person with a working grasp of the underlying subjects can direct it, check it and combine it.

The limits are real. Knowledge learnt in one setting does not move easily into another, and breadth without application turns into the rote learning that critics rightly object to. The conclusion is not "collect more subjects" or "collect more degrees". It is: build broad foundations, then practise applying them to real problems. Applied knowledge, not the certificate, is what employers, customers and businesses actually use.

Background: Why Does India Teach Every Subject?

For most of the past half century, a typical Indian school student has studied mathematics, the sciences, social sciences and languages side by side until at least the age of fifteen or sixteen, and only then narrowed into a stream. The approach has been criticised as rigid and exam-driven, and some of that criticism has been accepted at the highest level.

India's National Education Policy 2020, approved by the Union Cabinet in July 2020, did not abandon breadth. It doubled down on it while removing the walls between subjects. In the government's own summary, "There will be no rigid separations between arts and sciences, between curricular and extra-curricular activities, between vocational and academic streams", and "Students will have increased flexibility and choice of subjects". For higher education, the policy envisages "broad based, multi-disciplinary, holistic Under Graduate education with flexible curricula" and assessment that tests "analysis, critical thinking, and conceptual clarity".

In other words, the policy keeps the principle of learning widely and changes the method: less memorisation, more choice, more connection between subjects. That is very close to the argument of this paper.

The Complaint, and Why It Is Half Right

The complaint deserves a fair hearing. Many adults remember subjects they disliked, learnt by rote for an examination and never used again. When a graduate asks what the point of trigonometry was, they are often describing a real failure: the subject was taught as a set of answers to memorise, not as a tool for solving problems.

But the complaint confuses two things: the subject and the way it was taught. A civil engineer cannot design a safe structure without geometry. A software engineer cannot judge whether a system will cope with a million users without some feel for algorithms and growth rates. A trader who cannot reason about percentages and compound interest is at the mercy of whoever can. The subjects were not useless; their application was never shown.

There is also a second, quieter benefit of studying everything: it is how people find out what they are good at and what they enjoy. A student who never met economics or statistics at school is unlikely to discover a talent for either. Breadth early is a search process, and the later choice of specialism is better for it.

Market Landscape: What Employers and Founders Now Ask For

The labour market increasingly rewards the general, transferable abilities that broad schooling is meant to build. The World Economic Forum's Future of Jobs Report 2025 found that analytical thinking remains the top core skill employers look for, cited by nearly seven in ten employers surveyed, with resilience, leadership and creative thinking close behind, and curiosity and lifelong learning cited by about half and expected to rise in importance to 2030.

None of these is a subject on a timetable. Each is a capacity built by working across subjects: analytical thinking from mathematics and the sciences, communication from languages, judgement about people and markets from the social sciences. The report's direction of travel is towards people who can reason across domains, which is exactly what a narrow, single-subject education struggles to produce.

For founders, the demand is sharper still. A person starting a business is, in the early years, the finance team, the sales team, the product team and the compliance team at once. Every one of those roles draws on a school subject.

Analysis: Breadth as the Raw Material of Applied Knowledge

The Founder's Problem Is a Generalist's Problem

The economist Edward Lazear, in work published by the US National Bureau of Economic Research and later in the American Economic Review, proposed what is now called the "jack of all trades" view of entrepreneurship. In his framing, entrepreneurs "are jacks-of-all-trades who may not excel in any one skill, but are competent in many". Testing the idea on data from Stanford business school graduates, he reported that "by far the most important determinant of entrepreneurship is having background in a large number of different roles".

The intuition is simple. A specialist is paid for their strongest skill. A founder is held back by their weakest one: a brilliant product is still sunk by bad pricing, unpaid taxes or a cash-flow mistake. A commerce or management graduate who knows only one function well is exposed in every other one, and either has to learn the rest or depend entirely on others who know it. The less a founder depends on any one adviser, the better they can judge all of them.

Why the Algebra Did Not Go to Waste

Much of what businesses now automate is school mathematics in disguise. A sales funnel is a chain of multiplied probabilities. A pricing decision is an exercise in percentages and margins. A loan or an instalment plan is compound interest. A forecast is a weighted average with an error band. A ranking of options under uncertainty is expected value, a topic most students meet in their final school years and many dismiss as abstract.

The people who can build and check these models are rarely those who scored highest in one subject. They are those who can recognise which piece of earlier learning a business problem needs, and apply it. That recognition is the core of applied knowledge.

A Worked Case: One Formula, Several Businesses

Zrooth's own Intelligence Studio, the tool its teams use to decide which marketing and growth plays are worth funding, offers a concrete example. At its centre is an expected-value formula for a single customer or enrolment, built entirely from subjects taught at school:

Expected value per customer = (probability of conversion × probability of full payment × share kept × price) ÷ (1 + cost of capital)years until paid, minus the cost of delivery.

Each opportunity is then multiplied through the business's own funnel (how many people see an advertisement per rupee, click, enquire, apply and convert) to estimate the value produced per rupee spent. Because evidence is never complete, the tool calculates a low, base and high value, and ranks opportunities on the low estimate, so that a thinly supported idea has to clear a higher bar. Several AI models propose and challenge the opportunities; where a reviewing model finds no specific evidence behind a proposal, the proposal is held back rather than ranked.

Part of the formulaWhat it doesWhere it was first learnt
Conversion and payment probabilitiesSeparates an interested lead from money actually receivedProbability, school mathematics
Funnel multiplicationTurns advertising spend into expected customersRatios and percentages, middle school
Discounting by (1 + r)tValues money received later at less than money todayCompound interest, school mathematics
Low, base and high bandAdmits uncertainty instead of hiding itStatistics, range and spread
Share kept and delivery costCounts only the margin, not the headline priceCommerce and accountancy
Market scope adjustmentRecognises that a city, state and national audience cost different amounts to reachEconomics and geography

The striking point is not the sophistication of the formula; any part of it could be explained to a school leaver. It is its generality. Because it reasons about probabilities, prices, costs and time rather than about any one industry, the same formula has been applied across very different companies in the same group:

  • Zrooth, an education and business training company, for its AI Entrepreneurship Programme, where the "customer" is an enrolled participant.
  • Vyapitus Specialities, a speciality pharmaceutical company in nephrology, transplantation and renal nutrition, where the "customer" is a prescribing doctor, hospital pharmacy or trade partner.
  • Gensact, an IT services and AI automation company (see the Gensact case study), where the "customer" is a business client buying a website, application or automation.

The Vyapitus run is a useful illustration of the formula's discipline. Two AI models proposed and reviewed five plays, around renal nutrition, phosphate binders, transplant care and patient education. None reached unanimous agreement, and every one was held back for human judgement because the reviewing model found no specific evidence signal behind it. The figure below is taken from that report.

Intelligence Studio Report · Vyapitus Specialities · 14 November 2026

Executive summary (extract): "Vyapitus competes in a specialist, relationship-driven nephrology and transplant market against far larger Indian pharma houses, so every play below wins through scientific credibility, dependable supply and narrow, high-conviction HCP and trade targeting rather than outspending rivals."

Plays vetted: 5 Unanimous: 0 Held for judgement: 5 Models: 2

Held play: "CKD-MBD phosphate-binder specialty challenger to hospital pharmacies"
Reviewer's dissent: "No specific intel signal cited (e.g. tender, stockout, hospital shift); LinkedIn unsuitable for hospital pharmacy tenders; too generic to size budget."

Figure 1. Extract from a live Intelligence Studio report for Vyapitus Specialities. Values are omitted because offer economics had not yet been entered for this business; the extract shows the review step, not a financial result.

A tool that tells a founder "not yet, the evidence is thin" is worth more than one that ranks every idea as a winner. That discipline, not optimism, is what lets one formula travel from enrolments to prescriptions to software contracts.

Who Wrote the Formula, and Why

The formula's author is not a mathematician, and not an engineer. He is a business analyst and a graduate who looks at companies from the business side. The formula came out of a practical problem: the group's venture arm holds businesses as different as education, speciality pharmaceuticals and IT services, and he needed a common denominator to compare them and to normalise their expected future returns on a single scale.

The answer was not new mathematics. It was school probability, compound interest, percentages and basic statistics, combined with a commercial understanding of how each business actually makes money, and with AI used to research markets and challenge proposals. In his assessment the formula has held up, and been valid, in each business it has been applied to so far; that is an internal judgement, not an independent audit. It is, in a single example, the argument of this paper: knowledge learnt broadly, applied deliberately, by someone whose qualification says nothing about mathematics.

That is applied knowledge in practice: modest pieces of school learning, recognised, combined and applied to a real decision, then trusted because each piece can be explained.

Where AI Changes the Equation

AI has made it far cheaper to do specialist work a founder does not excel at: drafting a contract, writing code, analysing data, producing a forecast. That does not make the underlying subjects less relevant. It makes them more so, because the person directing the AI has to judge whether its output is right. A founder who never understood discounting will not notice when an AI-built model forgets it. A founder who never met probability will not question a funnel that assumes every lead pays.

Broad foundations are what allow a person to channel AI across areas where they are not an expert, and to catch its mistakes. This is the same argument made in an earlier piece on vibe coding and business building: the value lies in knowing what to build and what to check, not in typing it.

Risks and Limits of the Argument

Three limits deserve to be stated plainly.

Transfer is hard. A large body of research in psychology finds that knowledge learnt in one setting does not automatically carry over to another. In a widely cited review, Susan Barnett and Stephen Ceci note that the question of when and where people apply what they learn has resisted a simple answer for a century. Breadth on its own therefore guarantees nothing; it becomes applied knowledge only when learners repeatedly practise using it on unfamiliar problems.

The evidence on generalist founders is mixed. Later studies have supported Lazear's idea in some countries and data sets and weakened it in others, and some researchers suggest that general ability, rather than a balanced mix of skills, may explain part of the effect. The direction is consistent; the size of the effect is uncertain.

Breadth can become an excuse for rote learning. Studying twelve subjects badly, by memorisation, produces neither depth nor transferable skill. The defence of broad schooling in this paper is a defence of breadth with understanding, which is what NEP 2020 aims for, not of the exam-driven version many adults remember.

Recommendations

For each group with a stake in the question, a short set of practical steps:

  1. Students: treat disliked subjects as tools you may need later, and ask of each topic, "where would someone use this?". The question turns memorisation into understanding.
  2. Graduates: before adding another qualification, ask what it will let you build or decide that you cannot today. If the answer is unclear, open the books, use AI to study the subject directly, and apply it to a real problem first.
  3. Founders and managers: learn enough of every function (finance, sales, operations, law, technology) to judge the people and the AI tools doing that work. Depth in your product is necessary; it is not sufficient.
  4. Educators and trainers: pair every abstract topic with one real application. Compound interest with a loan; probability with a sales funnel; geometry with a floor plan.
  5. Employers: test for applied knowledge directly, with a real task, rather than inferring it from the number or names of degrees.

Practising this on real businesses is the approach of the AI Entrepreneurship Programme, in which participants cost a process before and after AI using exactly the mathematics described above. The full sequence is set out in what you learn, and Pathfinder shows how it applies to a given industry. For businesses, Zrooth Business Automation applies the same methods to an organisation's own processes.

Conclusion: Knowledge Is Not a Degree

A degree is evidence that someone completed a course. Applied knowledge is evidence that they can use what they learnt. The two overlap, but they are not the same, and in a period when AI can supply specialist output on demand, the gap between them matters more than ever. The skills that AI is rewarding rather than replacing are judgement, combination and verification, all of which depend on broad foundations.

India's habit of teaching every subject was not a mistake. Where it fell short, it was in teaching subjects as answers rather than tools. The remedy is not fewer subjects or more certificates; it is curiosity about how each subject is used, and the habit of applying it. The goal of education should be the ability to build something useful, not a longer list of qualifications.

Sources

Frequently Asked Questions

What is applied knowledge?

Applied knowledge is the ability to use what you have learnt, in any subject, to solve a real problem or make a real decision. It differs from knowing facts or holding a qualification in that subject.

Why does the Indian education system make students study every subject?

Broad schooling is meant to build general foundations and let students discover their strengths before specialising. NEP 2020 keeps that breadth while adding flexibility and removing rigid separations between streams.

Is it a waste to learn algebra and geometry if my career does not use them directly?

Rarely. They underpin percentages, compound interest, probability and measurement, which appear in pricing, finance, engineering, forecasting and software, even when nobody calls them algebra.

Do entrepreneurs need to be good at many subjects?

Research by the economist Edward Lazear suggests founders tend to be competent across many areas rather than excellent in one, because a business is held back by its weakest function. Later evidence is mixed on the size of the effect.

Does AI make broad knowledge less important?

No. AI makes specialist output cheap, so the scarce skill becomes directing and checking it. That needs a working grasp of the subjects involved, such as finance, statistics or law.

Is a degree the same as knowledge?

No. A degree shows that someone completed a course. Knowledge, and especially applied knowledge, shows that they can use what they learnt. The two overlap but are not interchangeable.

Should I pursue another degree or learn on my own?

Ask what the degree will let you build or decide that you cannot today. If it opens a regulated profession or fills a clear gap, it may be worth it; otherwise, studying and applying the subject directly is often faster.

How can I turn broad school learning into applied knowledge?

Practise on real problems: cost a process, price a product, model a sales funnel or read a balance sheet. Repeated application is what lets knowledge move from the classroom into work.

Filed Under India

Pathfinder · Free

Where Do You Fit in the AI Economy?

Type your industry and the work you do. See the live businesses you would learn inside, your twelve weeks, and three venture ideas made for you.

Find My Path

Join the Discussion

Agree, disagree or add your own experience. Every comment is reviewed before it appears; spam and abuse are removed, and disagreement is welcome.

    Add Your View

    0 / 2,000
    Links send a comment to manual review. Please do not share personal details about other people.