Education 2047 #Blog 67 (20 SEP 2026)
An Architecture for Indian Education in the AI Age: From Annual Promotion to Capability Progression
For more than a century, education has moved to the rhythm of the calendar. A child enters Class I, spends a year there, moves to Class II and continues climbing. Later come secondary school, undergraduate years, postgraduate programmes and perhaps doctoral study. We have become so accustomed to this arrangement that moving ahead by one academic year appears almost synonymous with educational progress.
But why should learning advance one year at a time? A year is a unit of time. It is not a unit of learning. Children do not become curious annually. Understanding does not deepen on 31 March. Creativity does not wait for the next semester. A learner does not suddenly become capable of handling greater uncertainty because the examination section has issued a promotion order.
Annual academic progression was immensely useful for administering mass education. It allowed schools to create batches, timetables, classrooms, examinations and teacher workloads. But what is administratively convenient need not remain educationally optimal. The AI age gives us an opportunity to separate the two, and India now has both the need and the emerging digital infrastructure to attempt it.
From TarunKul and NavKul to a Larger Architecture
In my earlier writings under Education2047/ on LinkedIn, I have argued for rethinking different stages of education rather than merely modernising inherited structures. TarunKul looks at adolescence differently. Secondary education should not remain predominantly a waiting room for board examinations and entrance tests. These are years in which young people should discover curiosity, capability and character. They should explore, make, investigate, experiment and gradually understand what they can contribute.
NavKul, in turn, is my attempt to reinterpret the principles of the Gurukul for an AI-enabled age—not by returning to the past, but by recovering principles that industrial education pushed aside. In NavKul, the learner navigates rather than merely receives. Questions matter. Mentorship matters. Problems precede chapters. AI can help with factual retrieval and verification, while teachers increasingly operate in the territory of judgement, interpretation and cognitive elevation. Learning becomes visible through projects, portfolios, peer interaction, authentic challenges and demonstrated capability.
But TarunKul and NavKul lead naturally to another question: what should connect the entire educational journey? If learning itself is becoming adaptive, why should progression remain fixed? Perhaps the next step is to rethink the very spine of education.
Annual Promotion Is Not the Same as Educational Progression
Our existing system largely follows a familiar sequence: age, class, curriculum, examination, promotion and then the next class. Students who are approximately the same age encounter approximately the same content for approximately the same period. At the end of that period, we determine whether they have acquired enough marks to move forward.
This architecture solved the industrial-era problem of scale, but it treats time as a proxy for learning. A learner who understands a concept quickly must often wait for the class. Another learner who requires more time must somehow keep pace. One student may already be capable of sophisticated problem-solving in electronics but still be classified simply as “Class XI”. Another may be an extraordinary communicator, organiser or creator but appear mediocre because the examination privileges only a narrow range of cognitive performance.
The calendar smooths these differences administratively. Education should reveal them. The question should therefore change from “Has this learner completed the year?” to “What has this learner become capable of doing, and what kind of challenge is the learner now ready to undertake?” That moves us from promotion to progression. Promotion is administrative. Progression is developmental.
Four Learning Territories
I have been using four learning quadrants to distinguish learning according to whether the question and answer are known or unknown: KAKQ—Known Answer, Known Question; KAUQ—Known Answer, Unknown Question; UAKQ—Unknown Answer, Known Question; and UAUQ—Unknown Answer, Unknown Question. These should not become four new rigid compartments replacing the old ones. They are better understood as overlapping learning territories through which the learner progressively becomes capable of operating in environments containing greater uncertainty.
.png)
KAKQ: Known Question, Known Answer
Early learning necessarily contains substantial KAKQ. What is 5 × 7? What is photosynthesis? What does this word mean? Where is Kolkata? Human civilisation cannot require every child to rediscover established knowledge. Foundations are essential. The problem is not KAKQ. The problem is keeping learners predominantly in KAKQ for fifteen or twenty years. This is also the territory where AI is becoming increasingly capable. A machine can retrieve, explain, calculate, summarise and often produce a competent response almost instantly. Human educational time should therefore progressively move beyond it.
KAUQ: Known Answer, Unknown Question
In KAUQ, established knowledge exists, but the learner is no longer handed the question. Instead of asking “What are the causes of water pollution?”, the learner may be shown a polluted water body and expected to investigate what is happening. Instead of receiving a perfectly framed mathematics problem, the learner encounters a real situation and must determine what needs to be calculated. The learner begins to discover questions, not merely answers. This is where I see much of secondary education moving. TarunKul belongs naturally here. The adolescent should increasingly explore, question, experiment and discover relationships rather than simply accumulate another layer of textbook knowledge.
UAKQ: Known Question, Unknown Answer
At the next stage, the problem is known but the answer is not. How can this village conserve more water? Can this device be redesigned at half the cost? Can waste from one process become an input into another? How can traffic outside a school be reduced? There is no answer key at the back of the book. The learner must investigate, design, test, fail, refine and finally defend a solution. This should become a major territory of undergraduate education. NavKul moves naturally into this space. Higher education should increasingly begin with problems rather than chapters.
UAUQ: Unknown Question, Unknown Answer
The deepest educational territory begins when neither the significant question nor its answer has been given. The learner must first notice something worth asking. This is the domain of discovery, advanced scholarship, entrepreneurship, original research and breakthrough innovation. Postgraduate and doctoral education should increasingly move here. Research is not simply answering a difficult question. Often the greater scholarship lies in identifying which question deserves to be asked.
The broad progression therefore becomes KAKQ → KAUQ → UAKQ → UAUQ, or approximately Primary → Secondary → Undergraduate → Postgraduate/Doctoral. But only approximately. There should be no sharp boundaries. A school student may reach UAKQ in an area of exceptional interest. An undergraduate learning an unfamiliar concept may temporarily return to KAKQ. A doctoral researcher does this routinely. Learning is not a railway timetable. It is navigation.
Replace the Ladder with a Landscape
This changes the metaphor of education. Today we think in terms of a ladder: Class I, Class II, Class III, secondary, undergraduate, postgraduate and PhD. Each rung supposedly represents advancement. I would rather imagine a learning landscape. The four quadrants overlap within it. Students navigate through the landscape at different speeds and along different routes. The question is not “Which rung have you reached?” It is “What terrain can you now navigate?” That is a fundamentally different conception of progression.
|
Present stage |
Predominant learning territory |
Developmental purpose |
|
Primary |
KAKQ |
Build foundations and confidence |
|
Secondary |
KAUQ |
Develop questioning and exploration |
|
Undergraduate |
UAKQ |
Solve authentic problems with unknown answers |
|
Postgraduate / Doctoral |
UAUQ |
Discover questions, create knowledge and innovate |
How, Then, Do We Measure Progress?
If years and semesters cease to be the dominant measure, we require another basis for progression. I suggest two ideas: traction and rigour.
Traction
Traction asks whether the learner is actually moving. Can the learner proceed with increasing independence? Can a useful question be identified? Can knowledge be transferred from one situation to another? Can feedback be incorporated? Can the learner persist after failure? Can humans, AI and other resources be used intelligently? Can meaningful outcomes eventually be produced? Traction tells us whether learning has acquired momentum.
Rigour
Rigour asks how demanding the learning environment has become. How complex is the problem? How much uncertainty does it contain? How strong must the evidence be? How many variables have to be reconciled? How much disciplinary depth is required? How seriously are alternatives examined? How much responsibility does the learner carry for the final judgement? A learner who demonstrates traction under increasing rigour is progressing. That tells us much more than “Completed Academic Year 2028–29”.
At what level of uncertainty can this learner operate with sustained traction, adequate rigour and increasing independence?
The Other Half of the Architecture: Heart, Hands and Mind
There is another inherited distortion that the AI age allows us to correct. Education has disproportionately privileged the cognitive domain. Knowledge dominates teaching, examinations, admissions and academic prestige. But education develops a complete human being through at least three broad domains: affective, psychomotor and cognitive. All three should remain present throughout education, but I propose that their dominance should change across the learning journey.
Primary Education: Let the Heart Lead
At the primary stage, I would make the affective domain dominant. Before asking how much the child remembers, let us ask whether the child is curious, can cooperate, cares for others, listens, develops confidence, shows empathy, accepts failure and tries again, and finds joy in learning. Language, numeracy and foundational knowledge remain essential, but their foundation should be a child who wants to learn. We have perhaps spent too much effort filling young minds and too little ensuring that the heart remains open.
Secondary Education: Let the Hands Lead
In adolescence, psychomotor learning should become much more prominent. Let students build, repair, grow, cook, measure, design, perform, code, experiment, operate instruments, work in teams, engage with communities, and create physical and digital artefacts. TarunKul should be a place where adolescents encounter the world rather than merely read about it. Skills enabled humans to survive on this planet long before organised academic knowledge became widespread. The AI age makes the recovery of doing even more important.
Higher Education: Let the Mind Rise Higher
At higher levels the cognitive domain can become dominant—but not the cognition of memorising and reproducing. Higher education must increasingly demand cognition at levels where machines do not make the human contribution redundant: analysis, synthesis, abstraction, judgement, problem formulation, model building, research and creation. NavKul belongs naturally here. The student should move progressively from being taught what humanity already knows towards participating in what humanity does not yet know.
Throughout these stages, none of the other domains disappears. The engineer still requires ethics. The physician requires empathy and psychomotor competence. The researcher requires integrity. The entrepreneur requires relationships. Eventually, the three domains should converge into integrated human capability.
Figure 1. An Adaptive Education Architecture for the AI Age: From Annual Promotion to Capability Progression
The architecture can therefore be visualised broadly as an affective-led foundation, followed by psychomotor-led exploration, cognitive-led higher learning and, ultimately, integrated human capability. Running alongside this is the progression from KAKQ to KAUQ to UAKQ to UAUQ, with movement judged through increasing traction and rigour. This, to me, is much closer to an educational architecture than a timetable divided into years.
Adaptive. Personalised. Purposed.
Three characteristics should define this system. It should be adaptive, because learning should respond to demonstrated readiness rather than forcing everyone through identical time blocks. A learner ready for greater challenge need not wait because the academic year has not finished, while another learner should not be branded a failure simply because the same capability requires more time. Time can be flexible while expectations remain demanding.
It should be personalised, but personalisation should mean much more than an AI application choosing the next question. The educational trajectory itself should become personalised. A learner could be operating at UAKQ in electronics, KAUQ in economics and KAKQ in a new language. Why should all these differences disappear behind one class number and one overall percentage? The educational record should increasingly describe the learner, not merely the programme completed.
And it should be purposed. The learner should increasingly be able to answer: Why am I learning this? What capability am I developing? What can I create with it? Whose problem might it solve? How does this help me understand myself, other people or the world? Not everything must have immediate commercial application. Literature, mathematics, history, philosophy, music and art enrich human existence beyond employment. But examination is not purpose.
AI Makes Personalised Progression Scalable
An obvious question arises: how can one teacher manage forty learners navigating forty different trajectories? The industrial classroom had no easy answer. AI changes this practical constraint. A learned AI companion can help track progress, explain difficult concepts, translate, recommend resources, generate practice, help document projects, identify possible gaps and support reflection. But AI should remain an enabler, not the centre of education.
The more machines can do, the more important the human teacher becomes in domains that should remain human: mentoring, motivation, relationship, ethical judgement, challenge selection, socialisation, observation of development and validation of genuine capability. In NavKul language, the teacher progressively moves beyond being merely the Sage on the Stage or even only the Guide by the Side. The teacher increasingly becomes the Pack at the Back—supporting, watching, challenging and allowing the learner to navigate independently.
The Transcript Must Also Change
A system based on capability progression cannot retain a twentieth-century transcript. Today’s academic record usually tells us the subjects studied, marks obtained, credits accumulated, semesters completed and degree awarded. It tells us very little about the quality of challenge the learner can handle.
A future learning record should increasingly reveal what kind of problems the learner has attempted, at what quadrant, with what degree of independence, what rigour was involved, what evidence was created, whether the capability was validated by teachers, peers, industry or community, and what cognitive, psychomotor and affective development is visible. Instead of only saying “Physics—4 credits—Grade A”, the record might establish that the learner identified an experimental problem, worked successfully within UAKQ, designed an investigation, used AI appropriately, constructed an experimental arrangement, analysed results, defended conclusions, worked collaboratively and produced an authenticated artefact. Now the record begins describing capability.
India Has Already Built Much of the Digital Backbone
This is where the Indian context becomes particularly interesting. APAAR, the Academic Bank of Credits, the National Academic Depository and the National Credit Framework together provide elements of a digital architecture through which learning identities, credits and authenticated achievements can become portable. At present, we are understandably using much of this infrastructure to make the existing education system more efficient and flexible. But its larger significance may lie elsewhere.
It gives us the possibility of changing the centre of gravity from Institution → Programme → Semester → Course → Examination → Credit → Degree towards Learner → Purpose → Challenge → Experience → Demonstrated Capability → Verified Evidence → Learning Record → Next Challenge. The learner becomes the persistent entity. Institutions become important contributors to the learner’s journey, rather than the only legitimate containers within which learning can occur. This is the architecture India should eventually explore.
Credit Should Follow Capability, Not Merely Time
There is little purpose in creating sophisticated digital credit banks if what we deposit into them remains merely an electronic representation of classroom hours. A learner may acquire genuine capability through a workplace, a research project, a community intervention, a prototype, a patent, entrepreneurship, professional practice, an online programme, a competition, an open-source contribution or independent learning. Why should learning become valid only when an institution has arranged a timetable around it?
The standard should remain demanding. In fact, once seat-time loses importance, verification becomes more important. The future principle should therefore be simple: do not lower the standard; change the evidence.
AI Has Lowered the Entry to the Club of Literates
There is another transformation that India should examine carefully. Historically, entry into the knowledge system required substantial textual literacy. One needed to read, write, often know English, navigate formal documents and possess enough confidence to interact with institutional systems. AI is lowering this entry barrier.
A person can now increasingly speak rather than type, listen rather than read, communicate in a local language, translate, photograph something and ask a question about it, dictate a formal document, ask for a difficult text to be explained simply, and even instruct software through natural language. Reading and writing remain foundational, but the threshold for functional participation has fallen.
This could bring many more Indians—and eventually many more people globally—into meaningful participation in the knowledge economy. An artisan with limited formal education can communicate professionally. A farmer can interrogate technical knowledge conversationally. A worker can acquire new capabilities outside a long formal programme. A person unable to code conventionally can increasingly create through natural language. The club of functional literates therefore expands.
AI lowers the floor of literacy while raising the ceiling of scholarship.
At one end, more people enter. At the other, human beings must climb higher. If everybody can obtain an ordinary answer from AI, then merely knowing the answer cannot represent advanced scholarship.
Do Not Train Humans to Compete With Machines
This should perhaps become one of the governing principles of Education2047. We should not spend twenty years educating humans to become inferior versions of machines. Machines will increasingly calculate, retrieve, translate, summarise, optimise and generate. Humans should increasingly contribute what gives these capabilities direction: asking, imagining, judging, caring, choosing, taking responsibility, forming relationships and deciding what ought to be done.
The purpose of education should therefore be cognitive and human elevation, not preservation of activities technology has rendered routine. Humans must be prepared to complement machines, not compete with them.
Education for Work—and Beyond Work
There is another possibility we rarely discuss. What if intelligent machines eventually reduce the amount of human labour required to produce what society needs? Our reflex is usually to ask where we will find more jobs. But perhaps another question deserves equal attention: what should humans do with time liberated from economically necessary work?
That time could be used to care for children and elderly people, support communities, mentor younger generations, restore the environment, create art, play, learn, explore, conduct research, participate in civic life or simply deepen human relationships. Industrial education prepared us overwhelmingly for work. Education2047 may also need to prepare us for the wise use of leisure. Work, livelihood, contribution and human worth need not remain synonyms.
Care for Each Other. Care for the Planet.
This is ultimately why the affective domain cannot remain the neglected part of education. If machines increasingly provide cognitive assistance, the future does not require humans merely to know more. It requires humans to become more human. We will need people who can live with others, care for others, exercise restraint, use technology responsibly, understand ecological limits, create prosperity without assuming unlimited consumption, and use the extraordinary intelligence becoming available to us to improve human wellbeing rather than simply accelerate production.
Education should therefore prepare a young person for three relationships: with machines, with other humans and with the planet. Each requires capability. Each also requires judgement.
The Academic Year Can Remain—Without Ruling Education
Schools will still need calendars. Universities will still need schedules. Teachers need holidays. Families need predictability. Institutions need budgets and administrative cycles. The argument is therefore not to abolish the academic year. It is to remove its educational authority.
Calendar time can organise institutions. It need not define human learning. We can retain the year for administration while allowing learners to progress educationally according to demonstrated capability. The batch can remain socially useful without becoming cognitively restrictive.
From TarunKul to NavKul—and Beyond
Seen this way, TarunKul and NavKul become parts of a larger continuum. TarunKul helps adolescents move beyond known-answer schooling towards exploration, skills, questioning and character. NavKul takes the learner further into authentic problems, independent inquiry, mentorship, interdisciplinary challenges and eventually the unknown.
The four quadrants provide the navigation. The three domains provide human development. Traction and rigour provide progression. AI provides personalised intellectual assistance. Teachers provide human mentorship. Digital academic infrastructure provides continuity and authenticated records. Purpose provides direction.
The resulting architecture is no longer simply Primary → Secondary → Higher → Degree → Job. It becomes Foundation → Exploration → Capability → Discovery → Contribution → Renewal. Education becomes lifelong, adaptive, personalised and purposed.
Perhaps We Have Been Measuring the Wrong Movement
Every year, we proudly announce that millions of students have been promoted to the next class. But perhaps the more important question is how far they have actually travelled. Have they moved from answering questions to asking them? From knowing to doing? From doing to solving? From solving to discovering? From dependence to agency? From learning for marks to learning with purpose? From individual advancement to contribution? From competing with machines to complementing them? From consuming the world to caring for it?
That is educational progression. The calendar can tell us how much time has passed. Only the learner’s demonstrated capability can tell us how far the learner has moved.
Perhaps that should become the organising principle of Indian education as we move towards 2047.
* * *
Author
Dr. Neeraj Saxena is a former Scientist at TIFAC/DST and co-author of Educational Roadmap of India's Technology Vision 2035, with subsequent advisory roles at AICTE spanning higher education policy and implementation. He is currently Pro-Chancellor of JIS University, Kolkata, and publishes on AI-induced transformation in education through the Education2047 platform [nrj2000.blogspot.com] & [nrjsaxenajisu.substack.com]
©Dr. Neeraj Saxena, 2026 | CC BY-NC-ND 4.0 | Attribution required | No commercial use | No derivatives without permission.