The central issue
Education has always been vocational in the broad sense: it prepares people to participate in economic and social life. But its purpose cannot be reduced to supplying labour for existing employers.
AI creates a deeper question:
What is education for when economically valuable intelligence is no longer exclusively human—and paid employment may no longer be the principal means through which everyone contributes or receives income?
The answer should not be to predict a list of “AI-proof jobs”. Those predictions will age quickly. The ILO’s current assessment is that about one-quarter of workers have some exposure to generative AI, but that job transformation is more likely than wholesale replacement. However, the distribution of productivity gains, bargaining power and income remains a political choice. (ilo.org)
1. The economic and social model being supported
The proposed education system should be designed to support four possible conditions simultaneously.
1. An AI-augmented economy
Most people continue working, but AI performs an increasing proportion of cognitive tasks. Human value shifts towards:
- defining worthwhile problems;
- exercising judgement under uncertainty;
- verifying AI-generated work;
- taking legal and ethical responsibility;
- integrating knowledge across domains;
- working with and caring for people;
- physical activity in unstructured environments;
- entrepreneurship and institution-building.
2. A turbulent transition
Jobs and tasks change faster than qualifications can be designed. People experience repeated periods of retraining, underemployment and occupational transition.
3. A more automated economy
Productivity increases while demand for human labour—or at least conventional full-time labour—declines. Working hours might fall, but only if productivity gains are distributed rather than concentrated.
4. A post-work or less-work-intensive society
Paid employment becomes only one recognised form of contribution alongside:
- caring;
- parenting;
- volunteering;
- cultural creation;
- environmental restoration;
- democratic participation;
- community activity;
- independent study.
The educational model must work reasonably well under all four scenarios, rather than depending on one forecast being correct.
2. Revised purpose of education
The future system would have five purposes.
Economic capability
Enable people to produce value with machines, not merely compete against them.
Human agency
Enable people to decide what should be done rather than simply execute tasks selected by employers or algorithms.
Social participation
Prepare people for relationships, citizenship, care and collective decision-making.
Adaptability
Make repeated learning and occupational change normal rather than treating retraining as evidence of failure.
Meaning and self-direction
Help people construct worthwhile lives even if paid employment occupies less of them.
This is where Sir Ken Robinson’s emphasis becomes even more relevant. Creativity is not an ornamental addition to economic education. It is the capacity to:
- generate possibilities;
- make connections;
- frame problems;
- experiment;
- create culture;
- develop an individual voice.
However, creativity must be combined with knowledge, disciplined practice and critical judgement. Otherwise, AI can generate plausible-looking outputs for people who lack the expertise to assess them.
3. Cohorts: motivation and means to actualise
Early childhood: approximately 0–7
Primary purpose
Develop the foundations of human agency:
- language;
- attachment and trust;
- physical development;
- curiosity;
- play;
- self-regulation;
- imagination;
- social interaction.
Motivation
Young children are motivated principally by:
- play;
- imitation;
- exploration;
- relationships;
- immediate feedback;
- a sense of growing competence.
The system should protect these intrinsic motivations rather than prematurely replacing them with performance anxiety.
Means
- Universal, high-quality early-years provision.
- Play-based learning.
- Rich language and story environments.
- Music, movement, construction, nature and art.
- Early identification of developmental needs.
- Extensive human interaction rather than automated childcare.
- Limited, supervised and purposeful use of AI.
- Support for parents as children’s first educators.
Economic and social contribution
This cohort is not being trained for a particular occupation. It is developing the cognitive, emotional and relational infrastructure from which all later adaptation proceeds.
Childhood: approximately 7–14
Primary purpose
Build powerful foundations while broadening—not narrowing—conceptions of ability.
Motivation
Children need:
- mastery;
- curiosity;
- belonging;
- recognition;
- increasing autonomy;
- opportunities to make things that matter.
Means
Every child should receive:
- strong literacy, numeracy and scientific reasoning;
- history, geography and civic understanding;
- art, music, drama, design and physical education;
- practical making and repair;
- computational and AI literacy;
- collaborative interdisciplinary projects;
- regular contact with nature, workplaces and community organisations.
AI should act as:
- tutor;
- practice partner;
- translator;
- accessibility aid;
- simulation environment.
It should not become the child’s principal teacher, decision-maker or source of emotional attachment. UNESCO has warned that educational AI can either enhance or diminish human agency depending on its design and governance. (unesco.org)
Actualisation
Each child would maintain a developmental portfolio containing:
- knowledge assessments;
- creative work;
- practical projects;
- performances;
- collaborative contributions;
- reflections;
- evidence of improvement.
This does not replace all testing, but prevents one examination format from becoming the definition of ability.
Adolescence: approximately 14–18
Primary purpose
Develop identity, capability and informed choice without forcing premature occupational specialisation.
Motivation
Adolescents are particularly motivated by:
- autonomy;
- identity;
- peer belonging;
- meaningful challenge;
- adult recognition;
- visible real-world consequences.
Means
A common core would include:
- English and communication;
- mathematics and statistics;
- science;
- humanities;
- civic, legal and political literacy;
- financial literacy;
- health and relationships;
- AI and data literacy;
- ethics and epistemology;
- creative and practical work.
Alongside this, each learner would pursue rotating pathways:
- academic inquiry;
- engineering and technical production;
- care and public service;
- business and entrepreneurship;
- arts and culture;
- environmental and community work.
Every learner would complete substantial real-world projects with employers, universities, councils, charities or community groups.
Actualisation
By 18, the learner should possess:
- foundational knowledge;
- a portfolio of completed work;
- some workplace experience;
- evidence of collaboration;
- evidence of independent creation;
- an understanding of their interests and strengths;
- several viable next steps.
The outcome is not “career certainty”. It is informed agency.
Emerging adulthood: approximately 18–25
Primary purpose
Replace the single university pipeline with a period of structured exploration, specialisation and contribution.
Motivation
Young adults need:
- independence;
- credible progression;
- income;
- social belonging;
- opportunities to test identities;
- evidence that effort produces capability.
Means
Every young adult would have access to a funded combination of:
- university modules;
- apprenticeships;
- paid work placements;
- technical institutes;
- public-service projects;
- enterprise incubators;
- creative practice;
- international or regional exchange.
Credits would be portable and stackable. A learner might combine engineering, business, ethics and an industrial placement rather than having to select one sealed institutional pathway.
Actualisation
Each person receives:
- a learning account;
- independent career guidance;
- access to AI tools and computing;
- a living-cost settlement while undertaking approved education, training or service;
- mentors from education and employment;
- transparent evidence about outcomes.
The motivation becomes:
“I am building a life and a portfolio of capability,”
rather than:
“I must obtain a degree because employers use it as a filter.”
Established adults: approximately 25–45
Primary purpose
Support professional growth, family formation, civic participation and transitions before displacement becomes a crisis.
Motivation
Adults generally learn when education is:
- relevant;
- compatible with their responsibilities;
- connected to income or meaningful activity;
- recognised by employers;
- immediately applicable.
Means
- Paid learning leave.
- Modular evening, online and workplace provision.
- Personal learning budgets.
- Employer co-investment.
- Recognition of existing capability.
- Conversion programmes between industries.
- Childcare and maintenance support.
- Access to diagnostic career guidance before redundancy.
- Worker participation in decisions about workplace AI adoption.
Training must not become an individual’s obligation to chase continually changing employer demands. Employers benefiting from automation should finance a significant part of transition provision.
Midlife: approximately 45–65
Primary purpose
Prevent accumulated experience from being discarded when particular technical skills become obsolete.
Motivation
Motivation is likely to come from:
- continued relevance;
- financial security;
- mastery;
- contribution;
- mentoring;
- the possibility of a meaningful second or third career.
Means
- Paid mid-career sabbaticals.
- Intensive conversion programmes.
- Recognition of professional and tacit knowledge.
- Routes into teaching, mentoring and public service.
- Support for enterprise and cooperatives.
- Flexible employment combined with study or care.
- AI training based on occupational context rather than generic courses.
A 50-year-old should not be required to imitate a 22-year-old entrant. Their experience in judgement, relationships, institutions and implementation should be recognised as an asset.
OECD work identifies mid-career as a critical intervention point because technological change, caring responsibilities and skills obsolescence converge there. (oecd.org)
Later life: approximately 65+
Primary purpose
Reject the assumption that education ends when conventional employment ends.
Motivation
- intellectual curiosity;
- social connection;
- contribution;
- health and independence;
- intergenerational exchange;
- legacy and purpose.
Means
- Free or low-cost access to colleges and universities.
- Community learning centres.
- Intergenerational teaching and mentoring.
- Support for civic research and local history.
- Digital and AI inclusion.
- Routes into part-time work, care, governance and volunteering.
- Recognition of community contribution.
The objective is not merely to keep older people economically productive. It is to retain knowledge, reduce isolation and sustain participation in society.
4. Does this require universal basic income?
Not necessarily—but it does require economic security for learning and transition.
As of July 2026, the UK operates a conditional, means-tested Universal Credit system rather than a national UBI. (gov.uk)
Several models are possible.
| Model | Strength | Risk |
|---|---|---|
| UBI | Simple income floor; supports risk-taking, care and learning | Cost, inflation and inadequate replacement of disability/housing support |
| Negative income tax | Targets support through the tax system | Delayed support and household-design problems |
| Universal basic services | Guarantees housing, health, transport, education and connectivity | Less individual flexibility |
| Job guarantee | Income, contribution and social structure | Risk of artificial or coercive work |
| Participation income | Recognises learning, care, volunteering and civic activity | Administrative monitoring and disputes over “valid” activity |
| Social dividend | Shares returns from collectively enabled AI productivity | Requires public ownership, taxation or capital funds |
| Learning income | Pays people during approved retraining | Too narrow if paid work itself contracts |
The strongest approach may be a hybrid social floor:
- universal basic services;
- a minimum income guarantee;
- learning and transition accounts;
- wage insurance during occupational change;
- portable benefits;
- additional support for disability, housing and care;
- potentially a public dividend from AI-related capital and productivity.
UBI alone does not create purpose, social connection or capability. Conversely, education accounts are of limited use if people cannot afford rent, childcare or time away from work.
5. What would motivate learning if work were optional?
This is a legitimate challenge. Much existing educational motivation comes from:
- fear of unemployment;
- qualification requirements;
- parental pressure;
- competition for income and status.
If economic compulsion weakens, some participation will decline. But that does not mean people cease to act. Motivation would shift towards:
- mastery;
- curiosity;
- recognition;
- belonging;
- service;
- creativity;
- status within communities;
- responsibility to others;
- access to interesting projects;
- a desire for greater-than-basic consumption.
The design should not assume either that everyone is naturally self-motivating or that coercion is the only motivator. It should create structured autonomy:
- secure foundations;
- genuine choices;
- mentors and communities;
- visible challenges;
- public recognition;
- deadlines and commitments;
- opportunities to produce something useful.
A UBI world would make Robinson’s conception of human diversity more feasible, but it would also make education’s role in cultivating purpose more important.
6. What remains invariant despite AI?
Some educational purposes survive nearly every plausible technological scenario:
- children still need attachment and language;
- citizens still need to understand power and evidence;
- humans still need relationships and belonging;
- society still requires ethical and political judgement;
- people still need enough knowledge to recognise machine error;
- concentrated wealth and informational power still require scrutiny;
- communities still need care, trust and collective action;
- individuals still need agency and meaning.
Therefore, the model remains valid—but its vocational purpose becomes broader:
Education prepares people not just to obtain employment, but to create economic value, govern technology, care for others, participate in democracy and construct meaningful lives.
7. The limits of the model
Yes, there are severe modelling constraints.
We cannot reliably know:
- the rate at which AI capabilities will improve;
- which occupations will expand or contract;
- whether productivity gains will reduce hours or increase inequality;
- whether robotics will match progress in software;
- how ownership of AI capital will be distributed;
- how governments will tax mobile capital;
- whether people will accept UBI or related settlements;
- how geopolitical, ecological and demographic shocks will interact;
- how human preferences change once work is less compulsory.
The danger is false precision. A curriculum designed around a confident 2040 job forecast may be obsolete before its first pupils graduate.
The appropriate response is not to abandon planning, but to design for resilience:
- Preserve strong foundational knowledge.
- Teach people how to learn and verify.
- Develop creativity, agency and social capability.
- Make qualifications modular and portable.
- maintain several academic, technical and creative pathways.
- Guarantee repeated access to learning.
- Link education policy to welfare, taxation and ownership policy.
- Review the system continuously against multiple economic scenarios.
Conclusion
The education model is valid only if it is not treated as a complete solution. Education can prepare people for transition, but it cannot determine who owns AI, who receives its economic gains or whether displaced workers have security.
Those are questions of political economy.
Education determines whether people possess the capability to flourish in an AI society. Economic institutions determine whether they have the resources, power and opportunity to use that capability.
Both systems must therefore be redesigned together.