Why Europe’s green and digital transitions depend on rebuilding the relationships through which knowledge becomes capability

SEO title: Europe’s Human Infrastructure in the Age of AI
Meta description: Europe needs more than AI, universities and online courses. Discover how mentorship, apprenticeships, local learning networks and community knowledge can build the skills needed for the green and digital transitions.
Suggested URL slug: /human-infrastructure-mentorship-skills-communities-ai/
Primary keyword: human infrastructure
Secondary keywords: future skills Europe, AI and mentorship, lifelong learning, apprenticeships, community learning, green skills, vocational education, future of work
Suggested category: Society / Future of Work / Education / European Transformation
Suggested reading time: 20–24 minutes

Introduction: Technology Does Not Transfer Knowledge by Itself

Europe is investing in artificial intelligence, advanced manufacturing, renewable energy, semiconductors, digital public services and climate-neutral buildings.

It is developing AI factories, data spaces, digital identity systems and new regulatory frameworks. Governments and industries increasingly understand that technological capability is essential to Europe’s economic resilience and strategic autonomy.

But technology does not implement itself.

A new energy system still requires electricians, engineers, planners, technicians, regulators and communities capable of maintaining it.

A digital building model still requires architects, craftspeople, contractors, inspectors and facility managers who understand what the information means.

An artificial-intelligence system still requires people who can evaluate its outputs, recognise its limitations, protect sensitive information and remain accountable for decisions.

A circular economy still requires someone who knows how a building was assembled, how a component can be removed without damage and whether a recovered material is safe to use again.

Europe’s transition is therefore not only a technological transformation.

It is a transformation of human capability.

The 2026 State of the Digital Decade reported that more than 60% of Europeans possessed at least basic digital skills. Yet ICT specialists represented only about 5% of European employment in 2025—half the EU’s 2030 target—with women still accounting for less than 20% of ICT specialists.

At the same time, shortages remain widespread across vocational occupations. Cedefop’s 2026 assessment identifies demographic ageing, low occupational attractiveness, gender imbalances, teacher shortages, working conditions and structural mismatches as major causes. Around 40% of vocational teachers are already aged fifty or older.

These are not isolated labour-market statistics.

They reveal a deeper structural problem.

Europe has spent decades improving the systems that produce information, but it has paid less attention to the relationships that turn information into confidence, judgement and practical skill.

Those relationships constitute human infrastructure.

Human infrastructure includes:

mentors who help others interpret unfamiliar situations;
experienced workers who transmit tacit knowledge;
teachers and trainers who connect theory to practice;
communities in which people learn from one another;
workplaces that allow experimentation without humiliation;
institutions that recognise skills acquired outside conventional education;
and local networks that connect talent with real social and economic needs.

Europe has not completely lost this infrastructure.

But it has fragmented it, undervalued it and too often treated it as an informal supplement rather than a strategic asset.

In the age of AI, Europe must rebuild it.

1. What Is Human Infrastructure?

Physical infrastructure allows energy, water, people and goods to move.

Digital infrastructure allows information and computation to move.

Human infrastructure allows knowledge, trust, responsibility and capability to move between people.

It is the social and institutional architecture through which a society learns.

A functioning human infrastructure helps someone move:

from curiosity to competence;
from education to employment;
from unemployment to participation;
from theoretical knowledge to practical judgement;
from one profession to another;
from novice to trusted practitioner;
and eventually from practitioner to mentor.

This process rarely occurs through course content alone.

People learn through observation, correction, repetition, dialogue, imitation, experimentation and participation in real communities of practice.

A textbook can describe how a structure transfers loads.

A calculation program can determine forces and utilisation ratios.

But recognising that an apparently minor site deviation may create a serious load-path problem requires experience, context and professional attention.

A BIM system can contain thousands of product properties.

But understanding whether a proposed assembly is buildable, maintainable and robust under local weather conditions requires more than access to information.

An AI assistant can explain how to perform a task.

But it may not recognise that the user has misunderstood the physical situation, lacks the necessary authorisation or is about to make an irreversible mistake.

Human infrastructure provides the context around knowledge.

It helps people know:

when a rule applies;
when an exception matters;
when evidence is weak;
when to ask for help;
when to stop;
and when they are ready to assume responsibility.

That is why human infrastructure cannot be reduced to an online learning platform.

It is an ecosystem of people, places, institutions, tools and relationships.

2. Europe’s Skills Paradox

Europe possesses highly developed education systems, research institutions and professional communities.

Yet employers across many sectors still struggle to find people with the required skills.

The European Labour Authority reports persistent and increasing labour-market imbalances across European countries, with shortages recurring in sectors such as healthcare, engineering, construction, transportation and information technology.

At the same time, many people remain underemployed, excluded from work or unable to convert existing experience into recognised opportunity.

This is Europe’s skills paradox:

Skills are scarce, while human capability remains underused.

The problem is not simply that Europe needs more people with degrees.

It is that the systems connecting people, education, employers and social needs are too slow, too fragmented and too difficult to navigate.

2.1 Formal qualifications move more slowly than work

Traditional education programmes require planning, accreditation, recruitment, teaching and evaluation.

These processes protect quality, but they can take years.

Technology, regulation and industrial practices may change faster.

By the time a full programme has been redesigned, employers may already require new combinations of:

digital competence;
environmental knowledge;
communication;
data literacy;
systems thinking;
practical experience;
and sector-specific judgement.

This does not make formal education obsolete.

It means that formal education must be connected to faster layers of learning.

2.2 Training reaches those who already have advantages

Adult participation in learning remains uneven.

Eurostat reported that 46.6% of adults aged 25–64 participated in formal or non-formal learning during the preceding twelve months in 2022, meaning that more than half did not participate.

The people who most need new opportunities often face the greatest barriers:

cost;
time;
transport;
family responsibilities;
insecure employment;
weak digital confidence;
previous negative experiences with education;
health limitations;
language barriers;
and uncertainty about which training will actually lead somewhere.

A course can be technically available while remaining practically inaccessible.

2.3 Employers ask for experience before providing experience

Many entry-level positions require applicants to demonstrate workplace competence.

Yet people cannot acquire workplace competence unless employers allow them to enter a workplace.

This creates a circular barrier.

Those with strong networks, financial support or conventional educational records may find a route through it.

Others remain outside despite possessing motivation and transferable ability.

2.4 Tacit knowledge disappears quietly

A large proportion of professional expertise is not fully written down.

It exists in habits, observations, sequences, caution, intuition and pattern recognition developed through repeated practice.

When an experienced craftsperson, engineer, nurse, technician or project leader retires without structured knowledge transfer, an institution loses more than one employee.

It loses part of its operational memory.

Demographic ageing makes this increasingly important. Cedefop’s 2026 analysis warns that workers nearing retirement already represent around 10% of total EU employment, while ageing is particularly pronounced among vocational teachers.

Europe cannot replace this knowledge simply by storing documents.

It must create relationships through which experienced people can demonstrate, explain and contextualise what they know.

3. Why the Course-Based Model Is Not Enough

Europe often responds to a skills shortage by proposing more courses.

Courses are necessary, but the assumption that a learning deficit can always be solved by adding content is incomplete.

A person can complete a course and still lack:

confidence;
practice;
professional networks;
contextual understanding;
access to equipment;
feedback;
recognised experience;
or a realistic pathway into work.

The distinction is between learning provision and capability formation.

Learning provision delivers information.

Capability formation enables a person to use knowledge reliably in real situations.

Capability formation usually requires several elements:

A clear objective.
Accessible foundational knowledge.
Demonstration by someone competent.
Opportunities to practise.
Feedback while mistakes remain reversible.
Increasing responsibility.
Reflection on results.
Recognition of the competence gained.
Connection to a real opportunity.

Most conventional courses provide only some of these elements.

Human infrastructure provides the rest.

4. Mentorship Is Infrastructure, Not Charity

Mentorship is often presented as a benevolent activity performed after normal work is complete.

That framing undervalues it.

Mentorship is a mechanism for:

transferring knowledge;
reducing avoidable mistakes;
building professional identity;
improving confidence;
retaining workers;
expanding networks;
identifying hidden talent;
and preserving institutional memory.

The European Commission is increasingly recognising mentoring as part of skills policy. Its basic-skills initiatives include mentoring and tutoring intended to support academic and social development while strengthening intergenerational solidarity. The Commission is also developing a mentoring system for early-career teachers and supporting STEM mentoring initiatives.

A strong mentorship model should not depend entirely on informal personal chemistry.

It needs structure.

4.1 A mentor is not merely an expert

Expertise matters, but mentoring also requires the ability to:

listen;
understand the learner’s starting point;
explain reasoning;
give proportionate feedback;
create psychological safety;
and recognise when another specialist is needed.

The best practitioner is not automatically the best mentor.

Mentors therefore need support and training of their own.

4.2 Mentoring must be reciprocal

The traditional image is one-directional: an older expert transfers knowledge to a younger novice.

In reality, effective mentoring can be reciprocal.

An experienced carpenter may teach material judgement, tolerances, sequencing and buildability.

A younger BIM specialist may teach digital coordination, reality capture and model-based quality control.

An architect may explain design intent.

A facility manager may reveal what repeatedly fails after handover.

A community representative may identify social and cultural consequences ignored by technical teams.

Each participant holds part of the knowledge.

This reciprocal model avoids treating experience as obsolete or technology as automatically superior.

4.3 Mentoring must be recognised as productive work

Organisations often claim to value knowledge transfer while allocating no time for it.

The result is predictable: experienced workers are expected to maintain full productivity while training others informally.

Mentorship becomes rushed and inconsistent.

A serious human-infrastructure strategy must recognise mentoring through:

paid time;
workload allocation;
professional credits;
career progression;
organisational performance measures;
and public support for smaller employers.

Knowledge transfer creates value even when it does not generate an immediately billable output.

5. Apprenticeship for More Than Traditional Trades

Apprenticeship remains one of the most effective models for connecting learning to real work.

Its principles are powerful:

participation in a real practice;
gradual responsibility;
close feedback;
recognised competence;
and a relationship between learning and productive contribution.

Cedefop and the OECD reported in 2026 that apprenticeship is expanding into occupations and sectors that have not traditionally used it, reflecting its relevance beyond established craft pathways.

Europe should apply apprenticeship principles more widely.

Potential fields include:

artificial-intelligence assurance;
cybersecurity;
digital heritage;
energy modelling;
circular procurement;
BIM coordination;
data stewardship;
climate adaptation;
community health;
robotics maintenance;
and public-sector digital transformation.

This does not mean weakening professional standards.

It means creating additional entry pathways into complex fields.

A person may begin through a structured project placement, complete modular learning, work under supervision and gradually earn recognised responsibility.

Such pathways can complement universities, vocational schools and professional certification.

Micro-apprenticeships

Not every learner or employer can commit immediately to a multi-year programme.

A micro-apprenticeship could last several weeks or months and focus on a defined, real-world problem.

Examples include:

creating a material inventory for a renovation project;
digitising maintenance records for a municipal building;
mapping accessibility barriers in a town centre;
validating selected BIM information requirements;
documenting local craft techniques;
or analysing energy data for a community facility.

The learner produces something useful.

The employer or community gains capacity.

The mentor can observe actual performance.

The resulting evidence can contribute to a skills portfolio or microcredential.

Micro-apprenticeships should not replace properly paid jobs or established apprenticeships.

They should serve as protected bridges into them.

6. AI Changes the Skills Problem—but Does Not Eliminate It

Artificial intelligence can explain concepts, translate material, simulate scenarios, generate exercises, analyse documents and provide feedback.

This could dramatically expand access to learning.

However, AI also creates new risks.

A learner may receive an answer that is fluent but wrong.

An employee may become dependent on a system without understanding the underlying task.

A workplace may automate entry-level activities that previously allowed novices to develop competence.

An organisation may mistake tool adoption for genuine capability.

The OECD’s 2025 analysis found that policies specifically targeting AI skills remained limited relative to employer demand. Its 2026 work also indicates that workers who receive training are more likely to report positive outcomes from AI adoption, including improved performance and working conditions.

This suggests an important principle:

AI creates the most value when it is embedded in a learning environment—not merely distributed as a tool.

6.1 AI as a learning companion

A well-designed AI learning companion can:

explain material at different levels;
translate terminology;
create practice exercises;
help learners formulate questions;
compare approaches;
provide immediate low-risk feedback;
identify gaps;
and recommend relevant human support.

6.2 AI as a mentor amplifier

AI can help mentors by:

preparing personalised learning plans;
summarising previous sessions;
documenting progress;
suggesting exercises;
retrieving relevant standards;
and reducing administrative work.

This enables human mentors to spend more time on judgement, motivation, context and relationships.

6.3 AI as a knowledge-capture assistant

Experienced workers often possess knowledge but lack time to document it.

AI can assist by:

transcribing demonstrations;
structuring interviews;
identifying repeated patterns;
linking explanations to images or model objects;
creating searchable knowledge entries;
and producing draft instructional material for human review.

The objective is not to extract knowledge from workers and discard them.

It is to preserve their contribution and extend their capacity to teach.

6.4 AI must not become the only teacher

AI cannot reliably determine every learner’s emotional state, workplace safety or readiness for responsibility.

It cannot replace social belonging.

It cannot provide professional legitimacy on its own.

It cannot accept moral or legal accountability.

Human learning must therefore remain human-governed, even when AI provides substantial assistance.

7. The Distributed European Learning Ecosystem

Europe needs an architecture that connects formal education, working life, technology and community.

This can be understood as a distributed learning ecosystem with eight interconnected layers.

Layer 1: Local learning hubs

Learning opportunities need physical and social presence.

Local hubs could operate through:

libraries;
vocational schools;
universities;
maker spaces;
community centres;
municipal buildings;
industry clusters;
cultural institutions;
and underused commercial premises.

A hub should provide more than classrooms.

It can provide:

equipment;
internet access;
mentors;
project rooms;
career guidance;
childcare coordination;
demonstrations;
and connections to employers.

Local presence matters because participation falls when learning requires difficult travel, unfamiliar institutions or complex digital navigation.

Layer 2: A trusted mentor network

Mentors should be discoverable by:

sector;
competence;
language;
location;
availability;
and type of support.

The network should include:

professionals;
retired specialists;
craftspeople;
educators;
students;
entrepreneurs;
community leaders;
and people with lived experience of career transitions.

Safeguarding, expectations and boundaries must be clear.

Mentors should not be expected to become social workers, therapists or unpaid recruitment agencies.

Their role should be defined and supported.

Layer 3: Real project opportunities

Learning becomes powerful when connected to something that matters.

Municipalities, SMEs, nonprofits, cultural organisations and research teams could publish bounded project challenges.

Learners could contribute under supervision.

This would help communities solve real problems while allowing participants to build demonstrable experience.

Layer 4: Modular and stackable learning

People need learning units that fit around work and life.

Microcredentials and modular training can help jobseekers and experienced workers adapt to changing labour-market needs, particularly when they are integrated with skills-based recruitment and trusted employer collaboration.

Modules should be stackable.

A learner might combine:

digital fundamentals;
sector safety;
BIM literacy;
environmental assessment;
communication;
and a supervised project.

Together, these could form a recognised competence profile.

Layer 5: Skills portfolios and evidence

A person’s capability should not be represented only by job titles and educational certificates.

A trusted skills portfolio could include:

completed learning;
project contributions;
mentor assessments;
verified work samples;
microcredentials;
professional licences;
and reflective documentation.

The portfolio must distinguish between self-asserted and independently verified claims.

It should also protect personal data and avoid creating a permanent surveillance record.

Layer 6: AI-assisted guidance

Many people do not know which skills they possess, which roles might suit them or what learning sequence would be realistic.

AI can assist with:

competence mapping;
opportunity matching;
gap analysis;
personalised learning routes;
and navigation of public support.

But recommendations must remain explainable.

A person should understand why a role or course was suggested and be able to challenge the assumptions behind it.

Layer 7: Employer and institutional partnerships

Human infrastructure cannot be built by education providers alone.

Employers must help define real needs, offer supervised practice and recognise alternative evidence of capability.

The EU Pact for Skills had expanded by March 2026 to around 4,000 members, with 20 large-scale sector partnerships and 22 regional partnerships.

This demonstrates the scale of interest.

The next step is to translate partnerships into local opportunities that ordinary people can access.

Layer 8: Public governance and durable financing

Learning ecosystems cannot rely solely on short pilot grants.

They need:

long-term governance;
quality standards;
outcome evaluation;
transparent funding;
participant protection;
and integration with employment and education systems.

Public investment should measure more than course completion.

It should examine:

sustained employment;
confidence;
competence application;
employer satisfaction;
social participation;
mentor retention;
and the continued value of community projects.

8. The Construction Sector as a Human-Infrastructure Test Case

Construction demonstrates why Europe needs integrated learning systems.

The sector must simultaneously respond to:

decarbonisation;
energy renovation;
digitalisation;
material reuse;
climate adaptation;
new product regulation;
automation;
demographic change;
and shortages of skilled workers.

No single profession can manage this transition alone.

Consider the renovation of an existing public building.

The project may require:

historical understanding;
moisture assessment;
structural analysis;
energy modelling;
asbestos or hazardous-material expertise;
BIM and reality capture;
circularity assessment;
craft repair;
accessibility improvements;
fire-safety evaluation;
procurement;
and engagement with users.

A conventional project may separate these disciplines into contractual packages.

A learning ecosystem can also treat the project as a knowledge-transfer environment.

An experienced craftsperson can teach repair techniques.

A younger technician can capture geometry digitally.

An energy specialist can explain thermal bridges.

A reuse coordinator can create a material inventory.

A facility manager can identify recurring operational problems.

An AI assistant can organise documentation and retrieve relevant requirements.

Students and career changers can participate in bounded tasks under supervision.

The building becomes both a physical project and a shared learning environment.

This model could help Europe increase renovation capacity without separating education from real delivery.

9. Norway: A Strong Foundation with Missing Connections

Norway already possesses several elements of a strong human-infrastructure system:

vocational education;
apprenticeship traditions;
relatively high trust;
organised working life;
universities and fagskoler;
public career services;
regional industry clusters;
and extensive professional associations.

The current Norwegian competence reform recognises the need for people to update their skills throughout working life. In March 2026, the government allocated NOK 142 million to industry programmes for continuing education, while also expanding the programme into additional sectors.

Norwegian authorities have also emphasised that more people need access to education through universities, vocational colleges and upper-secondary pathways while remaining connected to working life.

Yet the connections between these systems remain difficult for many people to navigate.

A person may need separate contact with:

NAV;
a municipality;
a county authority;
an educational provider;
an employer;
an industry organisation;
and a funding programme.

Each institution may provide useful services while the overall journey remains fragmented.

Norway’s opportunity is therefore not merely to create more programmes.

It is to create a clearer capability pathway.

A participant should be able to move through:

Initial conversation and competence mapping.
Identification of realistic goals.
Connection with a mentor.
Foundational or modular learning.
Participation in a supervised project.
Documentation of demonstrated competence.
Connection to employment, education or entrepreneurship.
Continued peer and mentor support.

Such a model can connect welfare, education and work without reducing the individual to a case number or course participant.

10. SamfunnsBro as a Practical Human-Infrastructure Model

Jarlhalla’s SamfunnsBro can become a real-world implementation of this architecture.

Its components can operate as connected functions rather than separate webpages.

MentorBro: The relationship layer

MentorBro can connect people seeking guidance with individuals willing to share professional, practical or life experience.

Matching should consider more than professional categories.

It should include:

goals;
communication style;
language;
availability;
geographical proximity;
lived experience;
and the boundaries of the mentoring relationship.

MentorBro should support both traditional and reverse mentoring.

ArbeidsBro: The opportunity layer

ArbeidsBro can connect competence to:

employment;
apprenticeships;
micro-apprenticeships;
short projects;
work trials;
collaborations;
and socially useful assignments.

The platform should enable employers to describe the actual capability needed rather than relying only on conventional credentials.

KunnskapsBro: The learning layer

KunnskapsBro can organise:

courses;
guides;
demonstrations;
standards;
recorded experience;
community knowledge;
and AI-assisted learning routes.

Content should be linked to specific goals and projects.

Knowledge becomes more useful when the learner understands where it can be applied.

SkaperBro: The project and innovation layer

SkaperBro can connect multidisciplinary teams around real challenges.

An architect, programmer, carpenter, student and local historian might collaborate on a heritage documentation project.

A municipality, energy adviser and community group might develop a neighbourhood renovation concept.

A retired engineer might mentor a younger team developing a circular-construction tool.

SkaperBro creates the arena in which learning becomes contribution.

One connected journey

The real strength of SamfunnsBro lies in the connections:

A person enters through curiosity, meets a mentor, follows a learning path, joins a project, demonstrates capability and reaches a real opportunity.

That is human infrastructure.

11. A European Roadmap for Human Infrastructure

Phase 1: Map capability and knowledge-loss risks — 2026–2027

Public authorities, industries and major employers should identify:

occupations facing retirement pressure;
critical tacit knowledge;
persistent recruitment gaps;
regions with weak learning access;
groups excluded from existing provision;
and technologies changing job content.

The mapping should examine not only how many workers are needed, but which knowledge relationships are disappearing.

Phase 2: Build mentor and project networks — 2027–2028

Regions should establish trusted mentor pools and catalogues of supervised project opportunities.

Mentors should receive guidance, recognition and compensation where appropriate.

Projects should be small enough to manage but real enough to generate credible evidence.

Phase 3: Connect modular learning to practice — 2028–2029

Education providers should connect microcredentials and short courses directly to:

projects;
apprenticeships;
employer needs;
and professional frameworks.

Learning records should become portable across institutions and borders.

Phase 4: Deploy human-governed AI support — 2028–2030

AI systems can support guidance, translation, tutoring, documentation and matching.

Governance should ensure:

transparency;
privacy;
human review;
accessibility;
and protection against automated exclusion.

Phase 5: Integrate local ecosystems across Europe — 2030–2032

Local and regional networks should remain locally rooted while using interoperable European standards.

A person’s verified learning and project experience should be understandable across borders without requiring every local community to use the same platform.

Europe should become a network of learning regions rather than one centralised education machine.

12. Principles for a Human-Centred Skills System

1. Start with the person, not the programme

The objective is not to fill courses.

It is to help people build meaningful capability and participation.

2. Recognise multiple forms of knowledge

Academic, vocational, professional, cultural and lived knowledge all have value.

They require different validation methods, not a single hierarchy.

3. Connect every learning activity to a possible next step

A learner should understand where an activity may lead.

4. Protect quality without protecting unnecessary barriers

Standards are essential where safety and public trust are involved.

But outdated credential requirements should not prevent capable people from demonstrating what they know.

5. Pay for knowledge transfer

Mentorship and supervision are productive activities.

They should not depend solely on unpaid goodwill.

6. Use AI to widen participation

AI should make knowledge more accessible and mentors more effective.

It should not centralise power or remove human accountability.

7. Build locally and connect openly

Human trust forms locally.

Knowledge and opportunity should still be able to move across regions and borders.

8. Measure long-term capability

Success is not the number of people who clicked through a course.

Success is the number who gained confidence, contributed, progressed and remained able to learn.

Conclusion: The Future Is Built Between People

Europe often speaks of the green transition and the digital transition as though they were parallel programmes.

In reality, both depend on a third transition:

the transformation of how people learn from one another.

Europe can finance advanced infrastructure and still fail to operate it.

It can deploy AI and still lack the judgement to use it responsibly.

It can produce regulations and still lack the professional capacity to implement them.

It can create thousands of courses while leaving people uncertain about how to enter real work.

The missing infrastructure is human.

It exists in the conversation between an experienced practitioner and someone taking their first uncertain step.

It exists when a workplace allows learning rather than demanding instant perfection.

It exists when a community treats its knowledge as something worth preserving.

It exists when retired professionals remain contributors rather than becoming invisible.

It exists when technical experts listen to craftspeople, residents and local cultures.

It exists when AI helps people ask better questions without pretending to replace responsibility.

It exists when education, work and civic contribution become parts of one continuous journey.

Europe’s future will not be secured by technology alone.

It will be secured by people who know how to understand it, question it, improve it, repair it and use it in service of society.

Mentorship is not a secondary social programme.

Apprenticeship is not an outdated educational form.

Community learning is not merely a local cultural activity.

Together, they are strategic infrastructure.

Europe has built networks for energy, transportation, finance and information.

Now it must build the network through which human capability can move.

That network begins wherever one person shares knowledge, another person is given room to practise and a community creates a meaningful place for both.

The future is not transferred through software.

It is built between people.

The Human Infrastructure Europe Forgot: Mentorship, Skills and Communities in the Age of AI