Leading with E.T.H.I.C.S. in the Age of Artificial Intelligence

"Weak leaders will use AI to justify decisions.

Strong leaders will use AI to improve decisions"

Artificial Intelligence (AI) is rapidly reshaping the landscape of leadership. While AI offers unprecedented capabilities in data analysis, predictive modeling, and operational optimization, it also introduces complex ethical challenges. The question for modern leaders is no longer whether AI should be used, but how it should be used responsibly.

Strong ethical leaders do not rely on AI to justify decisions. Instead, they leverage AI to enhance transparency, improve judgment, and strengthen accountability. A practical way to guide leadership in this evolving environment is to apply the ETHICS framework, an acronym that outlines six principles for responsible AI-enabled leadership: Evidence-based decision-making, Transparency, Human accountability, Integrity in AI governance, Continuous ethical learning, and Stakeholder-centered thinking.

Evidence-Based Decision Making

AI enables leaders to shift from intuition-driven decisions to evidence-based leadership. Advanced analytics can process large volumes of operational, environmental, and social data to reveal patterns that may not be immediately visible to decision-makers.

For example, AI systems can analyze workplace safety incidents, energy consumption trends, supply chain performance, or employee engagement data. By examining these patterns, leaders can detect emerging risks and opportunities earlier, enabling more informed and balanced decision-making.

Evidence-based leadership reduces reliance on subjective assumptions and helps organizations align strategic choices with measurable outcomes.

Transparency

Transparency is a cornerstone of ethical leadership. AI technologies can significantly enhance organizational transparency by enabling real-time monitoring and reporting.

AI-powered dashboards can track key indicators such as environmental performance, compliance metrics, safety records, and operational efficiency. This level of visibility enables leaders to identify issues promptly and communicate performance openly with stakeholders.

Transparent systems strengthen trust across the organization and with external partners. When decisions are supported by data and clearly explained, stakeholders are more likely to perceive leadership actions as fair and responsible.

Human Accountability

Despite the sophistication of AI systems, ethical responsibility cannot be delegated to algorithms. Leaders remain accountable for decisions made with the support of AI tools.

Strong leaders critically evaluate AI-generated insights by questioning assumptions, reviewing underlying data sources, and considering potential unintended consequences. Human judgment remains essential in interpreting recommendations and ensuring that decisions align with organizational values and societal expectations.

In this context, AI should be viewed as a decision-support instrument rather than a decision-maker.

Integrity in AI Governance

As organizations increasingly deploy AI systems, leaders must establish clear governance frameworks to ensure responsible use.

Effective AI governance includes measures such as algorithm transparency, bias detection, ethical data management, and robust cybersecurity protocols. Leaders must ensure that AI applications are aligned with legal requirements, organizational values, and ethical standards.

Without proper oversight, AI systems may inadvertently amplify bias or generate outcomes that undermine fairness. Strong governance safeguards against these risks and reinforces organizational integrity.

Continuous Ethical Learning

The rapid evolution of AI technologies requires leaders to adopt a mindset of continuous learning. Ethical leadership in the digital age involves staying informed about emerging technologies, regulatory developments, and evolving societal expectations.

Organizations should invest in training programs that enhance AI literacy across leadership teams. By understanding both the capabilities and limitations of AI, leaders can engage more effectively with technology experts and make more responsible decisions.

Continuous learning ensures that ethical considerations remain embedded in the organization's strategic development.

Stakeholder-Centered Thinking

Ethical leadership extends beyond financial performance. It requires leaders to consider the broader impact of decisions on employees, customers, communities, and the environment.

AI can assist leaders by modeling potential outcomes across multiple stakeholder groups. For example, AI can simulate the social and environmental implications of operational changes, supply chain restructuring, or infrastructure investments.

By incorporating these insights into decision-making processes, leaders can pursue strategies that balance economic performance with long-term societal value.

Conclusion

AI is transforming the tools available to leaders, but it does not replace the fundamental responsibilities of leadership. Instead, it amplifies the consequences of leadership decisions.

Organizations that integrate AI within a robust ethical framework will be better positioned to navigate complexity, maintain stakeholder trust, and achieve sustainable performance.

Ultimately, the effectiveness of AI in leadership depends on the values guiding its use. Ethical leaders do not ask AI to make decisions on their behalf. Rather, they use AI to deepen their understanding of consequences, challenge assumptions, and strengthen the integrity of their decisions.

In an era defined by technological acceleration, the ETHICS framework provides a structured approach for leaders seeking to harness the power of AI while upholding the highest standards of responsible leadership.

The Mathematics Behind AI-Enabled Facility Management

 

Facility Management is becoming more data-driven. Buildings today generate huge amounts of data from:

• sensors

• energy meters

• Building Management Systems (BMS)

• maintenance records

Artificial Intelligence promises to turn this data into insights. But there is an important point many people overlook:

Buildings follow physical laws.

If we combine engineering formulas with AI mathematics, we can unlock powerful building analytics.

This article explains how engineering mathematics and AI mathematics work together.

Step 1 – Turning Sensor Data into Engineering Insight

Buildings generate raw data such as:

• temperature

• water flow

• energy consumption

• equipment runtime

Raw data alone is not very useful.

Engineers convert data into meaningful indicators using formulas.

For example, cooling energy in a chilled water system depends on:

Water Flow × Temperature Difference

In other words:

More water flow or larger temperature differences mean more cooling energy is delivered.

This simple relationship allows facility managers to detect problems such as:

• low Delta-T syndrome

• inefficient coils

• excessive pumping

These calculated indicators become inputs for AI models.

Step 2 – Linear Algebra: Looking at Many Systems at Once

Large buildings may have:

• many air-handling units

• multiple chillers

• dozens of pumps

Instead of analyzing each system one by one, AI represents them as lists of numbers (vectors).

For example:

Cooling loads across systems might look like:

System 1: 120 kW
System 2: 95 kW
System 3: 110 kW
System 4: 140 kW

AI tools can analyze all systems simultaneously.

This allows:

• benchmarking across equipment

• detecting abnormal systems

• comparing buildings across portfolios

This mathematical approach comes from linear algebra, which is the foundation of machine learning.

Step 3 – Calculus: Optimizing Building Performance

Buildings constantly operate under changing conditions:

• outdoor weather

• occupancy levels

• equipment performance

AI systems try to find the most efficient operating point.

Think of it like adjusting controls to answer the question:

“What combination of pump speed, airflow, and temperature gives the lowest energy use?”

Calculus provides the mathematical tools that allow AI to gradually move toward the best operating condition.

This is similar to how navigation apps find the shortest route.

Step 4 – Probability: Predicting Equipment Failures

Equipment failure is never perfectly predictable.

However, patterns exist.

A pump may be more likely to fail if:

• it is old

• it operates under high load

• it has frequent past failures

AI models estimate the probability of failure.

Instead of asking:

“Will the pump fail tomorrow?”

The AI asks:

“What is the likelihood of failure given its age and operating conditions?”

This allows maintenance teams to move from:

reactive maintenance → predictive maintenance.

Step 5 – Graph Theory: Buildings as Networks

Buildings are not isolated systems.

Everything is connected.

For example:

Chiller → Pump → Air Handling Unit → Room

If a pump fails, cooling may be lost across multiple zones.

Graph theory is a branch of mathematics that studies networks of connected elements.

Using graph models, AI can:

• trace fault propagation

• identify root causes

• understand system interactions

This helps diagnose problems faster.

Step 6 – Digital Twins: The Mathematical Building

When engineering formulas, AI models, and sensor data are combined, we can build a digital twin.

A digital twin is a virtual representation of the building.

It continuously updates based on real data and predicts:

• energy performance

• equipment degradation

• future maintenance needs

Instead of reacting to problems, facility teams can anticipate them.

The Big Idea

AI in Facility Management is not just about algorithms.

It is about combining three elements:

1️⃣ Engineering knowledge
2️⃣ Mathematics used in AI
3️⃣ Real operational data

When these three elements come together, buildings become intelligent systems capable of:

• predicting failures

• optimizing energy use

• improving sustainability

• supporting better asset planning

Final Thought

The future of Facility Management will not be purely technical or purely digital.

It will be mathematical.

Facility managers who understand both engineering formulas and AI analytics will be best positioned to lead the next generation of smart buildings.

The AI Cannibalization Economy: When Every Generation of AI Eats the Last

Another important implication of the AI cannibalization cycle is its impact on the sustainability of Software-as-a-Service (SaaS) business models. Each generation of artificial intelligence tends to rapidly absorb and replicate the capabilities of the previous generation. Features that were once considered proprietary innovations are quickly incorporated into newer foundational models and platforms. As a result, the differentiation advantage of many AI-driven SaaS applications erodes rapidly. What previously required a specialized software product can often be reproduced by a newer generation of AI models integrated directly into broader platforms. This dynamic significantly compresses the traditional software business cycle. Instead of product lifecycles measured in years, AI-enabled services may experience competitive disruption within months. For SaaS companies whose value proposition relies primarily on algorithmic capabilities or data processing features, this creates a structural vulnerability. Unless such firms anchor their offerings in proprietary data, domain expertise, operational integration, or real-world infrastructure, their products risk being rapidly commoditized by successive waves of AI development. Consequently, sustainable competitive advantage in the AI era may depend less on the software itself and more on access to unique datasets, deep industry knowledge, and tightly integrated operational ecosystems.

A useful illustration can be seen in the rapid evolution of generative AI writing tools. Early applications built on large language models offered specialized services such as automated marketing copy generation, blog writing assistance, and email drafting. Several startups built SaaS products around these capabilities. However, when foundational models and productivity platforms such as OpenAI, Microsoft, and Google began integrating similar capabilities directly into widely used tools like Microsoft Word and Google Docs, the differentiation of many standalone writing assistants diminished rapidly. Functions that once required a separate subscription service could now be performed within mainstream productivity software. This illustrates how quickly AI-driven features can be absorbed into larger platforms, effectively shortening the competitive window for specialized SaaS products. Companies that relied primarily on access to generative algorithms found it increasingly difficult to maintain a sustainable advantage once those same capabilities became embedded in widely adopted digital ecosystems.

Breaking the AI Knowledge Loop: Why Organizations Must Anchor AI in Reality

Artificial Intelligence is widely perceived as a technology that improves continuously with scale. The prevailing assumption is that larger datasets and more sophisticated algorithms will naturally lead to better performance and deeper insights. However, a growing body of discussion among researchers and practitioners suggests that this development trajectory may contain a structural weakness. As AI systems become more widespread, increasing volumes of digital content are generated by AI itself. When new models are trained on large datasets that include AI generated material, the system may progressively learn from earlier machine outputs rather than from original human knowledge. This recursive learning process has been described as AI cannibalization or model collapse. Over time, the diversity, accuracy, and originality of knowledge embedded within AI systems may gradually decline.

The rapid proliferation of AI generated content across the internet has accelerated this concern. Articles, summaries, technical documentation, marketing material, software code, and visual media are increasingly produced by generative models. When future AI systems are trained on large internet datasets, a growing proportion of the training data may consist of machine generated content. This situation introduces several potential risks. First, knowledge diversity may decline because AI generated text tends to smooth over contradictions and uncertainties that are characteristic of human reasoning. Second, errors introduced by earlier models may propagate across successive generations of systems. Third, AI generated outputs often converge toward statistically averaged patterns, which may reduce the presence of unconventional or minority viewpoints. Finally, the accumulation of synthetic knowledge may reduce the system’s ability to generate genuinely novel insights. For organizations that rely on AI for operational analytics, decision support, and strategic planning, the implications are significant because the quality of insights produced by AI systems depends directly on the quality and diversity of the underlying training data.

Organizations that wish to benefit from AI while avoiding recursive self learning must therefore design systems that remain anchored in human knowledge and empirical observation. Artificial Intelligence should not be treated as a self contained intelligence engine but rather as one component within a broader knowledge ecosystem that integrates human expertise, operational data, and continuous experimentation. Three foundational elements are particularly important in this regard.

The first element is the preservation and systematic capture of human origin knowledge. Human expertise remains the most valuable source of contextual understanding, tacit judgment, and experiential learning. Many organizations possess large amounts of internal knowledge that are rarely curated or structured in ways that allow them to support AI learning. Maintenance reports, engineering design decisions, operational reviews, project documentation, and incident investigations all contain valuable insights derived from real experience. When these materials are systematically captured and organized, they form a rich dataset that reflects the complexity and nuance of real world operations. AI systems trained on such material can develop a deeper understanding of operational contexts, constraints, and decision processes that cannot easily be inferred from generic internet data.

The second element is the integration of real world operational data. Artificial Intelligence systems perform far more reliably when they learn from measured reality rather than from textual descriptions alone. Many modern organizations generate extensive operational datasets through sensors, monitoring systems, and digital infrastructure. Examples include energy consumption measurements in buildings, equipment telemetry from industrial machinery, environmental monitoring data, predictive maintenance vibration signals, and indoor air quality measurements. These datasets represent direct observations of physical systems and therefore provide empirical grounding for AI analysis. When AI models analyze such data streams, they remain closely connected to measurable operational conditions rather than drifting into purely synthetic knowledge domains. The combination of sensor networks, IoT infrastructure, and digital twin systems creates a powerful feedback mechanism that continuously refreshes AI models with new ground truth information.

The third element is the cultivation of continuous experimentation within the organization. AI systems remain effective only when they are supplied with new observations and validated outcomes. Organizations should therefore actively generate new knowledge through pilot projects, controlled trials, and operational experimentation. Examples include testing alternative energy optimization strategies, evaluating predictive maintenance approaches, experimenting with new operational workflows, or implementing technology trials in specific facilities. Each experiment generates new datasets that reflect actual system behavior under different conditions. These results enrich the organizational knowledge base and ensure that AI systems continue learning from real discoveries rather than from recycled or synthetic data.

Another important consideration involves the establishment of data provenance frameworks. Organizations should maintain clear visibility into the origin of datasets used for AI training and analysis. Training data can be broadly categorized into human authored content, operational measurement data, simulation datasets, and AI generated material. Each category has different characteristics and levels of reliability. Human authored and empirically measured data typically provide the most reliable knowledge foundations. Simulation datasets can be useful for exploring hypothetical scenarios or rare events that are difficult to observe directly. AI generated material can assist in certain modeling or scenario generation tasks but should not dominate the training process. Transparent data provenance helps organizations maintain confidence in the integrity and reliability of their AI systems.

Human oversight also remains an essential component of responsible AI deployment. Although machine learning algorithms can identify patterns and correlations across large datasets, they cannot fully replicate the contextual reasoning and ethical judgment of experienced professionals. Human experts play a critical role in validating AI generated insights, interpreting complex operational patterns, and ensuring that recommendations are feasible within real world constraints. Effective governance structures typically involve collaboration between domain experts, engineers, data scientists, and organizational leadership. Such human in the loop systems ensure that AI functions as an analytical support tool rather than an autonomous decision maker.

Organizations that successfully deploy AI at scale increasingly treat data as a form of strategic infrastructure. Just as physical infrastructure requires investment, maintenance, and governance, data ecosystems must also be carefully designed and managed. This includes establishing data governance frameworks, implementing data lineage tracking, developing knowledge repositories, and ensuring long term preservation of high quality datasets. Organizations that invest in robust data infrastructure are better positioned to develop AI systems that remain reliable, adaptive, and resilient over time.

Interestingly, sectors that operate physical systems may possess a natural advantage in avoiding AI cannibalization. Industries such as facility management, infrastructure operations, energy systems, manufacturing, and logistics generate continuous streams of operational data through sensor networks and monitoring systems. These industries can create a learning cycle in which real world measurements inform AI analysis, engineers validate the insights, operational improvements are implemented, and the resulting performance data becomes new training input for future models. This continuous interaction between physical systems and analytical models ensures that AI capabilities evolve alongside real world experience.

Artificial Intelligence should therefore be understood primarily as a knowledge amplification tool rather than an independent source of knowledge. The most effective organizational learning model can be described as a sequence in which human expertise generates knowledge, AI systems analyze and synthesize information, human experts validate the outputs, and real world feedback produces new datasets. This cycle ensures that knowledge continuously evolves rather than stagnates.

An analogy from agriculture provides a useful illustration of this concept. When farmers cultivate a single crop repeatedly without replenishing nutrients, soil quality gradually deteriorates and biodiversity declines. Similarly, a knowledge ecosystem that relies excessively on machine generated information risks becoming intellectually depleted. To maintain the health of the knowledge environment, organizations must introduce new insights, diverse perspectives, empirical observations, and experimental discoveries. This process can be viewed as a form of knowledge crop rotation that preserves the vitality and resilience of the intellectual ecosystem.

Artificial Intelligence will undoubtedly continue to transform industries and organizational practices. However, its long term effectiveness will depend on maintaining strong connections to human expertise and empirical observation. Organizations that design AI systems as part of an integrated learning ecosystem that combines expertise, operational data, experimentation, and governance will be able to avoid the trap of recursive self learning. Rather than becoming self referential systems, their AI capabilities will evolve alongside human discovery and real world experience. In the long term, the most valuable AI systems will not simply be those that process the largest quantities of data, but those that remain deeply grounded in how the world actually functions.

Designing a Resilient AI Nation: A 4-Driver Framework for Singapore

Singapore’s 2026 Budget signals a decisive national push into artificial intelligence. The government is investing in skills, enterprise adoption, infrastructure, and governance to ensure that AI strengthens economic competitiveness while protecting workers. But funding and ambition alone do not guarantee success. History shows that complex technological transformations fail not because of weak technology, but because of weak system design.

Two powerful lenses help explain why: Charles Perrow’s theory of complex system failure and David Hardoon’s concept of the “silent fracture” between strategy and execution. Together, they suggest that national AI success depends on architecture, not algorithms. Background on Perrow and Hardoon is found in Appendix. 

To translate these lessons into practice, Singapore should anchor its AI strategy on four structural drivers.

1. Structural Governance - Prevent Systemic Failure

AI must be governed like critical infrastructure, not treated as another IT tool. Strong oversight, auditability, and clear accountability reduce the risk of cascading failures in complex systems. National-level coordination bodies and sector-specific safeguards ensure that innovation does not outrun safety.

2. Workforce Adaptability - Prevent Social Instability

AI transformation is fundamentally a human-capital challenge. Training programs, mid-career pathways, and certification frameworks should be viewed as national infrastructure. A workforce that can adapt quickly reduces resistance, prevents displacement shocks, and increases execution capacity across industries.

3. Enterprise Enablement - Prevent Fragmented Adoption

Without structure, companies adopt AI unevenly, creating inefficiencies and hidden risks. Standardized toolkits, trusted solution libraries, and sandbox environments help firms deploy AI safely while controlling complexity. Adoption should be systematic, not experimental.

4. Ecosystem Coordination - Prevent National “Silent Fracture”

AI success depends on alignment across government, industry, academia, and society. Shared standards, interoperable platforms, and collaborative research environments prevent fragmentation and ensure that progress in one sector strengthens the whole ecosystem.

Why These Four Drivers Matter Together

Each driver protects against a different type of systemic risk:

Driver	Risk Prevented
Governance	Catastrophic failures
Workforce	Social disruption
Enterprise	Innovation fragmentation
Ecosystem	Strategic misalignment

Remove one, and the system becomes fragile.

The Strategic Insight

The global AI race will not be won by the country with the most models or the largest data centers. It will be won by the country with the most resilient AI ecosystem.

Singapore’s advantage is not size. It is system design capability. If it treats AI as a national systems-engineering challenge rather than a technology initiative, it can become one of the world’s most robust AI economies.

Closing thought

Robust AI is not achieved when systems never fail.
It is achieved when systems remain safe, stable, and trustworthy even when they do.

Appendix 

David Hardoon’s perspective highlights that most AI failures are not technical but organizational. His concept of the “silent fracture” describes the hidden gap between strategic ambition and operational capability. Organizations often invest heavily in AI tools yet lack aligned governance, clear accountability, skilled talent pipelines, and execution capacity. This mismatch leads to stalled projects, wasted resources, and leadership churn. Hardoon’s key insight is that successful AI adoption depends less on model sophistication and more on institutional readiness. In other words, AI transformation is fundamentally a systems-management challenge, not just a technology initiative.

Charles Perrow’s theory from Normal Accidents explains why complex technologies such as AI inevitably produce unexpected failures. Perrow argued that systems with high complexity and tight coupling will eventually experience breakdowns even when individual components work correctly. Failures arise from unpredictable interactions rather than single mistakes. Applied to AI, this means unintended behavior is not an anomaly but a structural property of advanced systems. His work emphasizes designing architectures that contain and recover from failure, rather than assuming perfect reliability. Together, Perrow’s framework shows that resilience must be built into the system’s design, not added afterward as a safeguard.

Summary of “Dollars and Sense of Safety” (1940)

In Dollars and Sense of Safety, F. J. Van Antwerpen argues that industrial safety should be viewed not only as a humanitarian responsibility but as a sound economic investment. Writing in 1940, the author challenges the belief that safety programs exist purely for moral reasons and demonstrates, using long-term industry data and case studies, that safety delivers substantial financial returns.

Drawing on accident statistics from as early as 1912 and longitudinal evidence from heavy and process industries, the paper shows that systematic safety programs reduce accident frequency, lost workdays, compensation payments, and insurance premiums. Van Antwerpen emphasizes that direct costs such as medical expenses and compensation represent only a fraction of the true economic burden of accidents. Indirect or hidden costs including lost productivity, supervisory time, production disruption, retraining, and material damage are estimated to be four to four-and-a-half times the direct costs.

Through multiple industry examples including chemical, steel, oil, and manufacturing firms, the paper documents reductions of 40 to 80 percent in accident-related costs following the introduction of structured safety and hygiene programs. While the chemical industry exhibits relatively low accident frequency, accident severity and fatality rates are higher, underscoring the importance of engineering design and hazard elimination rather than behavior alone. [Low probability, severe impact quadrant]

The paper concludes that investment in safety improves profitability, operational efficiency, and workforce stability, proving that good safety engineering saves both money and lives. 

Citation

Van Antwerpen, F. J. (1940). Dollars and sense of safety. Industrial & Engineering Chemistry, 32(11), 1437–1444.

https://doi.org/10.1021/ie50371a007

From Productivity to Purpose: Reframing AI Adoption and the Emergence of New Occupational Roles

 Artificial intelligence is presently deployed predominantly as a productivity-enhancing technology within existing occupational roles. Across sectors, AI systems are used to automate repetitive tasks, improve operational efficiency, reduce labour costs, and accelerate decision-making processes. These applications are typically embedded within established professions such as data analysis, engineering, operations management, finance, and marketing. The widespread adoption of AI in these contexts is neither legally nor morally problematic; rather, it reflects prevailing economic incentives that prioritise measurable returns on investment, scalability, and short-term efficiency gains. Consequently, AI adoption has largely reinforced existing organisational structures instead of challenging their underlying assumptions about work and value creation.

However, the current pattern of AI utilisation reveals a significant structural imbalance. While roles that leverage AI for productivity optimisation are well developed, there is a notable absence of formal roles responsible for addressing the broader societal, human, and systemic implications of AI deployment. In most organisations, decisions concerning automation are framed almost exclusively in terms of technical feasibility and economic efficiency. Questions regarding whether AI should be applied in specific contexts, particularly those with significant labour surpluses or social vulnerabilities, are rarely assigned institutional ownership. As a result, the consequences of AI adoption for job quality, human dignity, skill erosion, and social cohesion are often treated as secondary or external considerations rather than core design parameters.

This absence of responsibility becomes particularly consequential in labour-surplus and lower-income contexts, where the diffusion of labour-saving AI technologies may exacerbate unemployment rather than alleviate labour shortages. As highlighted in contemporary debates on economic inequality, technologies initially developed to address workforce deficits in high-income economies frequently migrate into regions where gainful employment, rather than automation, is the more pressing need. In such settings, AI systems may unintentionally displace formal employment and accelerate informalisation, thereby deepening economic precarity. Despite these risks, few roles exist that are explicitly tasked with adapting AI systems to local employment realities or evaluating their distributive impacts across different socioeconomic contexts.

The lack of dedicated roles also extends to the long-term systemic consequences of AI adoption. Current optimisation paradigms emphasise speed, accuracy, and cost reduction, yet often neglect resilience, trust, and intergenerational equity. AI systems may improve short-term performance while increasing long-term fragility by eroding human expertise, reducing organisational redundancy, or concentrating decision-making authority within opaque algorithms. In the absence of roles that explicitly prioritise resilience and societal value, these risks remain under-analysed and under-managed. This reflects not a failure of technology, but a failure of institutional design.

The emergence of these unaddressed gaps suggests the necessity for new categories of professional roles that extend beyond traditional productivity-oriented functions. Such roles would focus on defining the purpose of AI systems prior to their deployment, safeguarding human dignity within AI-mediated workflows, adapting technologies to diverse socioeconomic contexts, and ensuring that AI contributes to long-term societal resilience rather than short-term efficiency alone. Importantly, these roles do not arise from opposition to AI, but from recognition that technological capability must be matched by deliberate governance and human-centred design.

Fresh entrants to the labour market are uniquely positioned to contribute to the creation of these new roles. Because such positions sit at the intersection of technology, ethics, policy, and human systems, they are not easily claimed by established professions or legacy hierarchies. Rather than being constrained by predefined job descriptions, early-career professionals may identify emerging problems created by AI adoption and articulate roles that address these unmet needs. Historically, many now-established professions, such as sustainability management, data science, and cybersecurity, emerged in precisely this manner, following the recognition of systemic risks that existing roles failed to manage.

In this context, the future of work should not be framed solely in terms of job displacement or skill obsolescence. Instead, it should be understood as a period of occupational reconfiguration in which new forms of value creation become visible. While AI will continue to enhance productivity within existing roles, it simultaneously generates demand for new forms of human labour that are oriented toward judgment, contextual understanding, ethical stewardship, and social adaptation. The capacity to invent such roles, rather than merely occupy predefined ones, represents a critical source of agency and opportunity for the next generation entering the workforce.