The Agentic Organisation, Job Displacement and the Great Skills Recomposition

By Dr Luke Soon

We are no longer asking whether AI will change work.

We are now confronting how deeply it will reshape it — and whether we are prepared for the velocity of that change.

Across boardrooms, research labs and ministries of manpower, a new consensus is forming: AI will not simply augment work. In many domains, it will compress, substitute and recompose it.

The question is not technological.

It is structural.

And increasingly, it is societal.

The Labour Data: Early Signals of Recomposition

PwC’s AI Jobs Barometer has been tracking how AI exposure affects labour markets. Several themes are emerging:

AI-related job postings are growing significantly faster than overall job demand. Roles exposed to AI are seeing measurable productivity uplift. Many AI-exposed occupations command wage premiums. AI diffusion is broad-based — across finance, healthcare, legal, manufacturing and public services.

The World Economic Forum’s Future of Jobs Report echoes this, highlighting that analytical thinking, systems thinking and AI literacy are rapidly rising skills. At the same time, nearly half of workers globally will require reskilling within this decade.

Importantly, both bodies note that AI is currently reshaping tasks within occupations rather than eliminating whole professions overnight.

But we must look beyond first-order effects.

Generative and agentic AI introduce second-order dynamics: they reduce coordination friction, accelerate knowledge synthesis and enable semi-autonomous workflows.

That changes the equation.

When AI Pioneers Warn of Job Replacement

We should pay attention when the architects of these systems speak candidly.

Dario Amodei, CEO of Anthropic, has suggested that advanced AI systems could automate a very large proportion of routine white-collar cognitive tasks. He has warned that many knowledge-economy roles are structurally exposed.

Geoffrey Hinton has expressed concern that AI may replace more jobs than it creates in cognitive domains — particularly those involving pattern recognition, analysis and summarisation.

Mo Gawdat has argued that society is dramatically underestimating the speed of AI’s improvement curve. In his view, roles in customer support, administration, content generation and structured analysis are highly vulnerable.

Demis Hassabis takes a more optimistic lens, emphasising AI’s potential to accelerate scientific discovery and solve grand challenges. Yet even he acknowledges that large-scale deployment will reshape labour markets in profound ways.

The consistent theme is not dystopia.

It is speed.

AI capability is compounding exponentially, while institutional response remains incremental.

From Augmentation to Agency

Much of the current workforce conversation assumes AI as a productivity tool.

But we are entering the era of agentic AI:

Systems that plan multi-step workflows Agents that interact with other agents Autonomous task allocation Self-monitoring and exception escalation

In such environments, AI becomes an operational layer.

When coordination friction reduces, entire layers of middle management and reporting structures become economically redundant.

This does not mean wholesale unemployment.

It does mean structural compression.

The Vulnerable Zone: Coordination Roles

The roles most exposed are not manual labourers.

They are professionals whose core activities involve:

Synthesising structured information Producing routine reports Coordinating workflows Providing intermediary analysis Managing structured processes

PwC’s data shows productivity gains in AI-exposed roles. But productivity gains create strategic tension:

If output per worker rises materially, organisations must decide whether to redeploy capacity or reduce headcount.

This is where displacement risk emerges.

Singapore’s Response: Preparing for Structural Change

Singapore has been notably proactive.

National AI Strategy 2.0 and NAIC

Under the refreshed National AI Strategy (NAIS 2.0), Singapore is positioning AI as foundational national infrastructure. The establishment of governance frameworks and AI oversight bodies — including the National AI Council (NAIC) chaired by Prime Minister Lawrence Wong — reflects a recognition that AI transformation is not purely technological, but systemic.

Singapore’s AI governance approach emphasises:

Trusted AI deployment Public sector AI adoption Enterprise enablement Talent development pipelines Responsible AI verification frameworks

The underlying philosophy is clear: economic growth from AI must be coupled with trust and societal stability.

SkillsFuture: Lifelong Adaptation

SkillsFuture has evolved from general upskilling into targeted AI capability building:

AI-related training subsidies Mid-career conversion programmes Industry-aligned certifications Workforce digital literacy initiatives

The shift is from static education to continuous re-skilling.

Workforce Singapore (WSG)

WSG’s role has expanded in:

Career transition programmes Job redesign support for employers Funding for workforce transformation Sectoral manpower planning

WSG recognises that job redesign — not just reskilling — is central.

Singapore is effectively attempting to manage the recomposition of work before it becomes destabilising.

But even Singapore faces constraints.

Technology adoption can leapfrog institutional training cycles.

The Five Emerging Archetypes in the Agentic Organisation

In this environment, the future workforce clusters around five archetypes:

1. The Orchestrator

Designs human-AI systems across domains.

2. The Domain Guardian

Supervises AI outputs and manages exceptions.

3. The Trust Architect

Ensures governance, explainability and ethical compliance.

4. The Augmented Specialist

Delivers higher output through AI leverage.

5. The Human-Centric Creator

Provides empathy, ethics, meaning and relational value.

Notice what diminishes: roles dependent solely on coordination and reporting.

Productivity Gains and Social Stability

PwC’s broader economic modelling suggests AI could materially lift GDP if responsibly deployed.

But productivity growth without workforce transition planning can lead to:

Wage polarisation Skill stratification Middle-layer compression Psychological insecurity

Geoffrey Hinton has raised concerns that retraining may not fully absorb displaced workers.

Mo Gawdat has warned that universal income discussions may accelerate.

Amodei has emphasised the importance of alignment and controlled deployment.

Hassabis highlights the need for international cooperation and governance.

Singapore’s approach — combining AI growth with SkillsFuture, WSG and national governance frameworks — may represent one of the more coherent national responses.

But it must scale with AI velocity.

The Human Experience Imperative

As AI becomes embedded in workflows, employee experience becomes critical.

In my own framework — HX (Human Experience) = CX + EX — the success of AI transformation depends on:

Psychological safety Transparent governance Fair capability pathways Dignified transitions

AI adoption without trust will stall.

AI adoption with equitable design can uplift.

Trust is no longer a moral variable.

It is an economic multiplier.

The Strategic Question

The question leaders must ask is not:

“How many jobs will AI replace?”

It is:

“How do we redesign organisations when intelligence becomes scalable?”

That requires:

Systems thinking Organisational flattening Embedded governance Workforce transition planning Exponential mindset

The Decade Ahead

We stand at the threshold of the most cognitively disruptive technological wave in modern history.

If managed well:

Productivity rises New domains of growth emerge Work becomes more creative and meaningful

If mismanaged:

Trust erodes Skill polarisation deepens Social stability strains

Singapore’s early moves — NAIC leadership, SkillsFuture expansion, WSG job redesign support — signal that governments can act ahead of the curve.

But the window is narrow.

The agentic organisation is not a future abstraction.

It is forming now.

And the real work of leadership is not predicting disruption.

It is designing resilience before disruption becomes destabilisation.

The future of work is not about replacing humans.

It is about redefining human value in a world where intelligence itself is abundant.

And that responsibility rests with all of us.