Design for tomorrow’s work: talent strategy in the age of AI

Artificial intelligence is no longer an emerging technology. It is now woven into everyday organisational life - screening CVs, drafting reports, analysing data and supporting customer service. The dominant question remains: Which jobs will AI replace? It is understandable, but it is also the wrong place to start.

A more strategic question is this: What will AI actually be used for, and what human skill is it replacing, reducing or making more valuable?

Until leaders answer that, investments in training, job design and workforce planning risk being misaligned. Technology rarely removes work outright. It reshapes which tasks matter and where human judgement creates value.

We have seen this pattern before. In manufacturing, robotic welding replaced a highly skilled manual trade, reshaping entry pathways and professional identity. By contrast, computer‑aided design did not remove designers. It reduced repetitive drafting and enabled more creativity, iteration and precision. One technology displaced a defining skill; the other elevated a different one.

AI presents leaders with the same choice. It can be deployed to narrow human roles, or to shift people toward work requiring interpretation, judgement and creativity. The outcome depends on how deliberately organisations analyse what is changing.

Start with work, not job titles

Most workforce planning still operates at the level of jobs. But AI does not replace whole occupations in one step. It changes tasks unevenly.

Research by Autor and the OECD shows that automation affects tasks more than jobs. Routine, predictable activities are most exposed. Work involving judgement, social interaction and accountability is more resilient.

This means leaders need to begin with work itself. Break critical roles into tasks and ask:

• Which tasks can AI automate reliably?
• Which tasks can AI augment, where humans still decide but with better information?
• Which tasks must remain human because they involve ethics, trust or responsibility?

Without this clarity, organisations risk over‑automating and hollowing out roles, or under‑using technology and preserving low‑value work.

A modern example: recruitment

Recruitment illustrates this shift. AI tools can now screen CVs, analyse video interviews and predict job fit at scale. Used poorly, they simply replace human screening with opaque algorithms and risk reinforcing historical bias. Used well, they remove repetitive filtering and free recruiters to focus on relationship building, assessment quality and inclusion.

Algorithms excel at spotting patterns that resemble past success by matching keywords, credentials and career paths to an existing model of a “good candidate.” Humans, however, are better at recognising potential that falls outside the pattern.

A recruiter can see the career changer with transferable experience, the applicant whose path shows adaptability or resilience, the “square peg” who may be the innovation the organisation needs.

AI is not replacing that core professional skill. It is shifting the role from administrative sorting to judgement about potential, context and contribution. The task that shrinks is volume screening. The task that grows is interpretation.

The strategic question is not whether recruitment jobs will disappear, but what kind of recruiter the organisation is designing for.

The pattern repeats across professions

Across domains, AI reduces volume processing while elevating human judgement:

• In law, AI reviews contracts; lawyers interpret risk and advise clients.
• In healthcare, algorithms flag anomalies; clinicians decide treatment and communicate with patients.
• In finance, systems model scenarios; humans judge what is acceptable and ethical.
• In policy and research, models draft material; people assess evidence, context and consequences.

The disappearing skill is mechanical handling of information. The growing skill is deciding what the information means.

Skills become the real unit of strategy

Once work is understood at task level, skills become the foundation of talent strategy. Job titles conceal capability. Skills reveal it.

McKinsey’s research shows that most people transition into new roles through adjacent skills, not wholesale retraining. This matters for AI adoption. The goal is not to turn everyone into technical specialists, but to extend existing expertise with new tools.

An analyst may need stronger critical thinking and ethical reasoning alongside AI literacy. A customer service adviser may need emotional intelligence and conflict handling alongside system navigation. The technology changes the mix, not the need for human capability.

Training should therefore follow task redesign, not precede it. Teaching people how to use tools without redefining what good work looks like simply adds complexity.

Learning becomes organisational infrastructure

If tasks and skills are shifting continuously, learning cannot remain episodic. It must become part of how work is done.

Research on learning organisations shows that capability develops most effectively through application, feedback and reflection. Edmondson’s work on psychological safety demonstrates that people learn best when they feel able to experiment and admit uncertainty.

For leaders, this means:

• embedding learning into real work
• linking short learning to live projects
• enabling peer learning where expertise already exists
• protecting time for reflection and adaptation
• modelling learning themselves

AI adoption often fails for cultural reasons, not technical ones. When leaders treat AI as a tool for others to master rather than a shift they must also navigate, trust erodes.

Structure still matters

It is tempting to place responsibility for adaptation on individuals. But organisational design is equally important.

Rigid job boundaries and slow redeployment of skills prevent organisations from using human capability well. Hybrid human‑and‑AI roles require flexible team structures and clear accountability.

This is a socio‑technical design challenge, not just a training one.

Work must be reorganised so that judgement, creativity and relationship‑based tasks are not squeezed out by efficiency metrics designed for machines.

Ethics, identity and meaning

Work is not just a set of tasks. It is a source of identity and belonging. Poorly managed technological change increases anxiety and resistance even when productivity rises.

Leaders need to be transparent about what is changing and why. They must invest in those whose tasks are most affected. Organisations that treat AI purely as a cost‑reduction tool risk losing trust and institutional knowledge. Those that treat it as a redesign opportunity can preserve dignity while improving performance.

As machines become more capable, human qualities become more central. Empathy, judgement and creativity are harder to codify but more important to organisational life.

A practical sequence for leaders

A coherent talent strategy for the age of AI follows a simple logic:

1. Start with strategy. What work will matter most in the coming years?
2. Analyse tasks. What will AI automate, augment or leave to humans?
3. Map skills. What capabilities does this redesigned work require?
4. Redesign roles and teams. How will humans and AI collaborate?
5. Build learning systems. How will people develop these skills?
6. Model leadership behaviour. What norms will support adaptation?
7. Review continuously. How will assumptions be tested and updated?

This is not a one‑off project. It is a cycle. AI does not remove the need for human judgement. It raises the standard for how deliberately that judgement must be used.