Beyond execution: Why the AI revolution is scaling the human premium

One of the most persistent assumptions about AI in the workplace is that as the technology grows more capable, human judgment matters less. Tan Toi Chia, Chief of People, Organisation and Communications at StarHub, believes the opposite is happening.

“As AI makes execution easier, the real differentiator becomes judgment – knowing when to trust AI, when to challenge it, and how to apply context,” he tells HRM Asia.

The logic is straightforward: when producing an output becomes cheap, deciding what to do with that output becomes the scarce skill. AI can draft, summarise and generate at scale, but it cannot determine whether a result is right for the situation, the customer or the moment. “AI can generate outputs, but judgment, context and accountability remain fundamentally human,” Tan says.

At StarHub, that reasoning has sharpened the focus on two attributes: curiosity and agency. Curiosity, in Tan’s framing, drives people to challenge assumptions, explore new possibilities and continually learn. Agency is the willingness to take ownership and shape outcomes rather than wait to be told what to do.

The Singapore telco is not alone in elevating these traits. The World Economic Forum’s Future of Jobs Report 2025, drawing on responses from more than 1,000 employers worldwide, lists curiosity and lifelong learning among the skills expected to rise most in importance through 2030 – alongside creative thinking, resilience, flexibility and agility – and projects that 39% of employees’ core skills will change over the same period.

For Tan, these attributes come with a corollary that reward systems will need to reckon on: AI gives every employee greater capability, and as capability grows, responsibility grows with it. An employee who can now produce in hours what once took days is also the person accountable for whether that output should be used at all.

The measurement challenge is real – capabilities such as judgment rarely show up neatly in output metrics. StarHub’s response has been to widen the lens of evaluation. “We’re placing greater emphasis not just on what people deliver, but how they think, collaborate, exercise judgment and create value,” Tan says.

Defining roles by outcomes, not tasks

Asked what has changed in StarHub’s skills frameworks compared with two or three years ago, Tan points to something more fundamental than an updated competency list.

“The biggest shift isn’t simply adding new skills. It’s changing how we define work,” he says. StarHub is moving away from describing roles through the outcomes they are accountable for. The distinction matters because tasks are precisely what AI is absorbing; outcomes are what remain stubbornly human.

Functional expertise still counts, Tan stresses, but the organisation is increasingly building capability that travels – skills that can be applied across different roles, teams and business challenges, rather than preparing people for one static job.

Customer-facing roles offer the clearest illustration. Employees in these positions now spend less time retrieving information and more time resolving complex customer issues, building relationships and exercising discretion. “The opportunity isn’t simply to make work more efficient,” Tan says. “It’s to free people to solve more complex problems, build stronger relationships and contribute in higher-value ways.”

There is an optimistic vision of this story in which every employee glides “up the value chain” into higher-judgment work. Tan was careful not to tell it – particularly for employees whose roles have largely consisted of execution.

“Moving up the value chain doesn’t happen simply because people have access to AI,” he says. “It happens when organisations redesign how work happens and intentionally build new capabilities.”

StarHub’s approach centres on capability building rather than tool adoption, anchored in what the organisation calls The StarHub Way – a set of behaviours encouraging employees to stay curious, take ownership, experiment responsibly and continuously learn. Tan argues these behaviours matter even more in an AI-enabled environment, where technology can supply answers but people remain answerable for decisions.

READ MORE: In South-East Asia, AI moves at the speed leaders set – not the speed tools arrive

Notably, the destination he describes is modest. “The goal isn’t to make everyone an AI expert,” he says. “It’s to build a workforce that knows how to question outputs, apply context and exercise judgment responsibly.”

That framing quietly reorders the typical upskilling agenda. Rather than racing to certify employees in prompt engineering or data science, the priority becomes something closer to critical literacy: knowing when an AI-generated answer is good enough, when it is subtly wrong, and when it should be discarded.

The three-to-five-year horizon: Work design over hype

Looking past immediate industry hype toward a three-to-five-year horizon, Tan envisions a landscape where AI ceases to be viewed as a discrete, standalone tool. Instead, it will fade into the background architecture of daily work, becoming as ubiquitous and invisible as email or web search engines.

In this mature state of augmentation, routine administrative tasks, first-draft generation, and preliminary research will be fully automated. The remaining work – and the work that ultimately drives enterprise performance – will consist of human empathy, strategic creativity, relationship management, and nuanced decision-making.

“What many organisations underestimate is that this isn’t primarily a technology transformation. It’s a work design transformation. Technology scales quickly. People and capability don’t,” he concludes. “Many of the capabilities that create the greatest value, such as judgment, trust and collaboration, don’t appear neatly on a balance sheet. But when they’re missing, the impact on customers and business performance becomes immediately visible.”

When enterprise systems fail or output errors breach client trust, the root cause is rarely the underlying algorithm; it is almost always a breakdown in human oversight and context application. As organisations navigate the next era of workplace evolution, those that focus strictly on software procurement will quickly plateau. The organisations that thrive will be those that treat AI integration as a fundamental work design transformation, investing deeply in the judgment, agency, and accountability of their people.