In a world where workflows evolve every quarter and AI-driven updates are constant, human capabilities are eroding faster than most employers realise. Image: freepik.com.
Dan Schawbel fears that in a constantly changing, AI-driven workplace, human workers will become specialists in adaptation, learning just enough to function but not enough to achieve mastery over the machines they supervise.
For the past several years, the dominant workforce narrative has been about upskilling.
Organisations are investing in training programs, AI literacy workshops, and future-focused capability models.
Employees are encouraged to learn faster, adapt quicker, and expand their technical fluency. The message is clear: Acquire new skills or risk falling behind. However, there’s a quieter, less-discussed issue unfolding beneath the surface — skills decay.
In a world where workflows evolve every quarter and AI-driven updates are constant, existing capabilities are eroding faster than most employers realise.
The problem is not just that employees need new skills. It’s that the skills they already have can become partially obsolete within months. The half-life of expertise is shrinking, and organisations are not prepared for what that means.
Historically, professional skills remained relevant for years, sometimes decades. Software platforms evolved gradually, and business processes changed incrementally.
Today, AI upgrades, automation tools, and digital transformation initiatives are accelerating change.
A finance analyst who built complex spreadsheets may see predictive modelling tools generate outputs in seconds. A software developer who specialised in manual coding may now spend more time reviewing AI-generated code than writing it.
Employees may not notice immediately, but over time, the depth of expertise declines. Skills that are not regularly exercised weaken — just like muscles.
This erosion is not dramatic. It is gradual, quiet, and often invisible until it becomes consequential.
AI does not simply add tools to workflows; it redistributes cognitive effort. When automation handles routine tasks, employees spend less time practising foundational skills.
While this can increase efficiency, it also reduces repetition — the mechanism through which mastery is maintained.
For example, professionals who rely heavily on AI for writing, coding, analysis, or design may find that their independent ability to perform those tasks diminishes over time.
Judgment and oversight remain essential, but execution muscles weaken.
This does not mean AI is harmful. It means organisations must recognise a trade-off: Without deliberate practice, even high performers can experience skills drift.
Most workforce strategies emphasise acquiring new skills, particularly those aligned with emerging technologies, but adding capabilities does not automatically protect existing ones.
An employee might complete an AI certification while simultaneously losing proficiency in foundational processes they no longer perform regularly. Over time, this creates a subtle imbalance — broader knowledge, but shallower depth.
Skills decay becomes especially problematic when systems fail or when human judgment must override automation. If employees have not maintained core competencies, their ability to step in confidently may be compromised.
Skills decay is not only a technical issue; it is psychological. In fast-changing environments, employees often feel they are perpetually catching up.
This perception erodes confidence. Workers may hesitate to take risks or contribute ideas if they question their own expertise.
In some cases, fear of obsolescence drives overwork as employees attempt to compensate through longer hours rather than targeted learning. When skills feel unstable, career identity becomes unstable — and when identity feels uncertain, engagement suffers.
One of the defining characteristics of AI-driven workplaces is the frequency of change. New platforms are introduced, processes are automated, dashboards are redesigned, and reporting structures evolve.
While agility is essential, constant reconfiguration disrupts skill reinforcement. Employees barely master one system before it changes. Familiar routines are replaced before they become deeply embedded.
This creates a workforce that is adaptable but rarely fluent. People learn just enough to function, but not enough to achieve mastery. Over time, shallow proficiency becomes the norm.
For example, an AI tool may produce polished presentations, predictive forecasts, or detailed reports that appear impressive. Yet the employee overseeing the output may not fully understand the methodology behind it.
As long as outcomes appear satisfactory, the decline remains hidden, but this creates organisational fragility. Capability gaps may surface suddenly.
In previous eras, expertise was associated with deep knowledge in a specific domain. In today’s environment, expertise increasingly includes adaptability, systems thinking, and critical evaluation of AI outputs.
However, depth still matters. The challenge is maintaining domain knowledge while integrating new tools. Organisations must balance agility with stability, ensuring employees retain foundational understanding even as workflows evolve.
Expertise in the AI era may be less about memorising processes and more about understanding principles, but without consistent reinforcement, even principle-based knowledge can erode.
Most organisations track training participation, certification completion, and new skill acquisition. Few measure skill retention or decay.
There are rarely systems in place to assess whether employees maintain proficiency in core capabilities over time.
Addressing skills decay requires a shift in mindset. Instead of viewing learning as a one-time event, organisations must treat it as a continuous loop of acquisition, application, and reinforcement.
Deliberate practice must become part of the workflow. This might include rotating responsibilities, encouraging manual problem-solving before automation, or periodically testing foundational knowledge.
The goal is not to reject AI but to ensure humans remain capable partners within AI-enabled systems.
If organisations want to remain resilient, they must address not just what employees need to learn next, but what they cannot afford to forget.
Dan Schawbel is a bestselling author and managing partner of Workplace Intelligence, a research and advisory firm helping HR adapt to trends, drive performance and prepare for the future. This article is part of his Workplace Intelligence Weekly series.
