IT work in the age of AI and Gen Z: a double silent revolution

For at least two years now, we have witnessed the progressive establishment of Generative AI-based tools as daily support across all professions. According to the McKinsey & Company annual report (2024), 65% of global organizations report regularly using generative AI tools, a percentage that has nearly doubled compared to the previous year. In companies that design and develop IT solutions, this transformation is particularly visible and radical.

We are not witnessing a replacement of IT work. We are witnessing its profound reconfiguration. And this reconfiguration is intertwined with a second, equally powerful structural change: the massive entry of Generation Z into the labor market, with values, expectations, and operating modes radically different from those of previous generations.

The end of repetitive cognitive IT work

The most immediate and visible change concerns the very nature of the daily activities of software developers. Activities traditionally central to IT work—writing standard code, producing technical documentation, performing regression tests, systematic debugging—are today largely assisted or directly automated by tools such as GitHub Copilot, Amazon CodeWhisperer, and agentic systems based on Large Language Models (LLMs).

A GitHub study (2023) found that developers using Copilot complete tasks up to 55% faster than those who do not. Similarly, a Microsoft Research study (2024) documents how the use of AI assistants in coding significantly reduces time spent on repetitive tasks, freeing up cognitive resources for higher-level activities.

The critical point, however, is not speed. It is the type of value that remains a human prerogative. As Autor (2024) observes in a recent work on the future of work in the AI era, automation tends to erode routine activities—even cognitive ones—rather than adaptive, relational, and contextual skills. In the IT field, this means that a developer’s value is no longer measured in lines of code produced, but in the quality of the problem they can define, the ability to critically evaluate AI output, and the architectural design of complex systems.

The real change is not that fewer developers will be needed. It is that different developers will be needed.

Skill polarization and the disappearance of mediocrity

One of the most profound and least discussed effects of AI in IT work is the polarization of skills. In previous labor markets, there was a relatively homogeneous distribution of skills: junior, mid-level, senior. This hierarchy is now in a structural crisis.

A bimodal structure is emerging: on one side, highly specialized figures in software architecture, AI orchestration, system design, and the security of complex systems. On the other, figures with general skills but strongly empowered by AI—capable of producing higher quality output than their training would suggest thanks to the skillful use of tools. In the middle, what we might call the level of “un-updated average competence”—traditional technical profiles that have not integrated AI into their way of working—tends to lose competitive relevance.

This phenomenon is consistent with what was documented by the World Economic Forum in the Future of Jobs Report 2025, which identifies “skill obsolescence” as one of the main risks for knowledge workers over the next five years. It is not about losing one’s job to AI, but about losing competitiveness compared to—human—colleagues who use AI better.

AI does not eliminate work: it eliminates un-updated mediocrity.

From coding to problem solving: the new geography of value

The most profound consequence of the ongoing transformation is a shift in the center of gravity of professional value. For decades, in IT companies, the ability to write efficient and correct code was the primary yardstick for evaluating a technical professional. This paradigm is rapidly dissolving.

What becomes rare and precious is the ability to define problems with precision, articulate context so that AI can operate effectively, critically evaluate generated solutions, and ensure the systemic quality of the final result. In other words, IT work is shifting from syntax to semantics, from execution to direction.

This change is well captured by the concept of “AI fluency” proposed by Mollick (2024) in Co-Intelligence. It is not about knowing how to program better thanks to AI, but about developing a new form of hybrid intelligence—human and artificial—where the boundary between the developer’s thought and the model’s output becomes deliberately permeable. The most effective professionals in this scenario are not the best programmers: they are the best collaborators with AI.

IT work will no longer be ‘writing software’, but ‘making intelligent systems work’.

Implications for HR and Organizations

This transformative scenario has direct and urgent implications for the HR functions of IT companies. Three, in particular, deserve attention.

Continuous upskilling as infrastructure, not a project

Episodic technical training—the course, the certification, the annual workshop—is no longer sufficient. In a context where AI tool capabilities evolve every six to twelve months, upskilling must become an embedded process in the organizational culture. IBM Institute for Business Value (2023) estimates that the average time needed to retrain a knowledge worker has dropped from three years to less than one thanks to AI-powered learning tools—but only in organizations that have structured continuous and adaptive learning paths.

New emerging roles

The IT market is already producing new professional figures that did not exist until a few years ago. Among the most significant: the AI Orchestrator (responsible for integrating and coordinating AI agents in complex work pipelines), the Prompt Strategist (an evolution of the prompt engineer, with broader responsibilities for human-machine dialogue design), and the Model Evaluator (a specialist in evaluating the quality and reliability of AI outputs in a production context). These figures are not yet consolidated in corporate job descriptions, but they are emerging in the practices of the most advanced organizations.

Rethinking talent evaluation

If professional value shifts from the ability to write code to the ability to think in systems and collaborate with AI, traditional IT talent selection and evaluation tools—the technical CV, the coding test, the list of mastered languages—lose much of their predictive power. What becomes central is adaptive capacity: how the candidate learns, the speed at which they update their work model, and how they handle the uncertainty and ambiguity of partially autonomous systems.

The Double Accelerator: AI and Gen Z

As analyzed in the previous article dedicated to the ongoing generational leap, Generation Z does not only bring new digital skills: it brings a different conception of work, organizational authority, and professional meaning. For Gen Z, work is a tool for identity and growth, not just a source of income. Flexibility is not a benefit: it is a prerequisite. Transparency and continuous feedback are not virtuous HR practices: they are basic expectations.

AI and Gen Z are acting simultaneously as accelerators of change on different but complementary dimensions. AI accelerates the change of work: it redefines what is done, how it is done, and what value it has. Gen Z accelerates the change of organizations: it redefines why we work, in what context, and with what rules. The combination of the two vectors produces a multiplicative effect that many companies are still underestimating.

The real shock is not technological. It is not generational. It is the combination of the two.