Work Intelligence in Practice: How Software Engineering Is Changing
Understanding individual roles is essential, but jobs don't exist in isolation. They sit within job functions, organizations and wider labor markets. Software engineering is just one example of how AI and other forces are reshaping work.
The Lightcast Team

AI is reshaping work, but its impact is more complex than jobs simply growing or disappearing. To understand what is really changing, we need to look beyond jobs to the tasks and skills that make them up.
This is the central argument behind the Work Intelligence white paper we released recently: understanding the future of work requires looking at three dimensions together—jobs, tasks and skills.
In this blog, we apply this approach to software engineering to show why it matters in practice—and the insights it can unlock.
Demand for software engineers is falling while salaries are rising
On the surface, it may look like software engineers have been hard hit by AI, but looking at occupations alone is only giving us a partial picture of what is happening. While hiring demand fell by 46% between 2022 and 2025 for software engineers, over the same period, median advertised salaries rose by 21%, reaching $146,000.
Together, these figures paint a seemingly contradictory picture: employers are looking for fewer software engineers, but are willing to pay more for them.
Looking within the job—at the tasks software engineers perform and the skills they need—can help explain this apparent contradiction.
Software engineers spend 70% of their time on tasks with high AI potential
Tasks reveal what work actually gets done within an occupation. Software engineers divide their time across more than 15 tasks, with analyzing and monitoring system performance metrics (17%) and developing and integrating computer systems solutions (16%) taking up the largest shares.
Different tasks have different AI potential—and for software engineering, that potential is high. Software engineers spend approximately 70% of their working time on tasks highly exposed to AI. In fact, their five largest tasks by time spent all have high AI potential—meaning an LLM, alone or through AI-powered applications, could reduce the time needed to complete them by at least 50%.
Combined with declining hiring demand, it might be tempting to conclude that AI is rapidly taking over software engineering and putting these jobs at risk.
But AI potential is not the same as AI automation. Just because AI can accelerate a task doesn't mean the entire task—or the job itself—will be automated.
This is where skills come in. Looking at how skill requirements are evolving helps us understand how the role itself is changing.
Software engineering roles are declining in demand but broadening in scope
Despite falling hiring demand, software engineers are being asked to do more. In 2026, the average software engineering job posting required 29 skills—up 9% from 2022.
This suggests that the role is changing not just in demand, but also in scope. Looking at how skill requirements have evolved helps reveal the nature of that change.
While the core of software engineering remains relatively stable, the skill mix is shifting. Software engineering, computer science and software development remain the three most frequently requested skills. Beyond these, however, several changes stand out:
Skills moving up: Python, debugging, APIs and Amazon Web Services (AWS).
Skills moving down: Agile methodology, Java, JavaScript and SQL.
New entrants to the top 20: Workflow management, artificial intelligence, CI/CD, pipelines, Kubernetes and operations.
Skills leaving the top 20: Angular, Linux and full-stack development.
These shifts suggest that software engineering is evolving beyond developing code alone, with growing emphasis on managing workflows, infrastructure and increasingly complex systems. At the same time, skills moving down the rankings are not necessarily disappearing. They may simply be becoming less prominent relative to other requirements.
Taken together, these changes point to job enrichment. Software engineers face lower hiring demand, but broader and evolving skill requirements.
Looking only at occupation-level demand would have missed much of this transformation.
Understanding job redesign requires connecting the macro and the micro
Software engineering is just one example of how AI and other forces are reshaping work. While it sits at the center of the debate about AI's impact on employment, the same questions apply across the labor market: How are tasks changing? Which skills are becoming more important? And how are occupations evolving as a result?
Understanding individual roles is essential, but jobs don't exist in isolation. They sit within job functions, organizations and wider labor markets.
To prepare for change, we need to connect the micro to the macro: understand how individual roles are evolving, identify broader patterns across job functions and determine where organizations should prioritize their efforts.
This is where a skills-powered approach becomes essential. By breaking jobs down into their underlying tasks and skills, we can better understand how work is changing—and build more adaptable strategies for what comes next.
Learn more about Work Intelligence by downloading the report below.
Recent Posts

Redesigning Work: Are Business Leaders Asking the Wrong Question?

Carpenters’ Salaries in the US: Pay and Purchasing Power by State
