Software engineering is frequently cited as an occupation that AI will soon take over. It is often regarded as one of the fields most exposed to AI and not only that, as shown in the well-known Anthropic spider chart, it is an area where AI adoption - not just exposure - is exceptionally high.
So is AI really the end for Software Engineering?
Using online job postings data from Lightcast, we can actually start learning a little bit more about what is actually happening to this occupation.
First, let’s look at demand. Over the past year, software engineer job postings averaged 53k per quarter. Additionally, roles requiring software engineering as a skill averaged over 165k per quarter. Although demand has dipped since its 2022 peak, the field is far from being taken over by AI.
Second, consider salary, which can serve as a proxy for how employers value a role. A decline in salary for software engineers would likely indicate a cooling labour market, however, that is not happening here. The median advertised salary for software engineering—both as a specific role and as a required skill—has risen steadily over the past decade and currently exceeds $150k per year.
So should software engineers disregard AI claims as scaremongering? While AI won't wipe out software engineering, it is transforming the role. Key tasks like system performance monitoring are highly exposed to AI, so it is not surprising the required skill set is shifting.
Looking at the top 20 skills since 2015:
Enduring: 12 skills remain from 2015, with 10 becoming increasingly vital (e.g., communication rose from 33% to 42%).
Declining: Two original skills - both programming languages - are still in the top 20, but becoming less prominent.
Exiting: 7 of the 8 skills that dropped out were specific technical tools or languages, highlighting how specific technical requirements change while the core role remains.
Emerging: The new top 20 includes modern tools like AWS and APIs, plus crucial new competencies like AI knowledge, automation, and scalability.
Software engineering isn’t dead, but it is changing, and workers and organizations that understand this can better prepare for what is coming ahead. While the future of work is by definition uncertain, unpacking roles into their specific tasks and skills components allows for a more granular understanding of what is happening to an occupation and better preparation for the changes ahead.
