Researchers outline how engineers must adapt to AI
As artificial intelligence rapidly reshapes technical roles, researchers and industry experts are defining the specific adaptability skills engineers need to survive the transition.
The rapid integration of artificial intelligence is forcing a massive shift in technical roles, leaving educators and employers scrambling to define what adaptability means for modern engineers. According to a June 2026 report by PwC, the technology, media, and telecom sectors are experiencing the fastest pace of skill turnover. Furthermore, the World Economic Forum’s 2025 Future of Jobs Report revealed that employers across all sectors expect 39 percent of workers' core skills to change by 2030. To address this, researchers are urging a shift from teaching specific coding languages to fostering broader problem-solving capabilities.
Samantha Brunhaver, an associate professor at Arizona State University, received a National Science Foundation award in 2020 to study how to foster workplace adaptability among early-career engineers. Brunhaver defines adaptability as the capacity to recognize uncertainty and respond effectively, though she notes its application varies by field. While aerospace or biomedical engineers must track evolving regulations, software engineers face constant tool turnover. Jenna Butler, a research scientist at Microsoft, describes the current transition as an uncomfortable "chaos period" where software engineering is shifting from writing code line-by-line to prompting models and managing agents.
For practitioners, this evolution changes how they must approach their careers. Andy Hunt, co-author of the 1999 book "The Pragmatic Programmer," argues that engineers must prioritize systems thinking over specific tools, comparing a developer who only identifies with one language to a carpenter who only uses a hammer. To prevent burnout during this transition, Butler recommends that organizations allocate dedicated learning time, even just one hour a week, allowing engineers to explore new AI tools without the immediate pressure to produce code. Ultimately, engineers must actively direct how they use these models rather than letting the technology dictate their workflow.
This is our own summary of reporting by IEEE Spectrum AI
