Every weekday, a software engineer named Matt eagerly anticipates his four-hour train ride to Pawling, New York. This commute allows him uninterrupted time to focus on his personal project: developing a browser-based video game, entirely coded by himself.
“I am actively working to hone my skills,” said Matt, who requested anonymity to safeguard his job security. Over the past six months, he has noticed his role transitioning from coding and solving complex problems to mostly reviewing code produced by artificial intelligence. Fearing this shift will diminish his proficiency, he strives to minimize his reliance on AI tools. “I’m trying to avoid using AI whenever possible.”
Once a reliable profession with an annual salary exceeding $200,000, Matt’s software engineering career feels uncertain after experiencing a layoff last summer and receiving advice from his current employer to adopt more AI usage.
For many like Matt, software engineering represented a stable and lucrative career path. Yet, as AI rapidly transforms software development—Google reports that 75% of its code is now AI-generated—engineers find themselves grappling with unexpected changes to their profession. Frustration, anxiety, and the quest for relevance are common sentiments among engineers, who are now focusing on core fundamentals, pursuing new skills, advocating for better workplace protections, or even considering leaving the field altogether, as shared by numerous engineers in discussions with the Guardian.
The US Bureau of Labor Statistics highlighted software engineering as one of the highest-paying and most sought-after professions in 2022, with 1.5 million professionals earning double the national average salary. As competition for talent escalated, companies offered substantial bonuses to attract top developers. According to market research from SlashData, nearly 50 million developers were actively working worldwide last year.
In contrast, since the advent of OpenAI's ChatGPT in 2022, over 600,000 tech jobs in the US have been lost, as reported by Layoff.fyi. The unemployment rate for new computer science graduates climbed to 7% in 2024, up from 6.1% the previous year, with an underemployment rate exceeding 19%, based on New York Fed data. Job postings for tech positions on Indeed have also fallen by 36% from 2020 to 2025.
While experts are uncertain about the future of software engineering roles, they concur that while coding skills may be decreasing in importance, the ability to assess AI-generated code is becoming increasingly valuable. Bouke Klein Teeselink, an assistant professor at King’s College London, stated, “While it's difficult to predict the exact trajectory of the profession over the next couple of years, it's evident that the traditional skill of writing code is evolving.”
Ethan Mollick, an associate professor at the Wharton School of the University of Pennsylvania, emphasized that the focus has shifted from pure coding to effectively problem-solving, system design, and utilizing AI tools. “It changes the landscape, positioning these skills as what hold the most value now,” he explained.
For Matt, the nature of his contributions has shifted significantly. Whereas he once played a pivotal role in executing solutions, he now finds the line between his own decisions and the input supplied by AI increasingly hazy.
Engineers face a daunting choice: continue navigating a volatile industry or pivot to alternative careers.
George Dover, a Portland-based software engineer with six years of experience, took on a substitute teaching role for kindergarten students after being laid off from Inuit Mailchimp in late 2024. “Letting go of a career that defined me for so long was immensely difficult,” he shared. “What else is out there for someone like me?”
Despite the challenges, Dover persevered. Recognizing the necessity of understanding AI, he adopted AI tools to generate code for building websites, using his insights to identify strengths and weaknesses in the AI-generated code. “The quality needs thorough verification,” he noted, referring to the trade-offs he often encounters.
His efforts bore fruit: after submitting around 400 applications and attending numerous interviews, Dover recently secured a job as a software engineer with a focus on AI.
Dover's story is not unique as more non-technical individuals venture into coding, which could ultimately boost demand for software engineers, according to Teeselink. Assessing AI-generated code requires a skillset that non-coders typically lack, including identifying vulnerabilities and understanding errors.
That said, experts caution against definitive conclusions about the profession just yet, as high-quality AI-generated code is relatively new. Shriram Krishnamurthi, a computer science professor from Brown University, mentioned that the growing need for code assessments might leave some engineers behind. “Some are well-prepared for this paradigm shift; others will need to adapt quickly.”
While the role of human software engineers may change, they will continue to be necessary, especially considering the costs associated with developing AI technologies, as highlighted by David Malan, a computer science professor at Harvard University. Reports indicate that OpenAI spent approximately $8 billion, with Anthropic burning about $3 billion last year. As AI development costs likely pass on to consumers, Malan anticipates a balanced collaboration between human engineers and AI.
For over a decade, the call to learn coding resonated widely.
In December 2013, President Obama initiated a $4 billion program titled Computer Science for All to educate American students in coding, declaring it essential for economic advancement. Influential tech figures like Mark Zuckerberg and Bill Gates participated in tutorials, while coding bootcamps surged in number, yielding a significant increase in graduates—from 2,000 in 2013 to over 20,000 by 2020, according to Course Report.
As Sam, a Los Angeles resident who declined to disclose his last name for privacy, reflected on his own journey into software engineering a decade ago, he expressed concern about the profession’s future. After struggling with significant student debt while studying music and trying a variety of jobs, he invested in a formal education in software engineering.
Now, in just five years, he has gone from feeling secure in his career to fearing job loss. He laments how AI has taken away the most engaging aspects of his work, reducing it to “reviewing code I didn’t create,” he revealed.
Sam considers alternative career paths as he grapples with questions about the sustainability of his current role, especially with tech giants like Google and Amazon letting go of employees.
“I often find myself wondering, ‘Should I start a food truck? What if I switch to forestry?’” he confided.
Sam is not an isolated case. Over a decade after the coding boom, enthusiasm for the profession has waned. Enrollment in computer and information science programs at four-year universities decreased by 8.1% in the 2025-2026 academic year, while graduate enrollment plunged by 14%, according to the National Student Clearinghouse.
Amidst these challenges, software engineers are increasingly coming together to address their shared concerns about AI's impact on their careers.
Kaitlin Cort, feeling the pressure of AI's rapid evolution in her role, chose to step away and establish a resource center for tech workers navigating the industry's upheaval. “I may not be a senior engineer, but seeing how quickly AI was advancing made me anxious,” she expressed.
Initially drawn to software engineering to alleviate student debt, Cort also taught coding at bootcamps and witnessed the transformative power of coding on lives. However, as her responsibilities shifted from writing code to evaluating AI outputs, she reconsidered her career’s direction.
In February, Cort founded What We Will, a center aimed at assisting tech workers affected by layoffs and industry changes, helping them gain access to benefits, enhance skills, and engage in collective efforts. Her first initiative targeted Amazon workers impacted by AI adoption and offered support to those laid off from Oracle. In May, they worked with employees from Meta to discuss issues surrounding workplace oversight.
As job insecurity prompts growing interest in collective actions, Cort receives an influx of applications daily. "In recent months, an increasing number of individuals have expressed interest in unionization," she noted.
Despite the challenges of building this initiative without a clear precedent, she has sought guidance from organizations like the Alphabet Workers Union and the Washington Post Tech Guild.
“There’s a pressing demand for these services right now,” she stated. “As an industry, we still lack standards and regulations.”
