Since the Industrial Revolution, workers have feared that machines would replace them. However, in many industries, automation didn’t eliminate jobs but rather “degraded” them — breaking down complex tasks into repetitive, rapid-fire duties. This trend is now extending to white-collar jobs, especially software engineering, as artificial intelligence (AI) tools like GitHub Copilot become common in coding. While AI boosts productivity, it is also changing the nature of programming work — making it faster, more routine, and less reflective.

Historically, skilled roles were replaced by assembly line-style work. Similarly, personal secretaries were replaced by typists and clerks. Labor historian Jason Resnikoff notes workers complained of “speedup, work intensification, and work degradation.” The same is now being seen in programming: AI tools assist with code generation and productivity, but engineers say their jobs are becoming more repetitive and time-pressured. A Microsoft-led study found Copilot improved productivity by over 25%, but at the cost of creativity and job satisfaction.

At Amazon, which is heavily investing in generative AI, the culture of software development is changing rapidly. CEO Andy Jassy emphasized the importance of speed and productivity, noting that AI is shifting coding norms. Engineers report increased pressure to use AI and meet elevated output targets. One Amazon team, half its original size, is expected to produce the same amount of code with the help of AI tools. While Amazon claims teams are regularly evaluated and resized as needed, engineers feel expectations have intensified.

Other companies are following suit. Shopify now includes AI usage in performance reviews, and Google is hosting a company-wide AI productivity hackathon. These efforts demonstrate how central AI has become to workplace strategy in tech firms.

Some managers argue AI frees workers from mundane tasks, allowing them to focus on more engaging work. Amazon, for example, used AI to modernize legacy code, saving an estimated 4,500 developer-years. However, for junior developers, AI’s automation of basic coding and testing tasks risks undermining skill development and career growth. Engineers say that AI-generated code requires more review than writing original code, turning developers into code readers rather than creators. This shift can make them feel like bystanders in their own jobs.

The impact is comparable to the automation in Amazon’s warehouses, where robots improved productivity but made jobs more monotonous. Engineers worry they face a similar fate. Although AI use is optional in theory, it is becoming essential to meet output goals and perform well in evaluations.

Amazon employees have voiced growing concerns about how AI is affecting their jobs, careers, and overall work quality. Groups like Amazon Employees for Climate Justice have become spaces for workers to share stress about AI pressures, beyond just environmental concerns.

While unionization hasn’t occurred among Amazon engineers, the situation echoes historic labor movements like the United Auto Workers strike in 1936, which was sparked by a loss of control over work pace and methods. As AI reshapes coding, many engineers now fear a similar erosion of autonomy, creativity, and meaningful engagement in their work.