Software Engineers Wrestle With AI Code Overload and Identity Collapse

Software engineers are drowning in AI-generated code. This flood is tearing their profession apart.

“Most software engineers are facing an identity crisis bordering on depression,” says Deedy Das, a venture partner. The craft they once loved feels dead. The old pride in building clean, thoughtful code has faded.

Veteran engineers, known as ‘craftsmen,’ now spend their days sifting through and fixing horrendous AI-written code. Bugs seep into production. Their workload balloons while management pumps more AI tools at the problem. “The craftsmen are tired. Day after day, their workload grows. No one seems to care,” Das adds.

AI-generated code now makes up more than 25 percent of Google’s new codebase. About 25 percent of startups in Y Combinator rely on AI for 95 percent of their code. In the Stack Overflow 2025 survey, 84 percent of developers say they use or plan to use AI tools. Yet trust in these AI helpers has fallen to 29 percent, down 11 points from last year.

Workplaces push employees to maximize AI use. Meta even factors AI usage into performance reviews. This push has a name: “tokenmaxxing.” It rewards quantity of AI output over quality. Meanwhile, studies of 211 million lines of code reveal a 17.1 percent rise in copy-pasted code, an eightfold surge in duplicated blocks, and a 26 percent jump in code churn after AI adoption. The illusion of productivity masks growing mess.

This divide splits engineers into two camps. Less experienced “brainless vibe coders” lean heavily on AI crutches. Experienced craftsmen battle the fallout, wrestling with sloppy code. The clash breeds animosity. Veteran engineers feel their role slipping from skilled creator to cleanup crew.

Despite all this, software engineer job postings rose 11 percent year over year. Demand for skills persists. Adaptable engineers now focus on systems thinking, defining intent, and evaluating AI suggestions. They compare this shift to the industrial revolution’s move from hand-weaving to machine oversight.

The profession’s identity is at stake. The role shifts from pure implementation to orchestration and judgment. Some engineers double down on fundamentals. Others embrace AI mastery and system design. But if companies fail to redistribute work and restore ownership, burnout will spread. Many may quit the field altogether.

The software craft is evolving. Whether it survives this AI upheaval depends on how engineers and organizations adjust. For now, the crisis looms large, and the future looks uncertain.