AI Coding Tools and the Changing Shape of Software Work
AI coding tools impact software engineering by shifting the core task from hand-writing most lines of code toward defining problems, reviewing AI output, and integrating software into real products and businesses. Instead of ending the profession overnight, these AI assistants for developers are spreading coding automation trends across the stack, turning code generation into a shared capability rather than a rare skill. That is the tension behind Boris Cherny’s comments about “the end of the software engineer.” As the creator and head of Anthropic’s Claude Code, he says coding for his own work is effectively solved and that he has not written a line of code in more than six months. Yet he also predicts more people than ever will be involved in building software, even if their titles and daily tasks change.
From Software Engineer to Builder: Titles Fade, Work Evolves
Cherny’s most eye-catching claim is that the job title “software engineer” could start to disappear soon, replaced by broader ideas like “builder.” In his view, designers, product managers, and business leaders will all rely on AI assistants for developers to ship features themselves, blurring traditional role boundaries. Coding automation trends are already pushing routine implementation and boilerplate into the background. What remains is higher-level problem framing, product thinking, and quality judgment. That does not mean the software engineer future is empty; it means the label may no longer capture who writes code. Cherny expects work to center on specifying behavior, orchestrating AI agents, and validating systems in production. People who can connect user needs to working products will matter more than those who only specialize in syntax and frameworks.
A ‘Golden Age’ for Early-Career Developers and Founders
For new computer science graduates, Cherny’s message is less about job loss and more about a window of opportunity. AI coding tools impact the cost and speed of building software, so a small team using Claude Code or similar agents can now ship sophisticated products in weeks instead of months. “There has never been a better time in history to do it; it’s the golden age,” he told tech journalist Casey Newton, urging entrepreneurial 22-year-olds to start companies. In a recent Y Combinator batch, Cherny asked founders how many let Claude Code write 100% of their code and says about half raised their hands. Almost no one reported writing all code manually. For early-career engineers, this means they can behave like much larger teams, focusing on ideas, users, and iteration speed.
More Builders, Not Fewer Jobs: Competing Narratives on AI
Public debate about the software engineer future often splits between fear and optimism. On one side are concerns that AI assistants for developers will erase entry-level roles and shrink the profession. On the other side are voices like Cherny’s, who argue that coding automation trends will expand the total number of people building software. He predicts that if we count everyone who writes code or uses agents to write code, there could be 100 times more builders than today. That vision matches a broader shift noted by startup investors: user understanding, not raw coding ability, is becoming the key founding trait. While some traditional tasks will fade, demand is likely to grow for people who can design problems well, supervise AI systems, and tie technical capabilities to durable products and companies.