Defining the Debate: Is AI Replacing Engineers or Redefining Them?
The question of whether AI is replacing engineers focuses on how tools that generate or modify code change software engineering jobs, threaten traditional roles, and create new ways of building products and companies. In this debate, Anthropic’s agentic coding tool Claude Code sits at the center. Its creator, Boris Cherny, argues that for much of his own work, “coding is effectively solved,” and he has not written a line of code in more than six months. That experience informs his prediction that the title “software engineer” could begin to disappear as AI agents let designers, product managers, and other non-specialists ship production code. At the same time, Cherny maintains that people who write code or direct agents will not vanish; instead, their titles, workflows, and career paths will change.
Claude Code and the Claim that Coding Is Getting Solved
Claude Code is an AI coding agent built by Anthropic to automate large parts of the programming workflow, from generating new modules to maintaining existing systems. Cherny describes the tool as writing a growing share of real-world code, and points to cutting-edge founders as evidence. In a recent session with Y Combinator companies, he asked how many teams let Claude Code write 100% of their code, and about half of the hands went up. When he asked who does not use the model for any coding, only one hand rose among a few hundred people. This supports his view that “coding is getting solved for a bigger and bigger percentage of the code we write,” and explains why he believes classic engineering roles will undergo sharp pressure from AI replacing engineers in many routine tasks.
The End of the Software Engineer Title—and the Rise of Builders
Cherny’s boldest claim is not that software work disappears, but that the familiar job label does. He argues the title “software engineer” could start to fade, replaced by broader descriptions such as “builder,” as people in design, product, and leadership roles direct AI agents to ship code. In his view, coding expertise shifts from manual implementation to specifying requirements, reviewing outputs, and integrating systems. AI replaces engineers only in the sense that fewer people will spend their day hand-writing boilerplate or routine features. Instead, more people across a company will influence live software. That rebalancing fuels both anxiety and optimism in the developer job market: traditional junior roles may shrink, but a wider set of workers gain the power to build software-backed products.
Golden Age for Graduates: Why Cherny Pushes Startups Over Safe Jobs
Despite warning that classic engineering roles are under threat, Cherny delivers almost the opposite message to new computer science graduates. He tells 22-year-olds that they “can totally still” find entry-level software engineering jobs, but urges anyone with entrepreneurial instincts to found a startup. With AI coding agents like Claude Code, a small team—or even a solo founder plus a cluster of agents—can build and scale products that once required large engineering departments. Cherny calls this “the golden age” for such founders, predicting that if we count everyone who writes code or directs agents, “there will be 100 times more of them than there are today.” In that framing, AI does not eliminate software work; it spreads it, making company-building accessible to people who understand users even if they cannot code by hand.
What Claude Code Means for the Future Developer Job Market
Cherny’s stance captures a wider split in the tech industry: AI replacing engineers in some tasks while expanding the pool of people who can build software. For established developers, Claude Code’s impact likely means less time on routine coding and more on design, oversight, and system-level thinking. For new graduates, the choice is sharper: compete for fewer traditional junior roles, or treat AI agents as force multipliers for new ventures. According to Anthropic’s Boris Cherny, the broad work of “people writing code or using agents to write code will not” go away, even if the job titles change. That suggests a developer job market where success depends less on being the fastest coder and more on knowing what to build, how to guide AI systems, and how to turn software into a viable product or company.