The creator of the tool which appears to have played the biggest part in automating coding believes that AI could eventually make “software engineers” a thing of the past.
Boris Cherny, the Anthropic engineer who built Claude Code — the AI-powered command-line coding tool that has rapidly become one of the most talked-about developer products of the past year — made the remarks in a recent interview, offering a rare inside view of how those closest to frontier AI think about the future of technical work. What makes his perspective particularly striking is that it comes not from an industry observer or futurist, but from someone who has spent his career building the very tools that are driving this transformation.
“I think today coding is practically solved,” Cherny said. “For me — and I think it’ll be the case for everyone, regardless of domain — I think we’re going to start to see the title of software engineer go away. And I think it’s just going to be maybe ‘builder,’ maybe ‘product manager.’ Maybe we’ll keep the title as kind of a vestigial thing.”
For Cherny, the shift is already visible within Anthropic itself. The change isn’t just about what engineers do — it’s about who codes at all. “The work that people do, it’s not just going to be coding,” he explained. “Software engineers are also going to be writing specs. They’re going to be talking to users.” He pointed to what he described as a flattening of roles across his own team: “This thing that we’re starting to see right now on our team, where engineers are very much generalists and every single function on our team codes — our PMs code, our designers code, our EM codes. Our finance guy codes. Everyone on our team codes. We’re going to start to see this everywhere.”
Cherny described this as something of a floor — the minimum we should expect if current trends simply continue. But he was candid that the ceiling looks considerably more unsettling. “This is sort of the lower bound if we just continue with the trend,” he said. “The upper bound, I think, is a lot scarier.”
That upper bound, in his telling, is defined by what Anthropic internally calls ASL-4 — the next rung on the company’s AI Safety Level framework. “ASL-3 is where the models are right now,” he noted. “ASL-4 is the model is recursively self-improving.” At that point, Anthropic would need to meet a series of safety criteria before releasing any new model. The risks he flagged were stark: catastrophic misuse, including people using models to design biological viruses or software exploits. “This is something that we’re really, really actively working on,” he said, “so that doesn’t happen.”
Cherny’s comments land at a moment when the industry is already grappling with the implications of AI-assisted development. Cursor, the AI-first code editor, has attracted millions of users and a multi-billion-dollar valuation in just a few years. Claude Code itself — which allows developers to delegate entire coding tasks to an AI agent working in the terminal — has been credited by some engineering teams, including at rivals like Google, with dramatically cutting the time needed to ship software. Meanwhile, major technology companies including Google and Microsoft have begun publicly disclosing that a growing share of their code is now written or co-written by AI. The trend Cherny describes — where coding becomes a baseline skill distributed across an organisation rather than the exclusive domain of a dedicated engineering team — appears to already be forming in the companies best positioned to move fast. Whether the job title follows, or quietly fades into the background like roles that once defined earlier technological eras, may simply be a matter of time.