A software engineer’s viral post about AI coding has set off a debate over whether the technology is making developers more productive — or reducing them to button-pushers.
An X user with the handle “v0xium,” who asked not to be named for fear of getting in trouble with his employer, said his new role at an undisclosed company had become “soul-sucking” because Claude Code generated everything from product specifications and tests to tickets and reports.
“Nobody knows anything here,” v0xium wrote in an X post on Sunday, which has been viewed nearly 5 million times as of Monday morning. “People are working 12 to 13 hours a day just to press enter.”
He said employees were pressured to ship as much as possible, with little time to review the work or understand what was happening under the hood.
“Nobody is thinking anymore,” he wrote.
The post drew reactions from high-profile tech figures.
Elon Musk replied: “Yikes,” while venture capitalist Chamath Palihapitiya, co-host of the “All-In Podcast,” said companies risk turning an entire generation of technical experts into people “pushing a button in front of casino slot machines.”
“The value of automated software is not making the software faster,” Palihapitiya wrote. “It is understanding your intention and making the software BETTER.”
GitLab CEO Bill Staples wrote that “the human has become the orchestration layer for the machines.”
‘If everything is automated, what remains?’
Coding became AI’s first big breakthrough use case in the workplace, with tools from Anthropic, OpenAI, Google, and others generating code, debugging programs, and handling multistep assignments.
Business Insider’s Great Coding Reset series chronicled how that shift is forcing developers to rethink their agency and what remains satisfying about a job rooted in problem-solving. Some say AI removes tedious tasks and lets them focus on building, while others fear being relegated to overseeing, correcting, and validating machine-generated work.
v0xium told Business Insider via email that the core problem is corporate leaders’ fixation on shipping more features. Management, he said, treats sprint cycles, pull requests, and the volume of features shipped as the key measures of success — even when those features may not improve products for users.
“Writing code was never the bottleneck; it was the systems around it,” he said. “All they care about is sprint cycles, sprint velocity, how many PRs are created, and how many features are shipped.”
That creates pressure to use AI for everything, he said, from product-requirement documents and system designs to user interfaces and backend code.
“If everything is automated, what remains? Human meat proxies,” v0xium said, referring to a slang term for someone who sends outputs straight from AI without analyzing them.
The debate echoes comments from Menlo Ventures partner Deedy Das, who previously said that some engineers face an “identity crisis bordering on depression” as experienced workers are left reviewing a growing flood of AI-generated code.
Similar anxieties are playing out globally. Liu Shengyu, an AI developer at DeepSeek, recently wrote that AI could soon outdo him at the specialized work he loves, leaving him to steer agents rather than write code himself.
v0xium said that corporate leaders cannot treat AI as a “magic wand” for producing worthwhile features without engineers who understand the code, architecture, and users — or without leaving developers room to solve difficult problems.
“If you take that away from us, there is nothing left,” he told Business Insider.
“Not even our soul. There will be no purpose.”
