AI Coding’s Button-Pushing Crisis Is Here

The button is becoming the job. Powerful AI coding tools have made programmers heavily reliant on technology, raising fears that engineers are losing the ability to write code by hand and understand the software they build.
A viral post from software engineer v0xium describes an unnamed company where everything is being made by Claude Code, Anthropic’s AI coding assistant. The post presents a workplace in which engineers are instructed to use nothing but AI, turning software development into what the writer calls a race to the bottom.
“It has been half a month since I started a new role at a big company. Nobody knows anything here,” v0xium wrote. The post was liked tens of thousands of times and circulated over seven million times, turning one engineer’s complaint into a broader argument about what happens when coding becomes a sequence of prompts and approvals.
When Every Engineer Talks to Claude
At the unnamed company, people are working 12 to 13 hours a day just to press enter. “Everyone, literally everyone, from an L1 to an L7 engineer here is doing the same thing. Talk to Claude,” v0xium wrote.
The point is not that engineers have stopped working. It is that their work has changed shape. Coders who once wrote software, solved problems, and built products themselves now say they are pushing buttons instead of taking pride in the craft.
“In reality, nobody is thinking anymore,” v0xium wrote. That line captures the central fear: AI tools may produce more code while leaving the people responsible for it less connected to how the code works.
The result is an identity crisis bordering on depression for many software engineers, according to Deedy Das, a partner at Menlo Ventures. “Most software engineers are facing an identity crisis bordering on depression,” Das wrote.
The Cost of Making More Code
V0xium describes the craftsmen as tired. Their workload grows day after day, bugs seep into production, and no one seems to care; when problems appear, another round of AI is thrown at them instead of addressing the deeper failure.
That cycle also damages relationships inside engineering teams. As the workload grows and bugs remain, animosity toward colleagues rises. Eventually, the engineers give up: “The craft they loved is dead.”
Chamath Palihapitiya, a podcast host and venture capitalist, offered a broader warning about the same shift. “We are turning an entire generation of technical experts around the world into the equivalent of retirees pushing a button in front of casino slot machines,” he wrote.
His comparison is not subtle, but neither is the complaint from engineers working 12 to 13 hours a day to press enter. A tool meant to remove repetitive work can still create a new kind of repetition when every level of a company follows the same instruction: talk to Claude.
Palihapitiya also argued that automated software should not be judged by speed alone. “The value of automated software is not making the software faster, It is understanding your intention and making the software BETTER.” The distinction matters because generating more code does not guarantee better products, fewer bugs, or stronger understanding.
He added another warning: “Essentially the brain rot of doom scrolling social media is coming for your development team.” The comparison places AI-assisted coding beside a familiar form of passive consumption—an activity that fills time while demanding little thought from the person doing it.
Victor Tangermann, Senior Editor at Futurism, highlighted the concerns in coverage published Sep 23, 2026 8:07 AM EDT. The discussion also included Elon Musk, the owner of X, and the post carried the date 2026-09-21T14:54:53.972Z.
The argument is not that Claude Code has no value. The warning is that companies can use it in a way that strips software engineers of judgment, ownership, and the satisfaction of making something they understand.
AI can write the code. That does not mean anyone is still building the software.
Based on




