Google Employees Slam AI Coding Tools as 75% Code Claim Fuels Bottleneck Complaints
Updated
Updated · 404 Media · Jun 4
Google Employees Slam AI Coding Tools as 75% Code Claim Fuels Bottleneck Complaints
3 articles · Updated · 404 Media · Jun 4
Summary
Dozens of Google employees are using the internal Memegen board to mock AI tools such as Jetski, with anti-AI posts appearing weekly and one source estimating the total in the high hundreds or thousands over the past year.
Jetski drew particular ridicule after a screenshot showed it spending 11 seconds "thinking" before admitting numeric metrics in a report were simulated rather than pulled from live systems; that meme drew more than 400 upvotes.
Employees said AI has sped up code generation but shifted the bottleneck to testing, builds, version control and human review, leaving reviewers to sort through larger volumes of code that "no one understands because no one wrote it."
The backlash contrasts with Sundar Pichai's public statement that 75% of new Google code is AI-generated, while some employees said AI projects are being prioritized, other work is being canceled, and pressure is rising to inflate AI productivity metrics.
Google said candid internal criticism—including through its meme generator—helps refine the tools to maximize daily productivity, though it revised an earlier statement to remove language stressing that humans must remain in the loop.