Monologue: LLM Code Is Already Breaking Big Tech

Host Zitron warns that large language models (LLMs) are already destabilizing major tech companies by enabling rapid, low-quality code production, particularly when used by non-technical staff or with rushed reviews by engineers.

Non-Technical Workers Generating Production Code

Zitron describes a hyperscaler—likely a major cloud provider—where management encourages non-coders to build features for consumer products using generative AI, a practice he calls "vibe coding." These workers prompt LLMs without understanding code, producing outputs that engineers theoretically review before deployment. He argues this is "fucking insane," as it floods systems with code lacking clear intention. Reviewers often lack time for the verbose LLM output, and even adept engineers struggle under volume. Management mandates LLM use, tracking adoption, which burdens technical staff with fixes and monitoring.