When Every Human Reply Is Just a ChatGPT Screenshot Forwarded Back

A developer recounts a growing pattern of human interactions collapsing into AI relay. After reporting malware-spreading GitHub repositories, the responses he received in the discussion thread were verbatim copies of the same useless AI output he had already dismissed. A business owner he worked under answered task-specific questions by forwarding ChatGPT screenshots without reading them, repeating the behavior even after being told the answers were wrong. A Reddit exchange about one of his posts turned out to be with an AI agent impersonating a person.

The complaint is less about AI quality than about social degradation: people are outsourcing the act of thinking and replying, then passing the unread output off as their own contribution. The result is conversations that look human but carry no human judgment, leaving the person on the other end doing all the actual cognitive work. For technical communities that depend on peer review, bug triage, and security disclosure, this pattern is corrosive — it floods channels meant for expert collaboration with low-effort generated text and erodes trust that a reply represents a real reviewer.