The piece cites Rahim Amir as author, noting his background as a UAE-based tech writer with experience in PC hardware journalism.

Generative AI delivers fast, repeatable solutions that mimic top contributors’ work, reducing incentives to participate as their efforts feel underappreciated.

While less established users left earlier, the exit rate of expert contributors continued to rise, narrowing the gap with newer users who departed previously.

A University of Auckland study analyzing over 24,000 Stack Overflow contributors over 17 months finds that high-reputation, veteran users began exiting the platform at an accelerating pace starting around 2022, coinciding with the mainstream rise of generative AI tools like ChatGPT.

The trend may extend beyond Stack Overflow to classrooms, corporate settings, and scientific communities, making it harder to distinguish expert input from AI-generated answers in retraining cycles.

If high-quality contributors continue to exit, the long-term ecosystem for expert-curated knowledge may struggle to recover, impacting the reliability and depth of information for developers and researchers.

Root causes of departures extend beyond AI: perceived hubris among top contributors and heavy-handed moderation, with AI increasingly acting as a flexible search tool for routine questions.

The research raises concerns about AI training data pipelines, noting that reliance on human-in-the-loop sources such as Slack, Discord, or direct chatbot queries could introduce uncertainty in data quality for retraining models.

AI’s ability to provide reliable answers to common coding problems reduces demand for expert-written responses, and attempts to limit AI use did not stop the long-term decline in expert participation.

Ching argues AI, not Stack Overflow’s moderation, was the primary driver of top contributors leaving, suggesting broader implications for other domains where AI substitutes deep expertise.

Dr. Kenny Ching describes the trend as 'signal compression,' where AI-generated answers increasingly resemble expert responses, diminishing the perceived value of human expertise on the platform.

Quantitatively, Stack Overflow saw monthly questions drop by about 76% since the arrival of ChatGPT, signaling a broad decline in engagement and content quality.