Updated
Updated · TechRadar · Jul 12
Generative AI Drives 76% Drop in Stack Overflow Questions, Pushing Experts Out
Updated
Updated · TechRadar · Jul 12

Generative AI Drives 76% Drop in Stack Overflow Questions, Pushing Experts Out

2 articles · Updated · TechRadar · Jul 12

Summary

  • University of Auckland research says generative AI is accelerating the exit of Stack Overflow’s highest-skill contributors, as experts see less reward in sharing knowledge that chatbots now deliver instantly.
  • Nearly 76% fewer monthly questions have been posted since ChatGPT’s 2022 arrival, reflecting both shrinking user demand and a loss of answerers that the study says may be hard to replace.
  • Stack Overflow’s own problems also fed the decline: users cited heavy-handed moderation and a self-righteous culture, while AI became a faster tool for routine coding and syntax queries.
  • Dr. Kenny Ching calls the effect “signal compression” — AI makes expert and non-expert answers harder to distinguish, weakening incentives for specialists to contribute.
  • Researchers warn the pattern could spread beyond coding forums into classrooms, offices and scientific communities, raising questions about where future AI systems will find high-quality training data.

Insights

With public expertise fading, who is secretly teaching our AI, and at what cost?
AI can code, but can it lead? What human skills are now more valuable than technical mastery?
If AI learns from its own flawed code, are we building a future on a foundation of digital decay?