July 18, 2026
Ctrl+Alt+Defeat
What AI did to stackoverflow in a graph
That sad question graph has commenters blaming bots, burnout, and Stack Overflow’s own attitude
TLDR: A graph of Stack Overflow’s monthly questions sparked a messy debate over why the once-huge help site is shrinking. Commenters split between blaming AI tools and blaming the site’s own harsh user culture, turning a data chart into a public dragging.
A tiny Stack Exchange data page just dropped a giant internet mood swing: a graph of how many questions were posted to Stack Overflow each month, and yes, the crowd immediately turned it into a full-on blame game. The chart shows the coding help site rising fast for years, then sliding downward later on. And once people spotted the dip, the comments lit up with the digital equivalent of "we need to talk."
One camp says the graph clearly shows AI eating Stack Overflow’s lunch. Another camp was not having that neat little storyline at all. Several commenters argued the decline started long before ChatGPT showed up, with one saying you can even spot COVID in the graph and another insisting the chatbot era only gave the fall a small shove, not a dramatic push. In other words: was AI the killer blow, or just the latest villain to arrive after the party was already on fire?
But the spiciest reactions weren’t really about robots — they were about people. More than one commenter accused Stack Overflow of driving users away with a chilly culture, calling out "incredibly unhelpful users" and "ridiculously aggressive moderation." Ouch. The vibe was basically: sure, AI matters, but maybe the site also helped dig its own hole. The result? A graph became a roast session, with commenters treating the line chart like a breakup receipt for the internet’s once-beloved coding advice hub.
Key Points
- •The article content is a Stack Exchange Data Explorer page for a Stack Overflow query titled "What AI did to stackoverflow in a graph."
- •The SQL query counts monthly Stack Overflow questions by grouping `CreationDate` values by month and filtering `Posts` where `PostTypeId = 1`.
- •The results table reports 217 monthly rows returned.
- •Visible results begin in July 2008 with 4 questions and rise sharply through the following months.
- •The displayed portion of the table shows monthly counts exceeding 100,000 by March 2011, indicating strong early platform growth in the visible data.