I use AI tools every day. ChatGPT, Claude, whatever gets the job done faster. Most devs I know do the same now. And individually, it feels great. You solve problems quicker, you get unstuck faster, you ship more.
But I think something important is breaking in the background, and we’re not really talking about it.
For years, places like Stack Overflow were where knowledge piled up in public. Not just answers, but debates, edits, corrections, and context. You could see how solutions aged. You could see people disagree. You could learn how others thought.
Now most of that problem solving happens in private chats.
You ask an AI, you get an answer, you move on. No attribution. No discoverability. No trail for the next person who hits the same problem tomorrow.
Stack Overflow traffic didn’t drop because devs got worse. It dropped because AI replaced the interface, not because it replaced understanding.
https://stackoverflow.com
Wikipedia is another warning sign. It’s still one of the best sources of human knowledge ever created, yet it’s getting buried by AI summaries and SEO pages that were trained on it.
https://en.wikipedia.org/wiki/Knowledge_commons
The scary question is simple.
If we stop contributing to public knowledge, what does the next generation of AI train on?
We’re consuming the commons without replenishing it. That never ends well.
I don’t think the answer is “stop using AI”. That’s unrealistic. The answer might be using AI, then publishing the reasoning, not just the result. Turning private speed into public memory.
Right now we’re optimizing for personal efficiency. Long term, that might cost us collective understanding.
Top comments (0)