Stack Overflow has seen a substantial decline in traffic over the last year that appears to be accelerating. https://observablehq.com/@ayhanfuat/the-fall-of-stack-overflow
Stack Overflow has seen a substantial decline in traffic over the last year that appears to be accelerating. https://observablehq.com/@ayhanfuat/the-fall-of-stack-overflow
Most of the comments here seem to be arguing whether it’s better to get help now from SO or ChatGPT, but this is a pretty short-sighted mindset.
What happens when the next new standard comes out that ChatGPT hasn’t been trained on? If SO tanks and dies, where will you go?
I’m not saying use a lesser resource, I’m saying this is kinda tragic and I hope they can sustain themselves; AI is propped up by human input and can’t train itself.
Hey, if people are going to go back to reading manuals like we’re in the 1980’s again is it such a bad thing? /s
It’s insane how a single tool managed to completely destroy the value collectively created by people in over a decade.
That single tool is still propped up by that collective decade of knowledge. ChatGPT would be nothing without sites like stackoverflow
Yeah but will people still care about contributing that information if they’re not going to be compensated for it in any way? Like people get something out of contributing to stack overflow, even if it’s just recognition. This is gone with ChatGPT.
very good point! I find myself using ChatGPT more for references and I am also afraid what will happen if there isn’t enough “human generated content” to train on. I can picture an edge case a chunk of the internet is AI generated content (with even users at the wheel). The the next wave of AI will train on previous gen AI output
Crazy idea, what about a “federated” search. Hook up the websites’ internal search engines to an aggregator. Stop allowing random indexing spiders to scrape.
AI should be trained by itself though. I just wouldn’t call LLMs “AI” as a term
Also, it shall be possible in the future to just feed it the documentation and have answers. Obviously we are still nowhere near yet
We go back to expertsexchange