The "community" also includes the people who volunteer to answer the questions. The answerers don't want repetition. Jeff Atwood wanted to prioritize the SO experience for the answerers.
To paraphrase JA's reasoning:
- asking questions is easy which can lead to an increasing volume of bad questions that degrades the site for the answerers
- answering questions is hard and more valuable for StackOverflow so optimize the site for them. (JA wrote, ">We feel that the world is awash in questions, but not answers. Answers are the real unit of work in any Q&A; system. Therefore, the only logical thing to do is to maximize the happiness and enjoyment of answerers."[1])
I do understand how you would want SO to work differently and I'm not debating your opinion. I'm just relaying how the cofounders wanted StackOverflow to work.
[1] https://stackoverflow.blog/2011/06/13/optimizing-for-pearls-....
I think this assumption is not true anymore - ChatGPT can give an answer in less time and (human) effort than it takes to formulate and devise a question. So it really is about creating a set of curated question-answer pairs, where both the questions and the answers can be auto-generated. In this context, asking the same question different ways is really helpful to an AI. And there is no such thing as a bad question, there are just Q-A pairs that do not contribute significantly to training the model ("boring" questions). I've noticed Quora is experimenting with ChatGPT but they aren't at the point where you can flag the AI's answer as incorrect or misleading and fix it.
Can it? I mean, you still have to spend the time and effort required in order to ensure that ChatGPT's answer is correct and complete.
No, asking bad or thoughtless questions is easy. Asking a good question, complete with "this is what I've tried" and snippets and links to all revelant resources like API documentation, CodePens/Godbolts/standards and a summary of your thinking that has led to needing an answer in the first place is very hard.
A good question is probably at least an hour's work, and there are very few easy questions left that don't need substantial work even to frame.
An answer can be very in-depth with research and analysis, which also takes a long time, but it can also be anywhere on the spectrum down to fairly straightforward "you missed a function call" answer.
On the other hand, writing a good question sometimes solves your problem half way through when your Minimal Viable Example suddenly starts working and your realise the flaw, or the rubber-duck nature makes you think it through in a new way.
Because people standing from the sidelines telling me what I should and shouldn't be doing with my spare time leaves more than a little bad taste in my mouth.
I'm not saying that people answering questions are always right or always have the final say, but it's easy to have strong opinions on what's "better for the community" if you're not the one doing the legwork or have to deal with the fallout from that.
I certainly would never be saying that sort of stuff unless I had put in the work. My balls just aren't big enough I guess.