Comment by bravura | Hacker News Reader

Comment by bravura | Hacker News Reader

Parent Full thread

The main reason I launched my Q+A site is because, if Area 51 had their druthers, they would have fragmented the community into five niches:

machine learning (http://area51.stackexchange.com/proposals/7607/machine-learn...)

NLP (http://area51.stackexchange.com/proposals/2761/natural-langu...)

statistical analysis (http://area51.stackexchange.com/proposals/33/statistical-ana...)

data mining (http://area51.stackexchange.com/proposals/4229/data-mining)

AI (http://area51.stackexchange.com/proposals/6607/artificial-in...)

This niching is completely wrongheaded. As I wrote in a comment on this blog post, the last thing we need is for ML people, NLP practitioners, and statisticians to communicate less. The fact that these groups attend different conferences is a BUG not a feature. My site was designed so that these adjacent fields can crosspolinate information.

As we’ve learned from StackOverflow, it’s better to pick a broad topic and have a lot of experts in one place, communicating and helping with each other’s problems.

As we’ve learned from StackOverflow, it’s better to pick a broad topic and have a lot of experts in one place, communicating and helping with each other’s problems

Yeh, solid point. All the current sites are highly active because of their broad domain overview.

I don't think that it's a foregone conclusion that Area51 generates niche sites, just because it generates niche site proposals. I for one believe that it's the big blockbuster stackoverflow-size sites that will actually make it to the end of the Area 51 process and will actually get created.

And, if I were a betting man, I wouldn't bet against StackOverflow's monthly 7 million users. In fact I would be betting on the statistics Stack Exchange site launching in the next couple of weeks and getting critical mass very very quickly.

Not without an existing community to build off of you won't. You'll fail miserably. The people in the Area51 proposals are generally meeting there for the first time — they didn't know each other already.

MetaOptimize is building on top of an existing well-targeted community like MathOverflow did. Really, just like you did originally with your and Jeff's blog communities. AskMetafilter is successful because it was an outlet for an existing community — it wasn't attempt to start a new one.

Area51 is scavenging small clumps of people that happen to overlap with your existing main sites. That your sites are so QA-focused makes it very difficult for an actual community to grow within it — there's no outlet for chat and friend-making, for fucks sake even your moderation site is the same QA format! At least you've integrated it a bit more so that Meta isn't a completely separate community with it's own account system…

I also wonder when will they incorporate more "social network" features into SO. I feel how it is needed...

>if Area 51 had their druthers, they would have fragmented the community into five niches

I agree. I would see these new stackexchange clones every other day at /r/machinelearning. It would have hurt the ML community if all of the sites started competing for eyeballs.

Anyway thanks a lot for the initiative. I wish you all the best.

Reply on news.ycombinator.com