2
$\begingroup$

After 10 days, the metrics on Area 51 seem to indicate that the site is doing well, except for the lack of users ("experts") and visits.

  • 16.2 questions per day (10 questions per day is healthy, 5 questions need works)
  • 85% answered (90% is healthy, 80% need works)
  • 1.7 answers per questions (2.5 is healthy, 1 needs works)
  • 108 visits/per day (1500 is healthy, 500 need works)
  • 25 users with 200+ rep (250 users is healthy)

Yet, it doesn't seem as though the site is actually healthy, as the question Have we brought something new to the network?, suggests. It seems that there is a high chance of closure, despite the stats supposedly being on our side.

However, interest in AI is not going away. It hasn't gone away the last time an AI StackExchange was proposed and shut down. We still need to deal with the humanities aspect of this field...and to help try to "demystify" the field of artificial intelligence. Until we get a good answer to this problem of dealing with artificial intelligence in a unified manner, all that's going to happen is a fourth AI proposal.

Now, technically, there are many sites that can handle AI questions. I mentioned about them in this Area51 post, and how a programmer could build an AI simply by cobbling together answers from multiple StackExchange sites. I also stated that the de facto solution to the AI problem is a very problematic one:

The AI programmer is stuck traveling from one site to another (though Google is helping out), asking questions on each individual site, to try and figure out how to accomplish his single goal. This AI programmer is a migratory beast, out to stitch together random answers on Stack Exchange into a coherent 'whole' that he can then use. ... Is it sensible for AI programmers to be dependent on 6 to 8 different sites, at any given time? Is it sensible for these 6 to 8 sites to pander to these AI programmers instead of answering other questions? Or is it better to instead consolidate AI-specific questions from these 6 to 8 different sites, onto a single site that is easy to browse and look up? Some overlap between the communities may be inevitable, but it's better than the current status quo.

This idea has been resurrected by by kenorb recently very recently:

If it fails, the backup is to have 20 different accounts and ask the AI questions across the different sites such as Stats.SE, CSTheory.SE, CogSci.SE, Philosophy.SE, Worldbuilding.SE, SO.SE, CS.SE, HSM.SE, Robotics.SE, GameDev.SE, with no real AI experts focused in one place.

I have built "StackAI", a StackExchange aggregator of several AI-related StackExchange sites which I may plan on upgrading in the future to include other AI-related StackExchange sites (such as CogSci, Robotics, etc.). But I'm still dubious on whether this site would indeed be the best solution for our crisis. It's a good one-stop directory to help a migratory questioner know where they can ask their questions, but the process is still rather inefficient and we lack the consolidation that a standard AI site could provide for us...and, of course, I doubt that these other sites would even appreciate the influx of these migratory questioners.

Is there a better way to deal with the migratory AI questioner? Even if it is to just improve StackAI?

I get discouraged and cynical rather easily, so maybe I'm being overly pessimistic about this site. But I do care about helping out people, and will want to help them out, even if "ai.stackexchange.com" dies again.

$\endgroup$
  • 1
    $\begingroup$ Despite what other think, I believe most of the questions would be actually off-topic on other sites (because they're not so technically oriented for these specific sites like stats or data science), so migration probably won't happen, maybe just few of them. So they're likely to be just exported and dumped into zip file and forgotten as the previous AI dumps. $\endgroup$ – kenorb Aug 13 '16 at 19:58
6
$\begingroup$

I think it's too early to worry about what to do if the site closes. We're different from previous AI sites in a couple important ways, and though we're not completely free of issues, we're making progress. For what it's worth, the target statistics on Area 51 are what you should expect from a site about to fully graduate (not one about to continue into public beta).

Without this site, the topic of artificial intelligence is indeed split across the network. Fortunately, each site's scope is fairly well-defined, so even if finding a good site is tricky, checking whether a question is OK for a given site is quite doable. Meta Stack Exchange can help find a good home for questions - they even have a tag for such inquiries, and I think a question for routing people to the right AI site would be in order.

Again, don't get too worried. We'll cross that bridge if we come to it.

$\endgroup$
  • $\begingroup$ Additionally, there is now less pressure on a site to 'avoid closure' - as long as we are continuing to have questions and answers, and the community is stable or growing, public beta can now be effectively permanent. $\endgroup$ – Rory Alsop Aug 15 '16 at 6:55

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .