12
$\begingroup$

We've this old thread at Area 51 (related to older site proposal which failed):

In general accepted post says:

First of all, artificial intelligence is a much broader term than machine learning. While at the same time Cross Validated is not about machine learning, but about statistics.

Is it still valid point? Can we elaborate on this further more?

How this site is different from Cross Validated? Do we have now more arguments to it?

$\endgroup$

4 Answers 4

Reset to default
3
$\begingroup$

Clearly, Cross Validated is about statistics - it's even in the URL, stats.stackexchange.com. They're a very math-heavy and calculation-oriented site. MathJax is enabled there, and every question I scanned from their front page involves code or mathematical formulae. is their third most popular tag at the moment, and questions in it are about the stats/math involved in machine learning.

Questions here are not expected to involve that level of detail. MathJax is not enabled here, and that might be purposeful. Our questions should be about the science - not so much the technology or math or implementation of - artificial intelligence. (For machine learning implementation, see Data Science.)

$\endgroup$
4
  • 4
    $\begingroup$ MathJax probably should be enabled here. If you say we're about "the science of AI", you are pretty limited in talking about (real) science without using math. $\endgroup$
    – mindcrime Mod
    Commented Aug 18, 2016 at 15:31
  • $\begingroup$ @mindcrime I added a link to the discussion about MathJax formatting support. $\endgroup$
    – Ben N
    Commented Aug 18, 2016 at 15:32
  • $\begingroup$ I wonder how to make science without math? As soon as you want to quantify something you'll need it? So could you maybe elaborate a bit more on what you mean by "the science - not so much math - of artificial intelligence"? What is the science of AI? $\endgroup$
    – NoDataDumpNoContribution
    Commented Sep 6, 2016 at 19:32
  • $\begingroup$ @Trilarion That's a really good point, and the more I think about it, the more I feel that we need a better way to describe our scope. I encourage you to help us by participating in this other meta discussion! $\endgroup$
    – Ben N
    Commented Sep 6, 2016 at 20:30
4
$\begingroup$

AI is broader than Machine Learning and Statistical Learning. Yes, the probabilistic / statistical stuff dominates the conversation these days, but AI includes rule based systems, expert systems, symbolic processing, logic programming and other things that would not be on-topic at Cross Validated (or Data Science).

We've also been saying that we have more of a focus on the philosophy of AI and the humanities related aspects, as opposed to strictly the technology.

$\endgroup$
1
  • $\begingroup$ Being broader implies somehow that there is overlap. The question is probably how big the overlap is compared to the total scope. It also might be good to have standard migration paths to the overlapping sister sites, because askers will inevitably miss the finer points in distinguishing the scopes. $\endgroup$
    – NoDataDumpNoContribution
    Commented Sep 6, 2016 at 19:45
3
$\begingroup$

I'm of the opinion that we should allow ML and AI research-style questions here, of the sort that would also could be on-topic at Cross Validated but would be less likely to hit their intended audience there than they would here.

That is, I don't think there is a difference in topics so much as there is a difference between clusters of people who care about those topics, and the perspectives that they bring and the sort of questions and answers that they'll consider interestingly on-topic.

$\endgroup$
1
  • $\begingroup$ I think the comment on 'perspective' nails it. $\endgroup$
    – NietzscheanAI
    Commented Aug 12, 2016 at 19:27
-1
$\begingroup$

While at the same time Cross Validated is not about machine learning, but about statistics.

This is not a valid point, as all machine learning questions are on-topic on CV.

It's unclear to me what extent non-statistical AI is on-topic on CV, so I asked there https://stats.meta.stackexchange.com/questions/4257/what-is-our-stance-on-questions-about-non-statistical-artificial-intelligence If no, it makes sense to have a (non-stat?) AI site. If yes, I think we should merge and rename CV.

$\endgroup$

You must log in to answer this question.

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