1. 42
    Tag proposal: AI ai meta

There are a reasonable number of AI stories submitted here, but no good tag for it. Some fit in cogsci, but a lot of the AI stuff filed there is a bit of a stretch to call cognitive science, and just submitted there for lack of a better place (example, example). And machine learning often gets misfiled under ml, which is supposed to be for the ML family of programming languages.

Some more examples of things that imo should go under an ai tag: 1, 2, 3, 4, 5, 6.

Besides being useful to those of us interested in AI, it could conceivably also be useful for people who are bored of the AI hype and would like to filter it. :-)

  1.  

  2. 18

    If we do this we should also rename ml to ml-language so that people don’t continue to misfile it.

    ((metalanguage-language!)

    1. 3

      Agreed on both counts.

      1. 2

        I think metalanguage would be more clear than ml-language.

        1. 1

          I’d prefer if we kept ml as ml and called “ai” or “ml” or whatever is this weeks name for applied statistics “statistics” instead.

          1. 4

            But “ai” is more broad than machine learning / applied statistics. I would expect an “ai” tag to cover things like evolutionary algorithms, expert systems, etc, as well as your more typical (these days) machine learning.

        2. 10

          Added

          1. 7

            Very much agreed. I submitted a post a day or two ago about pathfinding that would’ve benefited from such a tag!

            1. 2

              IME lobsters isnt the best place for AI stuff. Youre better off going to a subreddit or arXiv or even HN and certain groups of people on twitter. Its unfortunate but true.

              1. 4

                But arxiv isn’t really a place for discussion. There is a MachineLearning subreddit but it’s pretty obsessed with Deep Learning, so any machine learning that’s not that gets treated in a rather hostile way. HN sometimes has people who know what’s going on, but you can pretty often get incorrect information arrogantly presented as fact.

                Twitter discussions are wonderful, but I feel are really ephemeral and hard to point to. Recently, Yoav Goldberg ranted https://medium.com/@yoav.goldberg/an-adversarial-review-of-adversarial-generation-of-natural-language-409ac3378bd7 about the flag-planting nature of current machine learning practice and it provoked a ton of back and forth by different groups of people. All that is really invisible if you aren’t following the right people.

              2. 1

                Is data-science too broad?