GNU social JP
  • FAQ
  • Login
GNU social JPは日本のGNU socialサーバーです。
Usage/ToS/admin/test/Pleroma FE
  • Public

    • Public
    • Network
    • Groups
    • Featured
    • Popular
    • People

Conversation

Notices

  1. Embed this notice
    isaac io schankler (io@merveilles.town)'s status on Friday, 14-Apr-2023 09:50:29 JST isaac io schankler isaac io schankler

    My semi-spicy take is that computer science depends on analogy, metaphor, and subtext just as much as the humanities or the arts. But because so many people in tech are willfully oblivious to this, they're especially prone to the kinds of cognitive distortions that people in the humanities and the arts are trained to recognize and understand.

    What makes a neural network "neural"? What assumptions does that analogy carry? What does that encourage us to see? What does that prevent us from seeing?

    In conversation Friday, 14-Apr-2023 09:50:29 JST from merveilles.town permalink
    • Embed this notice
      Prince Lucija (DRAGX̶FUNK) (luka@sonomu.club)'s status on Saturday, 15-Apr-2023 19:02:26 JST Prince Lucija (DRAGX̶FUNK) Prince Lucija (DRAGX̶FUNK)
      in reply to

      @io the interesting question is why are a majority of people in tech oblivious or even willfully oblivious to these cognitive distortions

      In conversation Saturday, 15-Apr-2023 19:02:26 JST permalink
      clacke likes this.
    • Embed this notice
      nnirror (nnirror@post.lurk.org)'s status on Saturday, 15-Apr-2023 19:03:09 JST nnirror nnirror
      in reply to

      @io My response obviously does not speak to the last three questions in your post. Those questions by themselves could be a PhD thesis. I think the questions you're asking are very interesting. I just don't think it's that unreasonable to label this sort of system "neural". Wouldn't it be more absurd if the inventors decided not to name it after a brain, distortion and all?

      In conversation Saturday, 15-Apr-2023 19:03:09 JST permalink
      clacke likes this.
    • Embed this notice
      nnirror (nnirror@post.lurk.org)'s status on Saturday, 15-Apr-2023 19:03:10 JST nnirror nnirror
      in reply to

      @io

      Googles "how does a neural network resemble a brain"

      From MIT: (https://news.mit.edu/2017/explained-neural-networks-deep-learning-0414)

      "Modeled loosely on the human brain, a neural net consists of thousands or even millions of simple processing nodes that are densely interconnected. Most of today’s neural nets are organized into layers of nodes, and they’re “feed-forward,” meaning that data moves through them in only one direction."

      Obviously it's not a perfect analogy. But, I think, given that this type of system needed a name that was a few words maximum, it seems like not that unreasonable? This sort of "distortion" is just par for the course when talking about labeling things, no?

      Maybe I'm wrong, not trying to be a tech-bro-reply-guy, just genuinely curious

      In conversation Saturday, 15-Apr-2023 19:03:10 JST permalink

      Attachments

      1. Domain not in remote thumbnail source whitelist: news.mit.edu
        Explained: Neural networks
        “Deep learning,” the machine-learning technique behind the best-performing artificial-intelligence systems of the past decade, is really a revival of the 70-year-old concept of neural networks.
    • Embed this notice
      Sexy Moon (moon@shitposter.club)'s status on Saturday, 15-Apr-2023 19:04:42 JST Sexy Moon Sexy Moon
      in reply to
      @io my spicy take is people that get seriously stuck on metaphors for things like perceptrons and neural nets are probably not that smart and we shouldn't change things to cater to them.
      In conversation Saturday, 15-Apr-2023 19:04:42 JST permalink
      clacke likes this.
    • Embed this notice
      cleopatra@c.im's status on Saturday, 15-Apr-2023 19:04:43 JST Cleopatra Cleopatra
      in reply to
      • Sexy Moon

      @Moon @io
      There used to be a whole trend of professional opinion-havers taking those new terms (such as 'neural network' or 'selflearning') very seriously and spinning a whole sloppy argument out of their misunderstandings.

      Haven't seen that in a while, thankfully. But it did leave me with a kneejerk 'please only comment on tech if you can code' reaction

      In conversation Saturday, 15-Apr-2023 19:04:43 JST permalink
    • Embed this notice
      clacke (clacke@libranet.de)'s status on Saturday, 15-Apr-2023 19:04:43 JST clacke clacke
      in reply to
      • Sexy Moon
      • Cleopatra
      @Cleopatra @Moon @io Like Deepak Chopra and "Quantum Consciousness", any model, any naming, will have people who take it and just run with it.
      In conversation Saturday, 15-Apr-2023 19:04:43 JST permalink
    • Embed this notice
      Josh Wells (josh@fediscience.org)'s status on Saturday, 15-Apr-2023 19:05:00 JST Josh Wells Josh Wells
      in reply to

      @io
      Jargon and milestones throughout CS comes from the arts.

      First programmer? Daughter of the poet Byron.

      Robots? From a Czech play in the 1920s.

      Cyber[insert concept]? Famous science fiction novel.

      And so on … obliviousness to these cultural debts is what makes the torment nexus possible

      In conversation Saturday, 15-Apr-2023 19:05:00 JST permalink

      Attachments


      1. https://fediscience.org/system/media_attachments/files/110/193/499/397/338/979/original/aa661ac7096ea9e8.jpeg
      clacke likes this.

Feeds

  • Activity Streams
  • RSS 2.0
  • Atom
  • Help
  • About
  • FAQ
  • TOS
  • Privacy
  • Source
  • Version
  • Contact

GNU social JP is a social network, courtesy of GNU social JP管理人. It runs on GNU social, version 2.0.2-dev, available under the GNU Affero General Public License.

Creative Commons Attribution 3.0 All GNU social JP content and data are available under the Creative Commons Attribution 3.0 license.