In a world-first, researchers from the GrapheneX-UTS Human-centric Artificial Intelligence Centre at the University of Technology Sydney (UTS) have developed a portable, non-invasive system that can decode silent thoughts and turn them into text.
  • mustbe3to20signs@feddit.de
    92·
    7 months ago

    This buzzword labeled bunch of algorithms isn’t reading minds, it’s interpreting ones EEG.
    Sensationalist headlines are a strong indicator of crap articles.

    • 🅿🅸🆇🅴🅻@lemmy.world
      4·
      7 months ago

      I’m sorry, but what is the difference? If you know the subject is “thinking” of a phrase, and the algorithm translates the EEG during that time into words, isn’t it mind reading?

      • SSUPII@sopuli.xyz
        3·
        7 months ago

        it is, as you are reading what is going on in the brain. Some think “reading minds” is only wireless like in current fiction.

        • mustbe3to20signs@feddit.de
          3·
          7 months ago

          I wasn’t thinking of science fiction at all.
          But this isn’t mind reading (which is impossible) it’s a statistical model giving the most likely answer based on an EEG.

          • angrystego@lemmy.world
            3·
            7 months ago

            Whether mind reading is possible or not depends on how you define it. I suspect your definition is different from that of other people in this comment section. It covers how I define mind reading. So how do you define it for yourself?

            • mustbe3to20signs@feddit.de
              3·
              7 months ago

              I go with a literal definition. Being able to identify the thought of a random stranger without calibration or them focussing on one specific thought.
              Don’t get me wrong, this is great for people who are unable to communicate otherwise. But in the end it is still an interpretation and therefore not error-proof.

              • angrystego@lemmy.world
                2·
                7 months ago

                Yeah, so by your definition this is definitely not mind reading. I wouldn’t expect it to be error-proof ever. Not even usual communication is. And this is just a beginning - their success rate seems to be 40 %.

                • 0ops@lemm.ee
                  4·
                  7 months ago

                  Honestly, I’m a good reader (regular old text reader, not mind), but I’m not perfect. Imo the line for this should be drawn at “accurate enough to be practically useful”