A world-first, non-invasive AI system can turn silent thoughts into text while only requiring users to wear a snug-fitting cap.
    • hansl@lemmy.worldEnglish
      3·
      7 months ago

      How much accuracy would you be happy with? Anything more than 25% in my book is better than anyone else. And the tech is just getting better.

      How much would it need to be at to beat a polygraph?

    • RainfallSonata@lemmy.worldEnglish
      01·
      7 months ago

      Although DeWave only achieved just over 40 percent accuracy based on one of two sets of metrics in experiments conducted by Lin and colleagues, this is a 3 percent improvement on the prior standard for thought translation from EEG recordings.

      The Australian researchers who developed the technology, called DeWave, tested the process using data from more than two dozen subjects. Participants read silently while wearing a cap that recorded their brain waves via electroencephalogram (EEG) and decoded them into text.

      Yep.

  • badbytes@lemmy.worldEnglish
    91·
    7 months ago

    Spent over 15yrs studying brain activity in EEG MRI and MEG. Seems like a far stretch, given our ability to accurately access signals. Brain is complicated, and signals like EEG are very poor reflector of specific signals. Like when you view city street lights at night. Pretty, but what can you decipher.

  • General_Effort@lemmy.worldEnglish
    92·
    7 months ago

    With further refinement, DeWave could help stroke and paralysis patients communicate and make it easier for people to direct machines like bionic arms or robots.

    The article doesn’t even hint at any use in a justice system. There’s nothing to suggest that this could even in principle be used as a lie detector.

  • TheWonderfool@lemmy.worldEnglish
    6·
    7 months ago

    Ignoring the technology itself, I found it interesting that it has a lot less trouble with verbs compared to nouns (tho the article does not give much information about it).

    Would it mean that humans keeps actions very separate (even if similar), while keeping things and concepts more clustered together? Is being precise on what is happening much more important than clearly specifying the subject and object of the action?

    • FooBarrington@lemmy.worldEnglish
      4·
      7 months ago

      I’d wager that humans have much more neural hardware relating to verbs, since they relate to the things you yourself do over longer periods. Let’s say I clean my bathroom, or my kitchen, or something else - the actions I take are very similar, and my head has to keep my body doing the right thing for long stretches of time. It’s much harder to clean the wrong thing than to clean the thing wrong.

    • Belgdore@lemm.eeEnglish
      1·
      7 months ago

      There are fewer verbs than nouns I’m sure our brains prioritize accordingly

  • phubarr@lemmy.worldEnglish
    4·
    7 months ago

    This is definitely progress, but we need to keep in mind that the particular language a person speaks can significantly influence how a person’s brain works.

  • Matriks404@lemmy.worldEnglish
    52·
    7 months ago

    So is it only useful for people who silently read? Because I don’t see any use case if so, it is not like we think using words, lol.

  • Smoogs@lemmy.worldEnglish
    2·
    7 months ago

    Later on they find the accuracy is about as good as the whole facial recognition fiasco

  • gregorum@lemm.eeEnglish
    1·
    7 months ago

    anyone remember the movie Strange Days? Or the movie Brainstorm?

  • A_A@lemmy.worldEnglish
    28·
    7 months ago

    Eventually this will streamline justice systems : freeing innocents and punishing culprits. At least it will be a powerful tool, just like DNA analysis before it and digital prints were also groundbreaking.