A new AI system’ can translate a person’s brain activity into text. Although it is not a word-for-word transcript, researchers at the University of Texas designed it to capture the gist of what is being thought, albeit imperfectly. Researchers warn that their ‘semantic decoder’ could be used for bad purposes, but also say it might help people who are mentally conscious yet unable to physically speak, such as those debilitated by strokes, to communicate intelligibly again.

  • Gaywallet (they/it)
    link
    fedilink
    811 months ago

    Hopefully incredibly useful as a medical device one day to those which are severely limited in physical movement and speech. Slightly sensationalized title and article, but to be expected.

    • @ailiphiliaOP
      link
      7
      edit-2
      11 months ago

      Slightly sensationalized title and article

      Yeah, maybe a bit. I tried to make it clear in the body that it’s not yet perfect, but it may also be hard for the researchers to communicate such a technology to a wider audience so that everyone understands.

      Edit for an addition: The Guardian wrote about it, citing two experts in the field who are not mentioned in the linked article:

      Prof Tim Behrens, a computational neuroscientist at the University of Oxford who was not involved in the work, described it as “technically extremely impressive” and said it opened up a host of experimental possibilities, including reading thoughts from someone dreaming or investigating how new ideas spring up from background brain activity. “These generative models are letting you see what’s in the brain at a new level,” he said. “It means you can really read out something deep from the fMRI.”

      Prof Shinji Nishimoto, of Osaka University, who has pioneered the reconstruction of visual images from brain activity, described the paper as a “significant advance”. “The paper showed that the brain represents continuous language information during perception and imagination in a compatible way,” he said. “This is a non-trivial finding and can be a basis for the development of brain-computer interfaces.

      That’s extremely impressive if I may say so.

      • Gaywallet (they/it)
        link
        fedilink
        611 months ago

        Yea no fault on you, just journalism doing its thing - they ultimately need to drive clicks and are less likely to fully understand the scope and implications of a device like this.

  • @loki@lemmy.ml
    link
    fedilink
    6
    edit-2
    11 months ago

    I wonder how it works with people that speak multiple language or mixed languages. Are the thoughts in the brain language agnostic? As we imagine the same thing just in different words.

    It’s very interesting research and can help people with good intentions but someone will inevitably use it for their own benefits.

    • @ailiphiliaOP
      link
      311 months ago

      That’s a good question. I understand from the article that the decoder must be trained on the individual, so I guess that might solve the problem? I’m not an expert on this, so I don’t know.

  • @tardigrada@beehaw.org
    link
    fedilink
    311 months ago

    I am wondering whether this technology could enable communication with non-human species. There’s a fair evidence from research that animals have someform of intelligence, e.g., the paper posted yesterday. I mean, if this decoder can be trained on an individual human being’s brain activity, why not on any non-human being’s?