Boston Dynamics turned its robot dog into a talking tour guide using AI as seen in a somewhat unsettling video posted by Boston Dynamics. Boston Dynamics used OpenAI’s ChatGPT API, along with some open-source large language models (LLM) to carefully train its responses. It then outfitted the bot with a speaker, added text-to-speech capabilities, and made its mouth mimic speech “like the mouth of a puppet.”

The version speaking in a British accent and the one of a Shakespearean time traveller had me 😂 but it’s certainly a little unsettling overall.

  • silverbax@lemmy.world
    link
    fedilink
    English
    arrow-up
    34
    arrow-down
    2
    ·
    edit-2
    1 year ago

    As a minor aside to the story, I find it interesting that the devs tried to make the robotic dog ‘talk’ by opening and closing it’s ‘mouth’ and failed. Cinema such as Star Wars has proven that humans don’t need an android or robot to actually move its mouth when it speaks in order to have a conversation.

    I also can’t help but wonder what has happened at Boston Dynamics - these robotic dogs were/are a wonder, but they aren’t new. They are scary with how agile they are, but it doesn’t seem like this is anything different than a thousand other ‘I integrated Chat GPT into my R/C car’ type things that have come out over the past year. Has innovation at Boston Dynamics stalled?

    • GreyBeard@lemmy.one
      link
      fedilink
      English
      arrow-up
      23
      ·
      1 year ago

      Boston Dynamics YouTube channel has been filled with silly videos. Often times they are duel function. 1. Build brand awareness through fun videos, and 2. Show the versatility of the onboard systems. In this case they are showing of the ability to navigate a real world human environment and the sensors/cameras that can be fed into other systems for advanced decision making and planning.

      • Bebo@literature.cafeOP
        link
        fedilink
        English
        arrow-up
        7
        arrow-down
        1
        ·
        1 year ago

        Also, the robots embracing the “personalities” was interesting (for someone like me who doesn’t have technical knowledge of LLMs etc) as well as entertaining.

        • kingthrillgore@lemmy.ml
          link
          fedilink
          English
          arrow-up
          1
          ·
          1 year ago

          They also uploaded this video a day after they showed off new functionality on their “Stretch” robot which more directly impacts livelihoods, as Stretch isn’t cute. Stretch is meant to replace menial labor.

          Spot is cute and proactive. Stretch is what Boston Dynamics is actually selling.

          • Bebo@literature.cafeOP
            link
            fedilink
            English
            arrow-up
            2
            arrow-down
            1
            ·
            1 year ago

            From what I understood, a short prompt regarding a personality was provided based on which the LLM generated the lines which were converted into speech conveyed to the listener through speakers. (If some technicalities are incorrect feel free to correct me). I used “embraced” kind of metaphorically. The robots themselves didn’t literally embrace a personality.

            • Z4rK@lemmy.world
              link
              fedilink
              English
              arrow-up
              1
              arrow-down
              2
              ·
              1 year ago

              Then I agree. I guess that’s why I didn’t find this very interesting - you could strap speakers and ChatGPT to anything really, it has very little to do with the robot.

              • R0cket_M00se@lemmy.world
                link
                fedilink
                English
                arrow-up
                1
                ·
                1 year ago

                Considering the robot is capable of locomotion and manipulating things with it’s arm, it’s not impossible to think something like this is the first step toward training an LLM on some specific physical task and giving it the robotics to accomplish the goal.

                You seem like the kind of person who hates the idea of AI so you go out of your way to make it seem like it’s just soooo boooring to you.

                • Z4rK@lemmy.world
                  link
                  fedilink
                  English
                  arrow-up
                  1
                  ·
                  1 year ago

                  Lol, I love AI and develop with it almost every day. AI wasn’t the problem here, it was just weird to me that it was Boston Dynamics making this video. It was really surprising tbh. It’s a video I expect from someone doing this on their hobby RC car project or something.

    • Merwyn@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      11
      ·
      1 year ago

      They are trying to move from “making an impressive video for the show” to “solving actual, real and usefull applications”. So this take obviously a long time to produce new results that show this.

  • Eggyhead@artemis.camp
    link
    fedilink
    arrow-up
    7
    ·
    1 year ago

    Someday we’ll all be dead and robot dogs will be giving tours to nothing. Some alien species will come across our planet and learn all about us through the ways in which the robots interact with nobody.

  • afraid_of_zombies@lemmy.world
    link
    fedilink
    English
    arrow-up
    9
    arrow-down
    2
    ·
    1 year ago

    Robots give me enough problems when they are stationary, I am not looking forward to the day I am dealing with a schizophrenic gazelle made of metal hurtling itself around a factory floor.

  • AutoTL;DR@lemmings.worldB
    link
    fedilink
    English
    arrow-up
    3
    ·
    1 year ago

    This is the best summary I could come up with:


    In a somewhat unsettling video posted by Boston Dynamics, we see its robot dog outfitted with a top hat, mustache, and googly eyes as it chats with staff members in a British accent, taking them on a tour of the company’s facilities.

    Follow me, gentlemen.” As shown in the demo, Spot is capable of answering questions and even opens its “mouth” to make it seem like it’s actually speaking.

    To make Spot “talk,” Boston Dynamics used OpenAI’s ChatGPT API, along with some open-source large language models (LLM) to carefully train its responses.

    Matt Klingensmith, the principal software engineer at Boston Dynamics, says the team gave Spot a “very brief script” for each of the rooms at its facilities.

    “We’re excited to continue exploring the intersection of artificial intelligence and robotics,” Klingensmith writes in a post on Boston Dynamics’ site.

    While Spot may sound and look pretty silly during the video, it’s still a bit hard not to think about the dog-like robot’s ability to open doors and spy on people.


    The original article contains 482 words, the summary contains 170 words. Saved 65%. I’m a bot and I’m open source!

  • iAvicenna@lemmy.world
    link
    fedilink
    English
    arrow-up
    2
    ·
    1 year ago

    1- Convince the dog to act like a racist bigot insulting you 2- Sue Boston Dynamics 3- Profit