home.social

#ameca — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #ameca, aggregated by home.social.

  1. I'm amazed at how people were falling over themselves at a recent conference to see Ameca as a real super smart nearly alive robot. They really want to believe.

    #robot #ameca #ai #chatgpt #llm

  2. No one lies with such grace as those who believe their lie…

    Ameca speaks multiple languages and translates. youtube.com/watch?v=wJyYMGswc1
    >This #ameca demo uses #gpt3 for conversation and translation. DeepL is used for language detection.

  3. This video is from a year ago. #AMECA is my favorite robot, *because* it's lifelike. I don't find it "creepy," I don't respond with words like "yikes," because I'm not threatened by it in any way.

    Let's discuss something that comes up in this video, and surrounding all types of AI and even computer animation: The "#uncannyvalley." This is a concept that as we approach a full simulacrum of a human face or body, people respond more positively the closer it gets to reality, until it gets too close. Then responses nosedive into a valley on the graph, where people recoil as if they just saw a ghost or an alien.

    Folks, the "uncanny valley" is "othering" bullshit from the depths of our unprocessed psychology about what it means to be #human. It's the exact same distrustful reaction as many people have when encountering people from a different nation or culture. i.e. #racism

    It's the same psychology of treating the "other" as "worse," because people are afraid that it might be *better.* We see this with hardworking immigrants now. Local-born Americans are indeed threatened because they know that immigrants and refugees were willing to go through absolute hell, to get what we got just by being born.

    As with immigrants who speak a different language, there's also an inherent mistrust of "uncanny" robots, as if they are inherently dishonest, by pretending to be something they're not.

    It's ridiculous, because even a perfect-looking or perfect-sounding humanoid robot is obviously not a human being (and not pretending to be).

    Just as a foreign-born non-English speaking immigrant is obviously not a native born citizen (and isn't pretending to be).

    Eventually such #robots may evolve consciousness and become *smarter than us.* Just as many immigrants end up *working harder than us.* And that's the basis of the fear. It's inadequacy--not being able to keep up--and it's ugly.

    I'll say it again. The uncanny valley is #robotracism. It's the exact same impulse that drives anti-immigrant racism. The jury is out as to whether society can coexist with non-human robots. But they're coming, and we'd better figure it out. It's a guarantee that people will physically attack them, just as they attack immigrants now. How we will deal with such attacks remains to be seen.

    I hope they are treated as hate crimes, and not mere property crimes.

    AMECA's inventor claims we won't see this type of humanoid robot in general use within a decade or even two. I disagree. It's only a question of cost and that gets driven down by mass production, which could start in the next few years.

    Even now if an AMECA (or similar) robot were functional at performing basic tasks, and even if it were only connected to the current version of #chatgpt it would be cost-effective at anything less than a year's salary for a human being. Which is to say, there would be a large market for this machine *right now* if it were available for under $50k.

    youtu.be/LzBUm31Vn3k

  4. Ameca sieht aus wie ein Roboter, besitzt aber ein menschlich wirkendes Gesicht, mit dem er menschliche Mimik nachahmen kann.
    Engineered Arts bringt Roboter "Ameca" realistische Gesichtsausdrücke bei