Google’s AI Chatbot Is Trained by Humans Who Say They’re Overworked, Underpaid and Frustrated::(Bloomberg) – Google’s Bard artificial intelligence chatbot will answer a question about how many pandas live in zoos quickly, and with a surfeit of confidence.Most Read from BloombergTesla Investor Rode a 14,800% Gain Thanks to 27-Year-Old AnalystSingapore Passport Is World’s Most Powerful, Replacing JapanMassive US Oil Caverns Sit Empty and Will Take Years to RefillMeet the Billionaire Who Convinced Messi to Pick MLS Over Saudi ArabiaWall Street Gets New ETF Offering 100% Downside ProtectionE

  • Dojan
    link
    English
    16
    edit-2
    1 year ago

    Tried Bard. It was fairly impressive with its image recognition, identifying the breed of my dog, but then confidently making up things like “he has very beautiful light blue eyes” (they’re brown) and “white feet” (feet weren’t visible).

    Prompted further on things it continued to make stuff up and proclaiming it with confidence, but at least it didn’t try to gaslight me when I pointed out that it was wrong.

    Basically, these products aren’t ready. People will view them as a source of information, which they’re not. They’re fancy text completion, not more.