Meta's New AI Chatbot , Spouts Uncomfortable , Truths and Lies.<br />On August 5, Meta's conversational <br />artificial intelligence prototype, BlenderBot, <br />launched on a demo website.<br />On August 5, Meta's conversational <br />artificial intelligence prototype, BlenderBot, <br />launched on a demo website.<br />'The Guardian' reports that members of the public are invited to help test the AI and share feedback with Facebook's parent company, Meta.<br />According to Meta, the bot <br />is capable of conversing <br />with users on nearly any topic.<br />'The Guardian' reports that the bot <br />uses material it finds on the internet <br />to offer users some interesting responses.<br />According to BuzzFeed, when asked <br />about Mark Zuckerberg, the AI replied, "he is a good businessman, <br />but his business practices <br />are not always ethical.".<br />According to BuzzFeed, when asked <br />about Mark Zuckerberg, the AI replied, "he is a good businessman, <br />but his business practices <br />are not always ethical.".<br />It is funny that he has all this money <br />and still wears the same clothes!, BlenderBot, via BuzzFeed.<br />Previous chatbots have tended to spout racist conspiracy theories, often forcing the company to apologize for , "inappropriate and reprehensible words and images.".<br />Previous chatbots have tended to spout racist conspiracy theories, often forcing the company to apologize for , "inappropriate and reprehensible words and images.".<br />In South Korea, a chatbot designed to mimic a 20-year-old college student was suspended after repeating racial slurs and anti-LGBTQ+ remarks.<br />In South Korea, a chatbot designed to mimic a 20-year-old college student was suspended after repeating racial slurs and anti-LGBTQ+ remarks.<br />'The Guardian' points out that users who log on to BlenderBot's test page need to acknowledge that the AI is , “likely to make untrue or offensive statements.”.<br />According to Meta, the AI has, “a high propensity to generate toxic language and reinforce harmful stereotypes, even when provided with a relatively innocuous prompt.”
