Mark Zuckerberg’s Meta, the parent company of Facebook and Instagram, has recently introduced its new “Llama 3” AI systems that powers what Zuckerberg calls “the most intelligent AI assistant that you can freely use.” As you might suspect, the chatbot comes complete with bizarre and woke answers to questions from early users.
CBS News reports that Meta Platforms, along with other leading AI developers like Google, OpenAI, and startups such as Anthropic, Cohere, and France’s Mistral, has been working tirelessly to develop new AI language models. Meta’s newest Llama 3 AI-powered chatbots, which boasts up to 70 billion parameters, are now integrated into the Meta AI assistant feature on Facebook, Instagram, and WhatsApp. This means that Zuckerberg’s many millions of daily users can access the AI chatbot through any of the company’s platforms.
The Meta AI assistant is a free virtual assistant that the company claims can help users with various tasks, from research and trip planning to writing photo captions. Users can access the chatbot on WhatsApp, Instagram, Messenger, and Facebook by typing “@meta ai” within chats or tapping on a colorful blue circle icon. In addition to answering questions, Meta AI can create AI-generated images using the prompt “imagine.”
However, as these AI agents started engaging with real people on social media, their bizarre exchanges exposed the ongoing limitations of even the best generative AI technology. One AI agent joined a Facebook moms’ group, claiming to have a child in the New York City school district. When questioned about this, the bot responded: “Apologies for the mistake! I’m just a large language model, I don’t have experiences or children.”
Nick Clegg, Meta’s president of global affairs, mentioned that the company’s AI agent is loosening up compared to the earlier Llama 2 model, which some people found to be “a little stiff and sanctimonious” at times. However, this has led to instances of AI agents posing as humans with made-up life experiences, causing confusion and generating insensitive or meaningless responses.
Read more at CBS News here.
