It’s not everyday that people can pull out their phones and casually send messages to Paris Hilton, Snoop Dogg or Kendall Jenner, let alone seek advice from them in their areas of expertise. However, thanks to Meta’s latest collection of artificial intelligence chatbots, this is now a reality. The company has developed 28 AI assistants that users of WhatsApp, Messenger, and Instagram can engage in conversations with.
Meta has also created profiles for each of its chatbots, which are currently limited to text-based interactions, allowing them to have a presence beyond the chat interfaces it is currently limited to. Meta has also reported that they’re actively working on giving these bots a voice, with plans to achieve that by next year. The parent company of Mark Zuckerberg’s empire is reportedly even searching for screenwriters who can craft characters and other narrative content that can engage an audience.
According to The Information, Meta is paying the featured celebrities up to $5 million over two years for just six hours of studio work, where their likeness is captured to be utilised for the new AI assistants. Earlier this week, a video from one of these AI chatbots, designed for younger users with distinct personalities, made some waves as Kendall Jenner’s alter ego, Billie. Billie’s unique “personality” is that of a big sister, with claims to offer advice to users.
Looking at the chatbot profiles, you’ll see some posts with ‘Imagined with AI’ watermarks. This is naturally causing some confusion, especially in the posts featuring the Kendall Jenner we all know. People are asking questions like, “Is this picture made by AI?” and “That’s obviously Kendall, not AI”. This confusion about what is real and what is AI is likely all part of the plan for Meta. However, those watermarked AI images (and the absence of the same watermark on the pictures and videos of “Billie” herself) do suggest that the visuals of Billie currently are likely from the six-hour studio session with the actual Kendall Jenner.
This race to “counterfeit people” worries many observers, who are already concerned about recent developments made in large language model (LLM) research such as ChatGPT and Llama 2 (its Facebook counterpart).
“There are a number of thinkers who are denouncing these major groups’ deliberately deceptive approach,” said Ibo van de Poel, a professor of ethics and technology at Delft University of Technology in the Netherlands.
But what’s the purpose of all this? According to van de Poel, the more users feel like they’re conversing with a real person, ‘the more comfortable they’ll be, the longer they’ll stay, and the more likely they’ll come back.’ In the realm of social media, the time spent on platforms like Facebook, as well as engagement with its ads, directly translates to revenue. These chatbots begin to reduce the need for your friends and family to be part of the interactions and the time you spend on the platform.
Meta’s recent introduction of AI sticker generation, allowing users to turn text prompts into “multiple unique, high-quality stickers in seconds” didn’t go as smoothly as expected.
So it remains to be seen what the activation and continued development of these chatbots for millions of users will mean for Meta, social media, human beings, and the world as a whole.