Key Insights
- Meta trialing AI chatbots with distinct “personas” to increase user engagement.
- One persona being explored: a surfer providing travel tips, inspired by CEO Mark Zuckerberg’s passion for surfing.
- Rise of ChatGPT’s popularity motivating big tech companies like Meta to develop their own conversational AI services.
Meta Corporation is developing several chatbots with artificial intelligence (AI) with images of various personalities, including Abraham Lincoln, reports the Financial Times.
What Will Meta’s AI Chatbots Offer?
Sources familiar with the matter say that Meta employees have referred to the chatbots as “characters” and they take on the form of different personalities. Chatbot prototypes are currently under development, and final products can have human-level discussions with users.
One of the bots speaks like former US President Abraham Lincoln, while the other gives advice to travelers in the voice of a surfer. Their goal is to offer recommendations and new search features, and to be “a fun product for people to play with.” The launch of the bots should take place in September.
Back in February, news emerged about Meta’s intentions to develop an AI-powered chatbot. CEO Mark Zuckerberg posted a message that Meta plans to allocate resources to develop its own chatbot. However, he did not provide any specific details on this topic, except that the company is going to pay increased attention to AI specialists.
Following the success of the popular ChatGPT chatbot, Meta joined other tech giants like Google and Microsoft in finding inspiration to create their own AI-based chatbot.
Read also: What are the Best AI Stocks to Buy Now in 2023?
AI lawsuit against Meta
Last month, Meta and chatbot developer OpenAI faced a lawsuit from writers Christopher Golden and Richard Kadri. According to their claim, both companies’ AI models were trained using the works of the authors. The authors’ works were obtained from pirated sites that have been historically utilized for AI training purposes.
However, the authors did not consent to the use of their copyrighted books as teaching material for ChatGPT. At Meta, they used the works of authors to train the LLaMA (Large Language Model Meta AI) language models.