Meta is the latest company that has decided to explore the social potential of AI-chatbots.
A screenshot , shared on Twitter by Alessandro Paluzzi, shows what appears to be the intro screen for Instagram’s new feature. The chatbots are said to be able answer questions, provide advice and assist users in writing messages. The company also claims that users can choose from “30 AI personalities to find the one they like best.”
Meta hasn’t made any official plans, but chatbots fit with previous statements about its AI ambitions. Mark Zuckerberg, Meta’s CEO, said in February that Meta is “developing AI persons that can help people in various ways.” The company is also exploring the possibility of making such bots available through text-based conversations “like those in WhatsApp or Messenger.”
Chatbots are also being used by other companies as a social tool. Snapchat launched “My AI”, a chatbot powered by OpenAI ChatGPT, in February. has gained significant popularity on sites such as Character.ai, which allow users to talk and train chatbots that are based on popular fictional character. This has been done with bots that were not designed for social interaction. Microsoft’s Bing chatbot was launched earlier this year. Many users were shocked, delighted, and disturbed by its strange conversational pattern.
It’s difficult for companies to create a bot that is engaging and fun, but does not cross the line to offensive or dangerous interactions. Snap’s “My AI” bot began giving disturbing advise not long after it was released. This included encouraging a person pretending to 13 years old to have sex a 31-year-old ‘boyfriend.
Meta has not made it clear whether they intend to launch these bots or what security measures will be taken. The company declined to comment when we contacted them. Paluzzi is a leaker who has been able to predict upcoming features of apps, such as a BeReal clone, and a Twitter co-author feature.