Meta's New AI Chatbot Makes Controversial Statements
Meta's New AI Chatbot Makes Controversial Statements

Meta's New AI Chatbot Makes Controversial Statements

9 august, 20221 minute to read
Follow [Durov's // Code] on Telegram

Last week, Meta released the BlenderBot 3 chatbot in the US. It has already made many blatant statements based on interactions with real people.

Meta's BlenderBot 3 can search the internet to talk to a person about almost everything. It has personality, empathy and the ability to have a long-term memory relating to the conversations it has had.

The ability to absorb information from its interlocutors led the chatbot to call Meta CEO Mark Zuckerberg "too creepy and manipulative" in a conversation with an Insider journalist, while it told Jeff Horwitz, Wall Street Journal journalist, that Trump will "always be" president, and said that it's "not implausible" that Jewish people control the economy.

In messages from some users, the bot admitted to being homophobic. Furthermore, it claimed to be a Pentecostal Christian.

Meta warned users that BlenderBot could still make rude or offensive comments, but the company is working to resolve the issue.

9 august, 2022
Follow [Durov's // Code] on Telegram