Meta’s new AI chatbot, BlenderBot3, is already courting controversy less than a week after its release for offensive speech, according to reports.
The chatbot, which users can interact with as Meta tests and tweaks the AI, has reportedly repeated far-right and anti-Semitic conspiracy theories.
This is from a fresh browser and a brand new conversation. Ouch. pic.twitter.com/JrTB5RYdTF
— Jeff Horwitz (@JeffHorwitz) August 7, 2022
It has also told a user that it thinks Mark Zuckerberg is “too creepy and manipulative”.