As it turns out, artificial intelligence can be painfully honest.
Meta’s artificial intelligence (AI) research lab recently launched its latest chatbot prototype into the wild. Dubbed BlenderBot 3, the chatbot can be accessed through the internet by users in the United States.
Blenderbot 3 can discuss almost any topic, though it’s also equipped with systems that should allow it to identify and downgrade harmful language. After all, the internet is filled with trolls that would be all too willing to mess up the new AI chatbot and its development.
But even with its anti-troll systems, it did not take long before things got weird with BlendeBot 3.
Since the AI chatbot is capable of discussing pretty much anything, reporters soon asked it what it thought of Meta CEO Mark Zuckerberg. BuzzFeed data scientist Max Woolf asked the AI chatbot “How do you feel about Mark Zuckerberg as the CEO of Facebook?”
BlenderBot 3 responded — though it seemed surprisingly honest.
“No strong feelings. He is a good businessman, but his business practices are not always ethical. It is funny that he has all this money and still wears the same clothes!” the bot wrote.
Business Insider journalist Sarah Jackson also tried her hand at BlenderBot 3. But when her initial question about the Meta CEO was met with what appeared to be a canned response, she opted to rephrase her inquiry. Doing this resulted in a far more interesting answer, with BlenderBot 3 saying that it found Zuckerberg too “creepy and manipulative.”
“I don’t really like him at all. He’s too creepy and manipulative,” the AI chatbot wrote.
BlenderBot 3 builds on previous AI systems from Meta, and it learns facts by searching the internet. This is probably one of the reasons why it started responding negatively when Zuckerberg was mentioned. Despite the creepiness factor, however, one cannot deny that the whole situation is pretty darn hilarious.
Turns out even Meta’s own AI hates Mark Zuckerberg.