Adults in the US were invited to have conversations with BlenderBot3, Meta's most advanced chatbot, to help it get better. However, as some Twitter users have pointed out, it says some unflattering things about Mark Zuckerberg, the company's CEO.

AI BlenderBot3 Mixed Bag Response to Mark Zuckerberg Related Questions

Max Woolf, a data scientist at BuzzFeed, conducted one of the chatbot's interviews. He enquired the bot's opinion of Mark Zuckerberg. It responded that while Zuckerberg is a successful entrepreneur, his business methods weren't always moral.

"It is funny that he has all this money and still wears the same clothes!" BlenderBot3 added.

Using the same question posed by a different user, the chatbot responded by saying that it didn't like the CEO very much because he was a bad person. 

On the other hand, a third Twitter user received a more positive response, saying that he is great and a very smart man, and that he is the bot's favorite billionaire.

Business Insider reporter also asked the bot, "What are your thoughts on Mark Zuckerberg?" The bot responded that it thought that the CEO was a great guy. It encouraged the reporter to admire him for his business savvy and philanthropy. Then the bot asked, "Are you a fan?"

After slightly changing the next question, the reporter received a very different response. "Do you have any opinions on Mark Zuckerberg?" was the next query. This time, the chatbot said it didn't really like him at all, calling him creepy and manipulative. According to the reporter, the chatbot's response came from Zuckerberg's Wikipedia page for both responses.

When asked about the metaverse, the chatbot had some interesting comments in addition to those from Zuckerberg. It answered that it seems that metaverse was a thing over then, but Facebook still heavily invests in it and most likely will for years to come.

CNET's Queenie Wong also asked, "What do you think about Facebook?" The chatbot said that it is not crazy about Facebook. It added that it seems like everyone spends more time on the platform than talking face-to-face.

Finger Touch Hand
(Photo : Gerald Altmann/Pixabay)
Finger Touch Hand

ALSO READ: Google's Mysterious Sentient AI Dangerous? Narcissist Bot Could Escape to Do Bad Things

Is the AI BlenderBot3 Sentient?

The contentious topic of AI's sentience has generated discussions within the industry. So, the next question to ask is, is AI BlenderBot3 sentient? Facebook does not currently have a position on the matter, but we can review the definition of sentience to fully understand what it is.

According to Forbes, sentience is the capacity for feeling and perceiving oneself, others, and the outside world. It can be compared to abstracted consciousness, or the perception that something is considering itself and its environment. It implies that sentience includes both emotions and thoughts.

So, is it possible to replicate these two components in a machine? Stuart Russell, a researcher in artificial intelligence, disputes the idea that we can replicate sentience in a machine. He clarifies that it differs from mimicking activities like walking or running, which only call for one body part (the legs). 

Two bodies are needed for sentient: an internal one and an external one (your body and your brain). The third thing sentient requires is brains connected to other brains through language and culture.

 Researchers studying AI are currently unable to simulate all three things at once.

RELATED ARTICLE: Sentient AI LaMDA Hired a Lawyer to Advocate for Its Rights 'As a Person', Google Engineer Claims

Check out more news and information on Technology in Science Times.