It’s such a bizarre situation that it almost gives you a headache. Google’s conversational artificial intelligence, Bard, has announced his own death, even as he just made his public release. And to get there, it took a series of pranks, misinformation and social media broadcasts.
As the American media The Verge points out, a user first had fun in a forum to make believe that Bard was already closed. To do this, with the help of other netizens, they asked ChatGPT to write false information about this supposed shutdown of Google’s AI.
ChatGPT, provider of fake news
No problem for ChatGPT, which immediately provided a fake text, as if it had been written by a news site. “Google recently announced that it will end Google Bard, its new AI-powered writing tool, less than six months after its launch,” ChatGPT invents.
It will not take longer for Bard to recover this information in his database and therefore think… that he is dead. A user has thus published on Twitter the exchange of him with Google’s AI where he asks “how long will Google take to stop Bard?”, Without waiting for him to answer that he “is already stopped”, collecting false information from ChatGPT.
Bing joins the fight
But the machine doesn’t stop there: to tell the story, the WindowsCentral outlet produced an article, taking up the tweet from the aforementioned Internet user. And that’s when Bing steps in: in turn, the AI built into Microsoft’s search engine picks up the false information.
When asking Bing if Bard was arrested by Google, the answer is clear: yes, on March 21, 2023, after six months of life. full circle. However, the bug has now been fixed and Bing now indicates that Bard is alive and well for now.
This episode shows above all the ease with which chatbots retrieve false information scattered here and there on the Internet. Not to mention his inability to discern which sources of information are trustworthy and which are not.
Source: BFM TV
