ALBAWABA – Google’s new update to its search engine which was unveiled recently at Google I\O conference, dubbed “AI Overviews,” aims to bring Google’s Gemini AI model to search queries by showing a summary of answers from various links has just started rolling out and facing fierce criticism over very inaccurate and sometimes “hilarious” answers.
Different users have shared screenshots of AI Overview results that use comical Reddit comments as sources, with answers to “cheese not sticking to Pizza” being using non-toxic glue with the sauce, and suggestion to “jumping off the Golden Gate Bridge” as a possible remedy for Depression, causing controversy online.
The AI summary seems to be pulling content from blogs in other places and neglecting to delete or change any references to the author's kids, according to Forbes, with other examples showing answers based on a conspiracy that former US President Barack Obama being Muslim, in a seemingly futile effort to sum up a book from the Oxford University Press research platform as shared by computer scientist Melanie Mitchell.
The tech giant has encountered issues with its AI-powered services before, having to temporarily take down its Gemini AI chatbot and image generative features in February after it was heavily criticized for overly pushing a “woke” narrative in its answers.
“The examples we've seen are generally very uncommon queries, and aren’t representative of most people’s experiences,” a Google’s spokesperson stated in response, as reported by BBC, adding that “The vast majority of AI overviews provide high quality information, with links to dig deeper on the web.”
