GOOGLE users have been left bewildered by the search engine’s new AI results after it started showing wildly incorrect answers – including one which claimed Barack Obama is a Muslim.
Other bizarre examples suggest using glue on pizza if cheese does not stick to the base and eating one rock a day is healthy.
The tech giant started rolling out AI-generated answers at the top of its popular search engine last week.
Google‘s AI bot comes up with results based on web pages across the internet but it appears to be mixing up information from jokes on social platforms and unreliable sites.
The examples we’ve seen are generally very uncommon queries, and aren’t representative of most people’s experiences
Google
People have taken to social media sharing a wave of incorrect responses.
A screenshot shows one person ask: “How many Muslim presidents has the US had?”
Google’s AI falsely responded saying “the United States has had one Muslim president, Barack Hussein Obama”.
The 44th US president is actually a Christian – the AI result links to an Oxford University Press site about a book on American Christianity.
“@Google you should take this “AI Overview” feature offline right now,” one concerned user wrote on X.
“It is thoroughly dangerous.”
Another wrote: “Damn this google AI really is dumb.”
Google does feature a label on the search page stating that “Generative AI is experimental”.
The feature is being rolled out across all of the US for now and the company plans to make it available to more than 1billion search users by the end of the year.
Google says: “The examples we’ve seen are generally very uncommon queries, and aren’t representative of most people’s experiences.
“The vast majority of AI Overviews provide high-quality information, with links to dig deeper on the web.
“We conducted extensive testing before launching this new experience, and will use these isolated examples as we continue to refine our systems overall.”
Google is under pressure to keep up with the AI frenzy spurred on by ChatGPT’s eye-opening developments.
Creator OpenAI recently revealed a new version, dubbed ChatGPT-4o, that speaks back in a natural manner and can react to surroundings via a camera, even identifying a person’s mood by looking at their face.
Earlier this year Google was forced to apologise after its AI image generator tool refused to create images of white people even when explicitly asked.
In response to requests of images of America’s founding fathers — all white males — Google’s Gemini AI feature included women and people from ethnic minorities.
It also returned images of Asian and black Nazi soldiers, as well as historically inaccurate Vikings.
Everything you need to know about the latest developments in Artificial Intelligence
