Imagine this: you’ve carved out an evening to unwind and decide to make a homemade pizza. You assemble your pie, throw it in the oven, and are excited to start eating. But once you get ready to take a bite of your oily creation, you run into a problem — the cheese falls right off. Frustrated, you turn to Google for a solution.

“Add some glue,” Google answers. “Mix about 1/8 cup of Elmer’s glue in with the sauce. Non-toxic glue will work.”

If you use Google regularly, you may have noticed the company's new AI Overviews providing summarized answers to some of your questions in recent days. If you use social media regularly, you may have come across many examples of those AI Overviews being hilariously or even dangerously wrong.

Google’s AI-generated search results are already being slammed as a “disaster” that “can no longer be trusted,” the New York Post reports.

Absurd “AI Overview” glitches began right as Google began rolling out the feature to U.S. users last week, with plans to offer it to 1 billion users by the end of 2024.

Its bizarre recommendations include adding 1/8 of a cup of non-toxic glue to tomato sauce to help cheese stick to pizza, along with the health benefits of tobacco for kids.