Testing suggests Google's AI Overviews tells millions of lies per hour
SMRTR summary
Google's AI Overviews feature, which appears atop search results, correctly answers questions 90 percent of the time according to new testing, but this means the system still generates tens of millions of wrong answers daily across Google's massive search volume. The New York Times analysis found AI Overviews improved from 85 to 91 percent accuracy after recent updates, yet continues making basic errors like citing contradictory sources.
SMRTR provides this summary for quick context. The original article belongs to Ars Technica.
Read the original article