Google’s AI Is Making Up Wise Sayings—and Fooling Us All

Google’s artificial intelligence (AI) has ignited a heated debate by interpreting non-existent proverbs, fueling concerns about AI hallucinations.

On Tuesday, Ars Technica reported that users have discovered Google’s AI interpreting made-up phrases as if they were genuine proverbs. For example, when someone enters the phrase “You can’t lick a badger twice,” Google’s AI explains it as meaning “You can’t fool someone twice who’s already been deceived.” However, this saying doesn’t actually exist.

Following the viral badger post, countless users took to social media to share Google AI’s responses to their invented proverbs. While some expressed dismay at Google’s erroneous interpretations, others noted instances where the AI derived more profound meanings than the original user’s phrase intended.

The danger lies in AI presenting false information as fact, which can seriously undermine information reliability. When AI confidently delivers misinformation, users are more likely to accept it as truth without question.

As Google’s AI capabilities continue to advance, these AI hallucinations are becoming increasingly common, raising serious questions about the future credibility of search engines.