Google Gemini chatbot, previously known as Bard, has been found to be generating false statistics for the just-concluded 2024 Super Bowl.
What Happened: Alphabet Inc.’s GOOG GOOGL Google Gemini chatbot has been answering inquiries about Super Bowl LVIII as if the game had already occurred. The bot even produced made-up statistics to back its assertions, according to Reddit posts.
What makes it worse is that Gemini made up statistics even after Super Bowl LVIII ended and the Kansas City Chiefs clinched the title in an overtime thriller with a 25-22 scoreline.
However, Gemini claims the San Francisco 49ers won the game.
As per a Reddit thread, Gemini predicted the Chiefs as the winners over the 49ers and provided a comprehensive player stats analysis.
This incident underscores the limitations of current GenAI models. Despite their ability to generate grammatically correct responses, these models can also produce nonsensical or incorrect information.
The Super Bowl 2024 fabrications by Gemini serve as a cautionary tale about the dangers of over-reliance on GenAI bots.
We also tried using the same prompt with Microsoft Corp.’s Copilot and OpenAI’s GPT-4. While Copilot answered the question accurately, GPT-4 did not answer it by saying that it’s likely that the game hasn’t concluded yet.
Why It Matters: This isn’t the first time Google’s Gemini has been under scrutiny.
In December, there were allegations that Google may have staged Gemini’s video demonstration by pre-feeding the questions to its AI model.
The recent Super Bowl 2024 fabrication by Gemini further highlights the potential risks and limitations of GenAI models, raising questions about the reliability and trustworthiness of these AI applications.
Check out more of Benzinga’s Consumer Tech coverage by following this link.
Disclaimer: This content was partially produced with the help of Benzinga Neuro and was reviewed and published by Benzinga editors.
Photo courtesy: Google