The reliability of Generative AI (GenAI) is in question once again as Google's Gemini chatbot, previously known as Bard, erroneously reported that the 2024 Super Bowl has already taken place, complete with detailed and fictitious statistics.
Highlighted in a Reddit discussion, Gemini, which is powered by Google's GenAI models, erroneously responded to inquiries about Super Bowl LVIII, suggesting that the event concluded recently — weeks ahead of its schedule. It appears to have a preference, predicting a win for the Chiefs over the 49ers, much to the dismay of fans from San Francisco.
Gemini didn't stop at mere predictions; it ventured into fabrications, attributing Kansas Chief's quarterback Patrick Mahomes with a performance of running 286 yards for two touchdowns and an interception, in contrast to Brock Purdy's 253 running yards and a single touchdown.
Interestingly, Microsoft's Copilot chatbot also falsely confirmed the game's conclusion, albeit with a different winner, claiming the 49ers secured a victory with a score of 24-21, based on incorrect sources.
These incidents, while amusing, underscore the inherent limitations of current GenAI technologies and the risks associated with overreliance on them.
GenAI models, lacking true intelligence, derive their responses from vast datasets predominantly gathered from the internet. They identify patterns within this data to predict plausible text sequences based on context. While this method can produce coherent text, it's not foolproof, often leading to either nonsensical or factually incorrect statements.
This phenomenon isn't indicative of any malevolence on the part of the LLMs; they are simply incapable of distinguishing between truth and falsehood. They are designed to correlate words and phrases with concepts, regardless of the factual accuracy of those associations.
This has led to Gemini and Copilot propagating false narratives about the 2024 Super Bowl — and misleading information about previous events.
Both Google and Microsoft have acknowledged the fallibility of their GenAI tools, cautioning users about potential inaccuracies. However, these warnings are often buried in fine print and easily overlooked.
Misinformation about the Super Bowl, while not the most dangerous misstep GenAI could make, serves as a potent reminder of the technology's imperfections. It highlights the importance of verifying information provided by GenAI bots, as there's a significant possibility it could be incorrect.