Google has an egg in its face as it is in a hurry to edit ads for Gemini AI tools before it airs during the Super Bowl. The ad has found that it contains false information about the consumption of cheese that could be “haptic” by the Bumble tech giant's Awakening AI.
Guardian Report Tech Giant's Super Bowl commercial shows how companies can leverage “AI for all businesses.” It portrays Gemini in supporting Wisconsin cheese makers in creating product descriptions. However, AD contained false statistics claiming Gouda accounted for “50% to 60% of global cheese consumption.”
The error was revealed by blogger Nate Hake, who posted to X that the statistics were “clearly false” with “AI hallucinations.” Hake noted that more reliable data suggests that Gouda is likely to be less popular worldwide than Cheddar and Mozzarella. He added: “I found the above example of the AI slop in 20 minutes. In my first Super Bowl ad, I tried to do fact checks.”
In response, Google executive Jerry Dischler said that inaccuracy is not a “hastique” and that it is not a phenomenon in which AI systems invent false information, but rather an incorrect one that exists on websites where Gemini scrapes data. He claimed it reflects information. Dischler said: “Gemini is grounded to the web, allowing users to see results and references at any time. In this case, multiple sites on the web have 50-60% stats.”
After that, Google issued a statement explaining that after consulting with the featured cheesemonger, it re-mapped the ad to remove the error after asking how it would prefer to handle the situation. “According to his suggestion to rewrite the product descriptions to Gemini without statistics, we have updated the user interface to reflect what the business is doing,” the statement read.
This is not the first time Google's AI tools have faced criticism for suppressing errors or providing useless advice. In May of the previous year, the company's AI overview search feature fired to suggest using “non-toxic adhesives” when users searched for ways to make cheese sticks better against pizza . Furthermore, the AI-generated responses claimed that geologists would recommend that humans eat one rock per day.
Google's Gemini tool came across an issue that was “paused” last year after the company admitted it was “undetectedly ruined.” The tool produced images depicting historical figures, including the Pope, the founding father of the United States, and German World War II soldiers as racial minorities. Additionally, the Gemini Chatbot provided an inconsistent response when comparing the harm caused by libertarians and Stalin. These incidents prompted negative commentary from conservatives, including Elon Musk.
Please read more Guardian here.
Lucas Nolan is a reporter for Breitbart News, which covers the issues of freedom of speech and online censorship.
