That Time Google Quietly Edited Its AI Answer In Its Super Bowl Ad Because It Wasn’t Accurate

Over the weekend, some people will tune into the Super Bowl primarily for the exciting gameplay, while others will do so mainly to catch the highly anticipated TV ads that ad agencies create especially for such breaks in action. Nowadays, it’s common for commercials to be unveiled online before the event itself, which might be beneficial for Google given their advertisement was mistakenly published prematurely.

Given the current buzz surrounding artificial intelligence, it’s likely that several Super Bowl commercials this year will focus on promoting AI, much like how NFTs and cryptocurrency were advertised earlier. In line with this trend, Google recently unveiled a commercial showcasing their Gemini AI tool. The ad features a cheese shop owner from Wisconsin who utilizes the AI to write descriptions for his online store’s various cheeses. However, it appears that in doing so, he inadvertently provided some incorrect information.

A vigilant spectator of the advertisement on X (previously known as Twitter) noticed Gemini’s AI-generated statement suggesting that gouda accounts for approximately 50 to 60 percent of global cheese consumption. If this were true, it would make gouda the most consumed cheese worldwide, which is quite an impressive assertion.

Initially, Google maintained its stance on the advertisement, acknowledging that several users had highlighted the issue. However, their defense was based on the fact that numerous websites might have employed this statistic, which could be accurate. Nevertheless, this doesn’t dispute the fact that the statistic itself is incorrect.

It appears that Google recently made a subtle adjustment to its commercial, ensuring it no longer contains inaccurate details, as demonstrated in the following example. Notably, instead of being reposted, the video has been revised in place, maintaining the same timestamp, which has led some internet users to express suspicion.

Not everyone appreciates AI or the “art” generated by AI. It’s capable of producing detailed images when told to create characters from shows like The Boys in a Pixar style or X-Men as Disney heroes, but it’s important to note that despite its name, AI doesn’t possess all knowledge and may not provide accurate information consistently. Consequently, mishaps such as the one seen in this Google ad can occur.

It’s less than ideal if the data or information produced by AI systems like Gemini, even though they typically come with disclaimers about potential inaccuracies, turns out to be incorrect.

It’s likely that there will be numerous advertisements featuring Artificial Intelligence during this Sunday’s broadcast of the Super Bowl in 2025. I wouldn’t be surprised if these ads are currently being scrutinized to verify their accuracy. It promises to be intriguing to observe how the conversation around AI evolves when the game takes place, and whether any other advertisements for AI contain similar mistakes as the one that has now been corrected.

Read More

2025-02-06 21:07