Google is facing backlash over its recent Super Bowl ad for Gemini AI, with critics accusing the company of exaggerating the model’s real capabilities.
Viewers and AI experts quickly pointed out inconsistencies between what was showcased in the commercial and what Gemini can actually do today.
The ad was designed to highlight Gemini’s ability to understand and reason across different types of media, featuring impressive examples of the AI interacting with images, videos, and real-time inputs.
But after a closer look, many believe the video wasn’t an accurate representation of Gemini’s current functionality.
In the initial ad, Gemini had hallucinated heavily, providing copy that said that Gouda cheese accounts for “50 to 60 percent of the world’s cheese consumption”. A false statement.
The video later magically appeared with the statement corrected.
Critics argue that the ad made it seem like Gemini AI could effortlessly analyze and respond to visual input in real time. In reality, the actual model reportedly struggles with some of these tasks.
Further scrutiny suggests that the responses shown in the ad weren’t generated on the spot but were either scripted or selectively edited to appear more seamless than they are.
Google hadn’t responded by the time of going to press.
