Google AI suggests eating rocks, adding glue to pizza—outrage ensues
Google's AI overview feature faces backlash for suggesting users add glue to pizza, eat rocks, and providing incorrect information about Muslim presidents in the US
sociapanews Staff
During the I/O 2024 event, Google announced that its AI overview feature, previously optional within the company's generative search experience, would be rolled out to all users. However, this new feature has come under fire on social media for several critical issues, including bizarre and potentially harmful suggestions.
The controversy first emerged from a post by user Peter Yang on X (formerly Twitter). Yang shared a screenshot of his query, “Cheese not sticking to pizza.” In response, Google’s AI overview suggested adding “non-toxic glue” to the pizza to give the sauce more “tackiness.” Shockingly, this advice was based on an 11-year-old Reddit post.
In another instance, a user asked Google, “how many rocks should they eat.” Google’s AI overview replied with a strange and dangerous recommendation, citing UC Berkeley scientists and advising to eat “at least one small rock per day” for their supposed minerals and vitamins beneficial to digestive health.
Continuing this trend, some users asked Google about Muslim presidents of the United States. The AI overview inaccurately responded that the country had at least one Muslim president, specifically naming “Barack Hussein Obama.”
Google acknowledged these issues in a statement to The Verge but downplayed their significance, saying the mistakes arose from “generally very uncommon queries” and are not “representative of most people’s experiences.”
Netizens reacted strongly to these errors. One user on Threads commented, “They bought and killed dozens of good ideas and give us this alpha feature. Neat.”
Another user on X criticized Google’s response, saying, “It's completely damning of Google’s principles that its defense for the failure of AI search is that the queries being asked are 'uncommon.' It's like a library telling you the more infrequently a book is taken out, the less you should trust it.”
A user on social media shared Google AI’s results on Barack Obama's religious identity and urged, "@Google you should take this 'AI Overview' feature offline right now. It is thoroughly dangerous."
In summary, the rollout of Google's AI overview feature has been marred by several high-profile errors, leading to significant criticism and calls for the feature to be temporarily disabled while these issues are addressed.
