Sundar Pichai-led Google defends its AI Overview results that asked people to use glue in pizza

Google’s AI Overview responses went viral earlier this week when it asked users to use glue to make cheese sticks on a pizza or eat rocks. Google’s new AI feature was meant to give better search results but when users asked it weird questions, it became a source of meme. The company has now blamed “data voids” or “information gaps” for these inaccurate results.

In a blog post, Google’s head of search, Liz Reid stated that AI Overview does “hallucinate” or make things up. When it gives wrong results, it is usually because of “misinterpreting queries, misinterpreting a nuance of language on the web, or not having a lot of great information available”.

She stated, “We saw AI Overviews that featured sarcastic or troll-y content from discussion forums. Forums are often a great source of authentic, first-hand information, but in some cases can lead to less-than-helpful advice, like using glue to get cheese to stick to pizza.”

Google has further clarified that when these inaccurate AI Overview responses were being shared on the internet, several fake screenshots were also making rounds. AI Overview results include leaving dogs in cars, smoking while pregnant, and depression never appearing.

Google has not dealt with a lot of nonsensical questions like “How many rocks should I eat?”. Reid admitted that AI Overview needs work when it comes to interpreting nonsensical queries and satirical content. The tech giant has also updated its systems “to limit the use of user-generated content in responses that could offer misleading advice”. Google has also added “triggering restrictions for queries where AI Overviews were not proving to be as helpful”.

The company has admitted, “In a small number of cases, we have seen AI Overviews misinterpret language on webpages and present inaccurate information. We worked quickly to address these issues, either through improvements to our algorithms or through established processes to remove responses that don't comply with our policies.”

Google AI Overview will not show up for “hard news topics” where “freshness and factuality are important”. Additionally, in terms of health, Google has launched “additional triggering refinements to enhance our quality protections”.

Also Read: 

OpenAI CEO Sam Altman eyes restructuring and turning into a regular for-profit company: Report

Drone startup TechEagle secures funding from Paytm’s Vijay Shekhar Sharma, Inflection Point Ventures and more