- Google’s new AI Overview product has faced criticism for generating unusual and nonsensical responses, such as suggesting users put glue on pizza or eat rocks.
- Users and social media platforms have expressed concerns over the strange responses, leading to the rapid removal of problematic AI overviews.
- Despite a year of testing and serving over a billion queries, concerns have been raised about the readiness of the AI system and the premature optimisation implemented by Google.
Google’s new AI Overview product has stirred controversy and garnered attention on social media platforms for generating peculiar and sometimes nonsensical responses. Users have reported instances where the AI suggests unusual actions like putting glue on pizza or eating rocks, leading to a flurry of memes and critiques.
Rollout challenges and response management
The rollout of the AI Overview product has been marked by challenges, with Google actively working to manually disable AI Overviews for specific searches to address the situation. This response comes as a result of various memes circulating on social networks, causing many of the unusual responses to be swiftly removed after being posted.
Also read: Google is using AI to answer search queries
Testing and optimisation journey
Despite undergoing testing for a year and serving over a billion queries, concerns have arisen regarding the readiness of the AI system. CEO Sundar Pichai highlighted an 80% reduction in the cost of delivering AI answers, achieved through hardware and technical breakthroughs. However, this optimisation may have been implemented prematurely, leading to the current issues faced by the AI Overview product.
Also read: Google CEO Sundar Pichai on AI search and the web’s future
Quality assurance and improvements
While Google maintains that the AI Overview product primarily provides high-quality information, instances of strange or inaccurate responses have prompted discussions about the overall quality of the output. The company is taking proactive measures to remove problematic AI overviews, leveraging feedback to enhance the system and ensure better user experiences.
