Google's 'AI Overview' Feature Struggles with Inaccurate Responses, Raises Concerns
Summary:
Google's artificial intelligence feature, “AI Overview,” has been providing inaccurate and potentially harmful responses to user queries. Google has yet to find a concrete fix to the problem, instead disabling certain queries that generate erroneous results. Missteps include suggestions to use glue to keep cheese on a pizza, and inaccurate claims about dogs owning hotels. Google seems to blame the issue on users for generating 'unusual queries', while queries from Cointelegraph for further clarification went unanswered. Meanwhile, Elon Musk, founder of AI firm xAI, predicts that machines will surpass human capabilities by 2025.
Google's recently-introduced artificial intelligence (AI) feature, "AI Overview," has been serving up incorrect and potentially dangerous responses to user inquiries, with the tech giant seeming to have no immediate solution to the issue. At the moment, Google has had to restrict certain search prompts on its "AI Overview" after the system started giving misleading and possibly hazardous suggestions.
The issue came to light when news and social media platforms reported users asking the search algorithm how to ensure cheese sticks to pizza, to which the AI responded by suggesting the use of glue. Other blunders included the system incorrectly claiming that dogs owned hotels and referenced a non-existent canine statue as proof. These mistakes may seem light-hearted, but the real issue lies in the fact that the AI continues to produce both accurate and inaccurate results with the same level of assurance.
Up to this point, Google's response has been to disable the specific queries that lead to inaccurate results, as they are identified. This approach, however, paints a picture of Google playing a game of catch-up with the shortcomings of its AI feature. On top of that, Google seems to be attributing the problem to the people asking the questions.
Meghann Farnsworth, a Google spokesperson, told The Verge via email, “A lot of the sample queries that we've come across are quite unusual. We've also come across cases that were either manipulated or could not be replicated.” How users can avoid asking "unusual questions" is not clear at this time.
Inquiries from Cointelegraph for further explanation from Google did not immediately get a response. It seems there's still a need for some adjustments on Google's AI system before it runs as smoothly as intended. In contrast, Elon Musk, founder of competing AI company, xAI, recently projected that artificial intelligence will surpass human abilities by the close of 2025. At the VivaTech 2024 conference in Paris, Musk predicted that xAI could match, if not surpass, the likes of OpenAI and Google's DeepMind by the end of 2024.
Published At
5/25/2024 7:34:22 PM
Disclaimer: Algoine does not endorse any content or product on this page. Readers should conduct their own research before taking any actions related to the asset, company, or any information in this article and assume full responsibility for their decisions. This article should not be considered as investment advice. Our news is prepared with AI support.
Do you suspect this content may be misleading, incomplete, or inappropriate in any way, requiring modification or removal?
We appreciate your report.