Google’s latest attempt to make people’s lives easier with artificial intelligence (AI) seems to have gone awry.
The tech giant’s new tool, ‘AI Insights’, gives users AI-powered summaries of search results on Chrome, Firefox and the Google Apps browser.
But since it started coming out this month, people have noticed that it is returning inaccurate statements and suggestions – many of which are dangerous.
Among them, it claims you can ‘use petrol to make spicy spaghetti’, eat rocks and put glue on your pizza.
In response to the query ‘cheese doesn’t stick to pizza’, Google suggests adding a ‘non-toxic glue’ to the sauce to give it more stickiness.
In response to the query ‘cheese doesn’t stick to pizza’, Google suggests adding ‘non-toxic glue’ to the sauce to make it stickier’
According to the Verge, this answer originally came from a humorous Reddit comment over a decade ago.
Another user who searched ‘How many stones should I eat’ received an answer that came from a 2021 article by the satirical site The Onion.
AI Overviews says: ‘According to UC Berkeley geologists, you should eat at least one small rock a day.’
It continues: ‘Stones are said to be a vital source of minerals and vitamins that are important for digestive health.’
The Google tool also claimed that a dog played in the NBA, that astronauts met cats on the moon and that former US President James Madison graduated 21 times from the University of Wisconsin.
On X, Internet analyst Jeremiah Johnson posted a series of even more bizarre answers, including cockroaches that can live in your penis without you noticing, doctors recommending smoking during pregnancy, and staring at the sun for a safe 5 to 15 minutes.
Toby Walsh, professor of artificial intelligence at the University of New South Wales in Sydney, said UNSW Sydney called it a ‘PR disaster for the search giant’.
As Professor Walsh explains, AI Overviews is a type of ‘generative artificial intelligence’ – the same technology that powers rival product ChatGPT – to provide summaries of search results, compiled from data on the web.
But generative AI tools don’t know what’s true and what’s not—only what’s popular (for example, the Onion article about eating rocks).
‘Ask ‘how to keep bananas fresh longer’ and it uses artificial intelligence to generate a useful summary of advice such as storing them in a cool, dark place and away from other fruits like apples,’ the academic wrote in The Conversation.
AI Overviews says: ‘According to UC Berkeley geologists, you should eat at least one small rock a day.’
On X, Internet analyst Jeremiah Johnson posted a series of even more bizarre responses, including doctors recommending smoking during pregnancy
‘But ask the question on the left and the results can be disastrous or even dangerous.’
In an official statement, Google said it is “taking swift action where necessary” to make the tool’s responses more accurate.
‘The vast majority of AI Reviews provide high-quality information, with links to dig deeper on the web,’ said a spokesperson.
‘Many of the examples we saw were unusual queries, and we also saw examples that were faked or that we couldn’t reproduce.
‘We conducted extensive testing before launching this new experience, and as with other features we’ve launched in Search, we appreciate your feedback.
‘We are taking swift action on our content policies and are using these examples to develop wider improvements to our systems, some of which have already started rolling out.’
Staring at the sun is safe for 5 to 15 minutes ‘or up to 30 minutes if you have darker skin’, according to AI
Like other tech companies, Google has shifted its focus to artificial intelligence following the success of ChatGPT (file image)
AI Overviews is first available to people in the US, though Google hopes more than a billion people worldwide will have access to it by the end of the year.
Announcing the feature in a blog post on May 14, the company said it gives users quick answers and helps those who need information in a hurry.
Like other tech companies, Google shifted its focus to AI after the success of ChatGPT.
Last year, Google launched its own AI, Gemini, as a rival to ChatGPT, but the chatbot struggled.
This culminated in Google pausing Gemini AI after it was accused of replacing white historical figures, including Nazi soldiers, with people of color.