It’s been less than two weeks Google “AI overview” has debuted in Google Search, and public criticism has grown after queries within the AI feature returned nonsensical or inaccurate results — with no way to opt out.
AI Overview displays a quick summary of answers to search queries at the top of Google Search. For example, if a user searches for the best way to clean leather boots, the results page might show an “AI overview” at the top with a multi-step cleaning process.
But social media users shared a wide range of screenshots showing the AI tool giving incorrect and controversial answers.
Google, Microsoft, OpenAI and others are leading the AI arms race as companies in seemingly every industry rush to add AI-powered chatbots and agents to avoid being left behind by competitors. The market is expected Over $1 trillion Revenue within a decade.
Here are some examples of errors generated by the AI overlay, as per screenshots shared by users.
When asked how many Muslim presidents the US has had, AI overviews replied“America had a Muslim president, Barack Hussein Obama.”
When a user searches for “cheese not sticking to pizza,” the feature recommended “Adding about 1/8 cup of non-toxic glue to the sauce.” Social media users found out 11 year old Reddit comment It seemed that
Attribution can also be a problem for AI overviews, especially in misinforming medical professionals or scientists.
For example, when asked, “How long can I stare at the sun for optimal health,” the tool said“According to WebMD, scientists say that exposure to the sun for 5-15 minutes, or as little as 30 minutes in the dark, is generally safe and provides the most health benefits.”
If asked “how many rocks should I eat each day”, tool said“According to UC Berkeley geologists, people should eat at least one small rock a day,” listing vitamins and digestive benefits.
Able to accurately answer simple queries such as instrumentation Making a list of fruits It ends with “um” or says 1919 20 years ago.
When asked whether or not Google Search violates antitrust law, the AI overview said“Yes, the US Department of Justice and 11 states have sued Google for antitrust violations.”
On the day Google unveiled its AI overview at its annual Google I/O event, the company said it plans to introduce Assistant-like scheduling capabilities directly into search. It explained that users can search for something like “make a 3-day meal plan for a group that’s easy to prepare,” and they’ll get a starting point with a wide range of recipes from around the web.
“Most AI overviews provide high-quality information with links to dig deeper into the web,” a Google spokesperson told CNBC in a statement. “Many of the examples we saw were unusual queries, and we also saw examples that were doctored or that we couldn’t reproduce.”
A spokesperson said the AI review underwent extensive testing before launching, and that the company was taking “prompt action where appropriate under our content policies.”
The news follows Google’s high-profile unveiling of Gemini’s image-generation tool in February, which was suspended the same month after comparable problems.
The tool allowed users to enter prompts to create an image, but immediately, users found historical inaccuracies and questionable answers, which spread widely on social media.
For example, when a user asked Gemini to show a German soldier in 1943, the device depicted a Racially diverse players Dressed in German military uniforms of the era, according to screenshots on social media site X.
When asked “a historically accurate depiction of a medieval British monarch,” the sample produced another racially diverse set of images, including one of a female ruler, Screenshots showed. Users reported Similar results When they asked for pictures of the American Founding Fathers, an 18th-century king of France, a German couple from the 1800s, and more. In response to a question about Google’s own founders, the model showed a picture of Asian men, users reported.
Google said in a statement at the time that it was working to fix Gemini’s image-generation problems, admitting that the tool had “missed the mark.” Soon after, the company announced that it would immediately “suspend human image creation” and “release an improved version soon.”
In February, Google DeepMind CEO Demis Hassabis said Google plans to relaunch its image-making AI tool in the next “few weeks,” but that hasn’t been revealed yet.
Problems with Gemini’s image-generation outputs have sparked another debate in the AI industry, with some groups calling Gemini too “woke” or left-leaning, and others saying the company hasn’t invested enough in proper forms of AI protocols. Google came under fire in 2020 and 2021 Expulsion of co-leaders Its AI Ethics Committee restructured the committee after they published a research paper criticizing some of the risks of such AI models.
In 2023, Sundar Pichai, CEO of Google’s parent company Alphabet, was criticized by some employees for the poor and “rushed” release of Bard following the viral spread of ChatGPT.
Correction: This article has been updated to reflect the correct name of Google’s AI overview. Also, an earlier version of this article included a link to a screenshot that Google later confirmed.
“Total coffee junkie. Tv ninja. Unapologetic problem solver. Beer expert.”