Google is abandoning the use of AI-generated answers in search results after the feature made notorious mistakes, including telling users to put glue in their pizza sauce.
Two weeks ago, Google launched AI Overviews, which places AI-generated summaries of search results at the top of the page for US users. Over the past few days, users, including an SEO expert, noticed fewer AI presentations and suspected the tech giant was taking them down a notch following criticism. It is not possible to deactivate the AI function when using the search engine.
Google’s head of search, Liz Reid, confirmed in a blog post on Thursday that the company was fixing some of these issues.
Changes come after recent examples of AI previews going haywire – and fake images of the feature – have flooded the internet. These included responses claiming that Barack Obama was a Muslim president, that Africa did not have countries beginning with the letter K, and that people should eat “at least one small stone a day.”
Google’s new safeguards include detecting “nonsense queries” that shouldn’t display AI results, limiting satirical or humorous content, and introducing restrictions for prompts where AI results would not be useful because there is not enough data on this subject.
Google’s own ads show that bad summaries aren’t limited to a few viral queries. In a demo video posted two weeks ago, the Presentation feature incorrectly advised the actor how to repair his camera.
Reid’s blog post also states that Google offers limited content from forums or social networks, which may contain misleading advice.
“Forums are often a great source of authentic, first-hand information, but in some cases they can result in less than helpful advice, like using glue to make cheese stick to pizza,” Reid wrote in his message.
Reid wrote that the company already has systems in place to not display AI-generated news or health-related results. She said the harmful findings which encouraged women to smoke while pregnant or leave their dogs in cars were “fake screenshots”.
The list of changes is the latest example of the Big Tech giant launching an AI product and coming back with restrictions after things get complicated.
Earlier this year, Google AI’s image generation feature was criticized for refuse to produce photos white people. He was criticized for being too “woke” and creating photos with historical inaccuracies like Asian Nazis and Black Founding Fathers. In a blog post a few weeks later, Google executives apologized and suspended the feature.
businessinsider
