Technology

Google rolls back AI search answers after it told users to eat glue

SAN FRANCISCO — Google said it is reducing the use of AI-generated answers in some search results, after the technology made high-profile mistakes, including telling users to put glue on their pizza and by affirming that Barack Obama was Muslim.

Google began putting artificial intelligence answers at the top of users’ search results in the United States two weeks ago, but users and search engine experts on Thursday began noticing significantly fewer queries triggered an AI response compared to previous days.

Google’s head of search, Liz Reid, confirmed in a blog post Thursday afternoon that the company is reducing some of the AI ​​responses, which she calls “AI previews.” The company reduced the use of social media posts as a source for AI responses, suspended some responses on health-related topics, and “added trigger restrictions for queries for which AI insights AI has not proven to be as useful,” Reid said in the post. . The company has made more than a dozen technical changes aimed at improving the system, Reid said.

The change is the latest example of Google launching an AI product with great fanfare, then canceling it after a problem. In February, the company blocked users from creating images of people with its AI image tool after conservative commentators accused it of anti-white bias.

The tech industry is in the grip of an AI revolution, with start-ups and Big Tech giants trying to find new ways to integrate the technology into their products and profit from it. Many tools have been released before they are ready for prime time, as companies scramble to be first to market and present themselves as cutting edge.

TO CATCH UP

Summarized stories to stay informed quickly

Google, whose employees invented much of the technology behind revolutionary AI tools like ChatGPT, is trying to prove to investors, consumers and its own employees that it remains the most important player in the sector. At its I/O conference this month, the company made more than 100 different AI-related announcements.

The biggest was confirmation that it would begin rolling out AI-generated answers in search results for most of its users. Google has been testing AI responses for a year with a select group of users, but by adding them to more search results, most people would now start interacting directly with generative AI on a tool that they use daily.

The technology works by reading websites that would otherwise appear in Google search results and then summarizing them into multi-paragraph answers. Publishers cried foul, accusing the company of hurting their businesses by taking their content and regurgitating it for users directly in search results, thereby depriving them of significant web traffic.

But journalists, search engine experts and social media users quickly began to spot problems with the responses. Some responses were funny while others were disturbing. They also presented themselves to sensitive requests, particularly related to health.

One answer, which Google has since corrected, asked people to drink lots of urine to help flush out kidney stones. Another said John F. Kennedy graduated from the University of Wisconsin at Madison in six different years, including three after his death.

Google tried to test the tool as much as possible before a wider rollout, but Reid said the large-scale launch revealed many situations the company wasn’t prepared for.

“There’s nothing like having millions of people using this feature,” Reid said.

News Source : www.washingtonpost.com
Gn tech

Back to top button