MOUNTAIN VIEW, Calif. (AP) - Google on Tuesday introduced a redesigned search engine that often favors artificial intelligence-generated answers over website links, a shift that promises to speed up the search for information while potentially disrupting the flow of money making internet traffic.
The makeover announced at Google's annual developer conference begins this week in the US, when hundreds of millions of people will periodically see conversation summaries generated by the company's AI technology at the top of the search engine results page.
The AI summaries would only show up if Google's technology determines that it is the fastest and most effective way to satisfy a user's curiosity - a resolution that will usually occur on complex topics or when people are brainstorming or making plans . People will likely still see Google's traditional website links and ads for simple searches for things like a shopping recommendation or weather forecast.
Google started testing AI summaries a year ago with a small subset of select users, but the company is now making it a staple in its search results in the US before rolling out the feature in other parts of the world. Google expects that the recurring AI overviews will be part of its search results for around 1 billion people by the end of this year.
In addition to injecting more AI into its dominant search engine, Google also used the packed conference at an amphitheater in Mountain View, California, near its headquarters, to showcase advances in a technology that is reshaping business and society.
The next AI steps include more advanced analytics powered by Gemini - a technology unveiled five months ago - and smarter assistants, or "agents," including an emerging version called "Astra" that will be able to do things to understand, explain and remember. is displayed through the camera lens of a smartphone. Google underlined its commitment to AI by having Demis Hassabis, the executive overseeing the technology, take the stage for the first time at its major conference.
The injection of more AI into Google's search engine marks one of the most dramatic changes the company has made to its foundation since its founding in the late 1990s. It's a step that opens the door for more growth and innovation, but also threatens to bring about a huge change in Internet surfing behavior.
"This bold and responsible approach is fundamental to achieving our mission and making AI more useful for everyone," Google CEO Sundar Pichai told a group of reporters.
Well aware of how much attention is being paid to the technology, Pichai ended a nearly two-hour series of presentations by asking Google's Gemini model how many times AI had been mentioned. The count: 120, and then the count went up by one more when Pichai said "AI" again.
The increased emphasis on AI will bring new risks to an internet ecosystem that relies heavily on digital advertising as its financial lifeline.
Google will suffer if the AI overviews undermine ads tied to its search engine - a business that raked in $175 billion in revenue last year alone. And website publishers - ranging from major media outlets to entrepreneurs and startups focused on narrower topics - will be hurt if the AI overviews are so informative that they result in fewer clicks on the website links that still appear lower on the results page.
Based on habits revealed during the testing phase of Google's AI summaries over the past year, about 25% of traffic could be negatively impacted by de-emphasizing website links, says Marc McCollum, chief innovation officer at Raptive , which helps about 5,000 websites Publishers make money from their content.
A drop in traffic of that magnitude could translate into billions of dollars in lost advertising revenue, a devastating blow that would be dealt by a form of AI technology that removes information harvested from many of the websites that threaten to lose revenue. to lose.
"The relationship between Google and publishers has been quite symbiotic, but when it comes to AI, what has actually happened is that the Big Tech companies have used this creative content to train their AI models," said McCollum. "We are now seeing this being used for their own commercial purposes, in what is effectively a transfer of wealth from small, independent companies to Big Tech."
But Google found that the AI summaries caused people to perform even more searches during testing of the technology "because they can suddenly ask questions that were previously too difficult," said Liz Reid, who oversees the company's search operations. during an interview with The Associated Press. . She declined to provide specific numbers on link click volume during AI summary testing.
"The reality is that people do want to click to the web, even if they have an AI overview," says Reid. "They start with the AI overview and then want to dig deeper into it. We will continue to innovate in AI oversight and also in how we send the most useful traffic to the web."
The increasing use of AI technology to summarize information in chatbots such as Google's Gemini and OpenAI's ChatGPT over the past 18 months has already raised legal questions about whether the companies behind the services are illegally using copyrighted material to improve their services . It's an allegation at the heart of a high-profile lawsuit The New York Times filed late last year against OpenAI and its largest backer, Microsoft.
Google's AI overviews could also lead to lawsuits, especially if they siphon traffic and ad sales away from websites that believe the company is unfairly profiting from their content. But it's a risk the company has had to take as the technology evolves and is used in competing services like ChatGPT and upstart search engines like Perplexity, said Jim Yu, executive chairman of BrightEdge, which helps websites rank higher in Google search results .
"This is definitely the next chapter in the search," Yu said. "It's almost as if they're tuning three key variables at once: search quality, traffic flow in the ecosystem, and then monetizing that traffic. There hasn't been a moment in the quest bigger than this in a long time."
Outside the amphitheater, several dozen protesters chained themselves together and blocked one of the entrances to the conference. Protesters targeted a $1.2 billion deal known as Project Nimbus, which provides artificial intelligence technology to the Israeli government. They claim that the system is being used lethally in the Gaza war - a claim that Google refutes. The demonstration did not appear to have an impact on conference attendance or the enthusiasm of the crowd in the room.
