The future of Google Search is a big green box.
That’s exactly what Google showed off at its annual developer conference, Google I/O, earlier this month. The theme for 2023 was AI. The term was mentioned by him over 140 times during his two-hour keynote. Google has indeed announced a publicly available AI product, which marks a turning point for the internet giant as it fears how to counter increasing competition.
Late last year, OpenAI announced ChatGPT to near-universal acclaim. Suddenly, everyone had access to a generative AI engine, ready to answer any question with a novel response. It’s powered by the Large Language Model (LLM) and essentially works as a “”.powerful autocompleteuses large amounts of text data to identify the next best word.
The power and ease of use of ChatGPT has made it the fastest growing consumer web platform ever. That’s why Microsoft increased its investment in OpenAI earlier this year, integrating ChatGPT’s technology directly into Bing search, which boosted the company’s traffic by 16%. The day before Microsoft announced Bing AI, Google announced Bard, its own generative AI engine, but failed to do so, losing $100 billion in stock market value in the process. After that, the stock price recovered to the highest level since the beginning of the year.
In many ways, Google I/O was a referendum against the company’s shaky entry into consumer AI, and the company is determined to stay at the forefront of Internet search, even if that means turning the tables. It was a clear message to skeptics (and investors) that the company was ready to take drastic steps. its flagship product. Google search has long been the engine through which we find product information, find the latest news, interact with the Internet, and generate revenue for many businesses.
The new Search Generative Experience (SGE) is an experimental version of search that deprioritizes the 10 blue links that have defined Google over the past quarter century. Instead, any question, no matter how specific, is answered in a green box that expands as the user’s answer fills the screen.
Kathy Edwards, Google’s vice president of engineering, said in I/O, “Search now does the heavy lifting.” She said that search as we know it today requires you, the user, to break down complex queries into smaller questions that require you, the user, to sift through the information and assemble the answer in your head. I was. SGE does all this automatically and can even ask you follow-up questions.
Google Search Generative Experience is an experimental version of search that integrates AI-generated results, similar to Bing and ChatGPT.
CNET
At the same time, this means no need to visit multiple sites, meaning that web pages are click-dependent, potentially upending the Internet’s advertising-driven business model.
Google is by far the largest player in online search, with 93% market share, according to Statcounter. According to his 2019 report from Brightedge Research, online search engines are also the biggest driver of his website traffic, with 68% of his online experiences originating from search engines. At one point, Google dominated search, and its valuation reached $2 trillion.
With SGE, Google could usher Internet users and businesses into a new future. That future will require rethinking how quality information continues to permeate while motivating people to create worthwhile content to feed into the company’s AI machines.
Sign-ups for SGE have just started, so we don’t have any data to share about user experience yet. However, Microsoft has been collecting feedback on Bing AI over the past three months, and it could potentially offer a lens on how consumers react to his AI-driven Google searches.
A Microsoft spokesperson said, “Feedback on the new Bing-generated answers has been largely positive, with 71% of responses in the preview saying they liked AI-powered answers.” . , multiple questions will be asked during the session to discover new information. “
It is unclear how the AI-generated news articles will be incorporated into Google and Bing’s AI results. Already, publications, including CNET, are experimenting with articles written by AI. Unfortunately, the AI itself isn’t always accurate and can “hallucinate” confidently saying it’s right when something is wrong.
If the hallucination problem is finally solved, generative AI in search could be faster and ultimately better for consumers. But it’s still unclear what impact that will have on the digital publishing industry, especially if people stop clicking links en masse.
“As we experiment with new LLM-powered features in search, we continue to prioritize approaches that drive valuable traffic to a wide range of creators and support a healthy and open web,” a Google spokesperson said. . It’s true that Google links to the source prominently in his SGE, but it’s unclear if SGE will lead to more traffic or better quality for your site.
Microsoft did not respond to questions about traffic to sources when using AI search. Google said it has no plans to disclose publisher compensation, but that it will “continue to work with the broader ecosystem.”
“In my opinion [generative AI] Monica Ho, chief marketing officer at digital marketing firm SOCi, will reduce the amount of outbound traffic. Because that’s what it’s for,” said Monica Ho, chief marketing officer at his SOCi, a digital marketing firm. Users are looking for specific information rather than jumping back and forth between sites.
If Google can no longer rely on your traffic, there may not be viable alternatives. According to Rasmus Kreis Nielsen, director of the Reuters Institute for Journalism and professor of political communication at the University of Oxford, social media platforms such as Facebook are unreliable partners for publications and whimsical downgrades of news. It turned out. He added that platforms like Instagram and TikTok “have relatively few referrals and don’t feature links as often as search and social do.”
Currently, search engines “crawl” websites every day to gather new information and index the results. Websites can be crawled by the engine for free with traffic conversion. But if AI search reduces clicks, the search economy could require a whole new rethink.
“We anticipate that original content will be placed inside a paywall and require LLM models to pay to read it,” said Don White, CEO of conversational AI company Satisfi Labs. White sees a future where sites are paid per view in a “Spotify-style reward model.”
Ultimately, we believe Google will need to find a way to deliver revenue to creators and publishers so that it maintains an incentive to create quality content.
“We need to feed the engine with quality data, and Google doesn’t create all of our own authentic, original content,” Ho said. “It has to come from the creators. They know they have to somehow feed that engine and make the content worthy of being served continuously.”
Editor’s Note: CNET uses an AI engine to create personalized financial explanations that are edited and fact-checked by our editors. For more information, see this mailbox.
