May 10 (Reuters) – Alphabet Inc’s (GOOGL.O) Google is adding more artificial intelligence to its core search product, with Microsoft’s (MSFT.O) rival search engine Bing We hope to generate the same kind of consumer excitement that was generated by the . in the last few months.
At its annual I/O conference in Mountain View, Calif., on Wednesday, Google announced a new version of its eponymous engine. Google’s revamped Search Generative Experience allows you to create responses to free-form queries while maintaining a recognizable list of links to the web.
“We’re reimagining all of our core products, including search,” Alphabet CEO Sundar Pichai said after speaking at the event.
He said that Google is working on search and other products like Gmail, which lets you draft messages, and Google Photos, which lets you make changes to images like centering pictures and coloring empty spaces. It is said that it integrates generation AI into
Alphabet shares rose nearly 4%.
U.S. consumers will be able to access the Search Generative Experience in the coming weeks via a waitlist, Kathy Edwards said in an interview, and this pilot phase will see Google improve the quality, speed and speed of search results. He said he would monitor costs.
Google’s foray into so-called generative AI comes after startup OpenAI introduced ChatGPT, Silicon Valley’s beloved chatbot, starting a fierce funding race among competitors. Generative AI can use historical data to create entirely new content such as fully formalized text, images, and software code.
OpenAI, backed by billions of dollars from Microsoft and now integrated into Bing search, has become the default version of generative AI for many, allowing users to read term papers, contracts, travel plans, and even Helps create an entire novel.
Google, the top portal to the Internet for many years, has been questioning its own perch since rivals began abusing the technology as an alternative way to view content from the web. At stake is Google’s part of the huge online advertising pie, which research firm Magna estimates at $286 billion this year.
Edwards told Reuters that Google’s aim is to employ generative AI to reduce the decision-making steps for consumers and allow people to ask a wide range of questions, including creative ones. Regarding how the AI spits out false information, Edwards said the company prioritized accuracy and cited reliable sources.
“AI can provide insight,” says Edwards. “But basically what people ultimately want is to get information from real people and organizations, knowing that this health information came from, say, the WHO or the World Health Organization. That’s it.”
what to wear
With AI built in, Google still looks and behaves like the familiar empty search bar.
However, according to previous Reuters demonstrations, a search for “weather San Francisco” would show users the usual eight-day weather forecast, but the question of what to wear in California would be answered. Instead, it encourages long AI-generated responses. week.
The result states, “You should bring layers such as a short-sleeved shirt and a light sweater or jacket for the day,” and includes a link to a website with such advice. .
Searching for news or navigating to a particular website did not get a response from the generative AI, but demonstrated that the user could prompt a chat and press a button to ask a follow-up question. indicated by
Through its partnership with OpenAI, Bing search and Microsoft’s corresponding Edge browser can summarize web pages, synthesize disparate sources, compose emails, and translate. Microsoft said it could generate an additional $2 billion in revenue for every 1 percent share it gains in search advertising.
Another challenge is the high cost of using such AI, known as large language models. “We and other companies are working on different ways to bring costs down over time,” Edwards said.
Edwards said advertising will continue to be central to the experience. “You only get paid for clicks.”
Bards of 180 countries
In further I/O news, Google Cloud CEO Thomas Kurian told Reuters the division is arranging customers to test the latest technology. Among them are Deutsche Bank (DBKGn.DE) and Uber Technologies (UBER.N).
But in recent years, Google’s rivals have made breakthroughs in Google’s research, incorporated them into their products, and overtaken their inventors.
ChatGPT came to light after Google announced its AI system in 2017. Chatbots grew faster than any other consumer application in history, and Google, with much reflection, urged staff to rush the project forward and get it ready for public release.
Google announced a rival chatbot, Bard, in February. That same month, a promotional video showing Mr. Byrd answering a question incorrectly triggered a stock price drop that cut Google’s market capitalization by $100 billion.
Going forward, Bard will be multimodal like OpenAI’s GPT-4, allowing people in more than 180 countries and territories to access its chatbots, the company announced Wednesday.
This means customers will be able to prompt Bard with images as well as text. For example, you can ask a chatbot to write a caption for a photo you pass to it.
Behind Bard is also Google’s announcement of a more powerful AI model called PaLM 2, which the company says could solve more difficult problems. Pichai also said one of the company’s PaLM 2 models is lightweight enough to work with smartphones.
Reported by Jeffrey Dustin, who lives in San Francisco. Additional reporting by Sheila Dang.Editing: David Gregorio
Our standards: Thomson Reuters Trust Principles.
