Search king responds to Microsoft’s AI challenge By Reuters

Written by Jeffrey Dustin and Greg Bensinger

MOUNTAIN VIEW, Calif. (Reuters) – Alphabet Inc.’s Google is rolling out more artificial intelligence to its core search product, hoping to create some of the same consumer excitement generated by Microsoft Corporation (NASDAQ:) rival search engine Bing has updated in recent months.

At its annual I/O conference in Mountain View, California, on Wednesday, Google showed off a new version of its engine of the same name. Dubbed the Generative Search Experience, Google’s revamp can craft responses to open queries while maintaining a recognizable list of links to the web.

“We’re reimagining all of our core products, including search,” said Sundar Pichai, CEO of Alphabet (NASDAQ:), after taking the stage at the event.

He said Google is integrating AI into search and other products, including Gmail, where it can create draft messages, and Google Photos, where it can make changes to images such as centering shapes and coloring in blank space.

Alphabet shares rose about 4%.

Vice President Cathy Edwards said in an interview that American consumers will get access to the generative search experience in the coming weeks through a waiting list, a beta phase during which Google will monitor the quality, speed, and cost of search results.

Google’s foray into what’s known as generative AI comes after startup OpenAI introduced ChatGPT, the chatbot beloved in Silicon Valley that set off a furious race for funding among potential rivals. Generative AI can, using past data, generate entirely new content such as complete text, images, and program code.

OpenAI, backed by billions of dollars from Microsoft and now integrated into Bing search, has become for many the default version of generative AI, helping users prepare term papers, contracts, travel itineraries and even entire novels.

For years, the premier gateway to the Internet, Google has found its own niche in the question ever since competitors began exploiting the technology as an alternative to delivering content from the web. At stake is Google’s slice of the massive online advertising pie that research firm MAGNA estimated at $286 billion this year.

Edwards told Reuters that Google’s goal is to use generative AI to reduce consumers’ steps to make decisions and allow people to ask a broader range of inquiries, including creative ones. Addressing how artificial intelligence can divulge incorrect information, Edwards said the company prioritized accuracy and cited reliable sources.

“Artificial intelligence can provide insight,” Edwards said. “But what people fundamentally want at the end of the day is to be connected to information from real people and organizations, knowing, for example, that this health information comes from the World Health Organization, or WHO.

What is the appropriate outfit to wear?

With built-in artificial intelligence, Google still looks and behaves like the familiar blank search bar.

But while a search for “San Francisco weather” would normally direct the user to an eight-day forecast, a query asking what outfit they should wear in the California city calls up a drawn-out response generated by artificial intelligence, according to a Reuters demonstration earlier this year. week.

The conclusion reads, “You should bring layers, including a short-sleeved shirt and a light jacket or jacket for the day,” including links to websites where such advice was drawn.

The demo showed that searching for news or navigating to a specific website did not result in a response by the generative AI, although users could prompt to chat and press a button for a follow-up query.

Bing search and Microsoft’s corresponding Edge browser, through its partnership with OpenAI, can abstract web pages, synthesize various sources, and compose and translate email messages. Microsoft said every percentage point earned in search advertising could bring in another $2 billion in revenue.

Another challenge is the high cost of relying on such artificial intelligence, known as large language models. Edwards said, “We and others are working on a variety of different ways to bring cost down over time.”

Edwards said ads will continue to be an essential part of the experience. “We only get paid when there is a click.”

BARD for 180 countries

In more I/O news, Google Cloud CEO Thomas Kurian told Reuters that the division is courting customers to test its latest technology, among them Deutsche Bank AG (NYSE:) and Uber Technologies (NYSE:) Inc.

However, in recent years, Google’s competitors have made their own search breakthroughs and worked with them in products, bypassing their inventor.

ChatGPT emerged after Google revealed an artificial intelligence system in 2017. The speed at which the chatbot grew—faster than any consumer app in history—spurred the oft-traded Google to get employees to rush projects so they were ready for public consumption.

In February, Google announced its competing chatbot called Bard. A promotional video that month, which showed Bard incorrectly answering a question, prompted a discounted $100 billion stock slice of Google’s market capitalization.

Now, Bard will be multimedia like OpenAI’s GPT-4, the company said Wednesday, and will make the chatbot available to people in more than 180 countries and regions.

This means customers will be able to ask the Bard for photos, not just text messages — for example, asking the chatbot to write a comment on a photo it hands it to.

Behind Bard is also a more powerful AI model that Google announced called PaLM 2, which it said could solve tougher problems. Pichai also said that one of the PaLM 2 models is lightweight enough to run on smartphones.