Google CEO promises new AI features are coming to search ‘very soon’ amid competition from ChatGPT

Alphabet and Google CEO Sundar Pichai mentioned Thursday that the corporate will quickly add superior AI options to its search engine.

On Tuesday, CNBC reported that Google is testing a few of these options with workers as a part of a “code red” plan to answer ChatGPT, the favored chatbot backed partially by Microsoft. They embrace a chatbot referred to as “Apprentice Bard,” in addition to new search desktop designs that may very well be utilized in a question-and-answer format. 

“Very quickly, folks will have the ability to work together straight with our latest, strongest language mannequin as a companion to Search, in experimental and progressive methods,” he said, referring to Google’s conversation technology LaMDA, or Language Model for Dialogue Applications.

Pichai said that it will release the large language model “within the coming weeks and months” so the company can get more feedback.

Executives frequently returned to the subject of artificial intelligence on the company’s fourth-quarter earnings call. “AI is the most profound technology we are working on today,” Pichai said in his opening remarks.

The effort to direct attention to AI comes as the company faces pressure on its core advertising business and a competitive threat from one of its historic archrivals.

Thursday’s earnings report marked the fourth consecutive quarter in which the company missed Wall Street’s expectations for both earnings and revenue, according to expectation estimates provided by Refinitiv. Weakness in the advertising business appeared in an 8% revenue decline in YouTube’s advertising revenue and a 2% fall in Google’s Search and Other revenue.

Google is also facing pressure from ChatGPT, which was launched late last year by Microsoft-backed OpenAI. Google’s prime business is web search, and the company has long touted itself as a pioneer in AI. But generative AI products like ChatGPT could pose a threat to the entire model of internet search, as they can provide creative answers to more complicated queries.

Microsoft is reportedly considering adding ChatGPT functionality into its own search engine, Bing. The threat of falling behind in AI has even reportedly spurred Google co-founders Larry Page and Sergey Brin to take a direct interest in the efforts years after they stepped down from day-to-day work at the company in 2019.

In addition to touting forthcoming search improvements, the company also said that starting in the first quarter, it will change the financial reporting structure for its DeepMind artificial intelligence segment so it rolls up to Google, instead of the Other Bets segment that includes long-payoff projects like self-driving cars and venture capital investments.

Google acquired the London-based company in 2014 for more than $500 million and then placed it under the Other Bets umbrella when the company reorganized as Alphabet in 2015. DeepMind turned a profit for the first time in 2021.

This reporting change “reflects the strategic focus in DeepMind to support each one of our segments,” Alphabet’s finance chief Ruth Porat said on Thursday’s earnings call.

“To be very clear, we consolidate Other Bets into Google solely when that guess helps services inside Google or Alphabet broadly,” Porat added, pointing to cybersecurity company Chronicle, which it rolled into Google’s cloud unit in 2019. “That was very efficient.”

Pichai also said the company will also provide “new instruments” and APIs for developers, creators and partners to “empower them to find new prospects with AI. He added, “these models are particularly amazing for composing, constructing and summarizing.”

But Pichai additionally warned that it might want to scale slowly, saying he views giant language utilization as nonetheless being in “early days.”

CNBC beforehand reported that workers had requested in regards to the risk from ChatGPT in an inner assembly, and Google’s Jeff Dean informed workers that Google has rather more “reputational risk” in offering improper info, and thus was transferring “more conservatively than a small startup.”