Last month, Larry Page and Sergey Brin, Google’s founders, held respective meetings with institution executives. The topic: a rival’s caller chatbot, a clever A.I. merchandise that looked arsenic if it could beryllium the archetypal notable menace successful decades to Google’s $149 cardinal search business.

Mr. Page and Mr. Brin, who had not spent overmuch clip astatine Google since they near their regular roles with the institution successful 2019, reviewed Google’s artificial quality merchandise strategy, according to 2 radical with cognition of the meetings who were not allowed to sermon them. They approved plans and pitched ideas to enactment much chatbot features into Google’s hunt engine. And they offered proposal to institution leaders, who person enactment A.I. beforehand and halfway successful their plans.

The re-engagement of Google’s founders, astatine the invitation of the company’s existent main executive, Sundar Pichai, emphasized the urgency felt among galore Google executives astir artificial quality and that chatbot, ChatGPT.

The bot, which was released by the tiny San Francisco institution OpenAI 2 months ago, amazed users by simply explaining analyzable concepts and generating ideas from scratch. More important to Google, it looked arsenic if it could connection a caller mode to hunt for accusation connected the internet.

The caller A.I. exertion has shaken Google retired of its routine. Mr. Pichai declared a “code red,” upending existing plans and jump-starting A.I. development. Google present intends to unveil much than 20 caller products and show a mentation of its hunt motor with chatbot features this year, according to a descent presumption reviewed by The New York Times and 2 radical with cognition of the plans who were not authorized to sermon them.

At the aforesaid time, Alphabet is scaling backmost its enactment force. On Friday, the institution said it would chopped astir 12,000 jobs aft a hiring spree during the pandemic and amid concerns of a slowing economy. The layoffs were designed “to guarantee that our radical and roles are aligned with our highest priorities arsenic a company,” Mr. Pichai wrote successful a enactment to employees.

“This is simply a infinitesimal of important vulnerability for Google,” said D. Sivakumar, a erstwhile Google probe manager who helped recovered a start-up called Tonita, which makes hunt exertion for e-commerce companies. “ChatGPT has enactment a involvement successful the ground, saying, ‘Here’s what a compelling caller hunt acquisition could look like.’” Mr. Sivakumar added that Google had flooded erstwhile challenges and could deploy its arsenal of A.I. to enactment competitive.

Since stepping backmost from day-to-day duties, Mr. Page and Mr. Brin person taken a laissez-faire attack to Google, 2 radical acquainted with the substance said. They person fto Mr. Pichai tally the institution and its genitor company, Alphabet, portion they person pursued different projects, specified arsenic flying car start-ups and catastrophe alleviation efforts.

Their visits to the company’s Silicon Valley offices successful the past fewer years person mostly been to cheque successful connected the alleged moonshot projects that Alphabet calls ‘Other Bets,’ 1 idiosyncratic said. Until recently, they person not been precise progressive with the hunt engine.

But they person agelong been keen connected bringing A.I. into Google’s products. Vic Gundotra, a erstwhile elder vice president astatine Google, recounted that helium gave Mr. Page a objection of a caller Gmail diagnostic astir 2008. But Mr. Page was unimpressed by the effort, asking, “Why can’t it automatically constitute that email for you?” In 2014, Google besides acquired DeepMind, a starring A.I. probe laboratory based successful London.

Google’s Advanced Technology Review Council, a sheet of executives that includes Jeff Dean, the company’s elder vice president of probe and artificial intelligence, and Kent Walker, Google’s president of planetary affairs and main ineligible officer, met little than 2 weeks aft ChatGPT debuted to sermon their company’s initiatives, according to the descent presentation.

They reviewed plans for products that were expected to debut astatine Google’s institution league successful May, including Image Generation Studio, which creates and edits images, and a 3rd mentation of A.I. Test Kitchen, an experimental app for investigating merchandise prototypes.

Other representation and video projects successful the works included a diagnostic called Shopping Try-on, a YouTube greenish surface diagnostic to make backgrounds; a wallpaper shaper for the Pixel smartphone; an exertion called Maya that visualizes three-dimensional shoes; and a instrumentality that could summarize videos by generating a caller one, according to the slides.

Google has a database of A.I. programs it plans to connection bundle developers and different companies, including image-creation technology, which could bolster gross to Google’s Cloud division. There are besides tools to assistance different businesses make their ain A.I. prototypes successful net browsers, called MakerSuite, which volition person 2 “Pro” versions, according to the presentation.

In May, Google besides expects to denote a instrumentality to marque it easier to physique apps for Android smartphones, called Colab + Android Studio, that volition generate, implicit and hole code, according to the presentation. Another codification procreation and completion tool, called PaLM-Coder 2, has besides been successful the works.

Google executives anticipation to reassert their company’s presumption arsenic a pioneer of A.I. The institution aggressively worked connected A.I. implicit the past decennary and already has offered to a tiny fig of radical a chatbot that could rival ChatGPT, called LaMDA, oregon Language Model for Dialogue Applications.

Google, OpenAI and others make their A.I. with alleged ample connection models that trust connected online information, truthful they tin sometimes stock mendacious statements and amusement racist, sexist and different biased attitudes.

That had been capable to marque companies cautious astir offering the exertion to the public. But respective caller companies, including You.com and Perplexity.ai, are already offering online hunt engines that fto you inquire questions done an online chatbot, overmuch similar ChatGPT. Microsoft is besides moving connected a caller mentation of its Bing hunt motor that would see akin technology, according to a report from The Information.

Mr. Pichai has tried to accelerate merchandise support reviews, according to the presumption reviewed by The Times. The institution established a fast-track reappraisal process called the “Green Lane” initiative, pushing groups of employees who effort to guarantee that technology is just and ethical to much rapidly o.k. its upcoming A.I. technology.

The consequences of Google’s much streamlined attack are not yet clear. Its exertion has lagged OpenAI’s self-reported metrics erstwhile it comes to identifying contented that is hateful, toxic, intersexual oregon violent, according to an investigation that Google compiled. In each category, OpenAI bested Google tools, which besides fell abbreviated of quality accuracy successful assessing content.

Google listed copyright, privateness and antitrust arsenic the superior risks of the exertion successful the descent presentation. It said that actions, specified arsenic filtering answers to weed retired copyrighted worldly and stopping A.I. from sharing personally identifiable information, are needed to trim those risks.

For the chatbot hunt objection that Google plans for this year, getting facts right, ensuring information and getting escaped of misinformation are priorities. For different upcoming services and products, the institution has a little barroom and volition effort to curb issues relating to hatred and toxicity, information and misinformation alternatively than preventing them, according to the presentation.

The institution intends, for example, to artifact definite words to debar hatred code and volition effort to minimize different imaginable issues.

Google expects governments to scrutinize its A.I. products for signs of these issues. The institution has precocious been the taxable of galore government inquiries and lawsuits accusing it of anti-competitive concern practices. It anticipates, according to the presentation, “increased unit connected A.l. regulatory efforts due to the fact that of rising concerns astir misinformation, harmful content, bias, copyright.”

Tripp Mickle and Cade Metz contributed reporting.