Google brings in Larry Page and Sergey Brin to solve ChatGPT and AI chatbot problems

Last month, Google founders Larry Page and Sergey Brin held several meetings with company executives. The subject: A rival’s new chatbot, a clever artificial intelligence product that looks like it could be the first significant threat to Google’s $149 billion search business in decades.

gentlemen. Page and Mr. Brin, who hasn’t spent much time at Google since leaving his day job at the company in 2019, reviewed Google’s AI product strategy, according to two people familiar with the matter. They approved plans to put more chatbot functionality into Google’s search engine and came up with ideas. They also offer advice to company leaders who have put AI at the center of their initiatives.

The Google founder’s re-engagement, at the invitation of the company’s current CEO, Sundar Pichai, underscores the sense of urgency many Google executives feel about artificial intelligence and chatbot ChatGPT.

Released two months ago by the small San Francisco-based company OpenAI, the bot wows users by simply explaining complex concepts and generating ideas from scratch. More importantly for Google, it appears to offer a new way to search for information on the Internet.

New artificial intelligence technology is taking Google out of the norm. gentlemen. Pichai announced “Code Red,” upending existing plans and launching AI development. Google now intends to launch more than 20 new products this year and showcase a chatbot-enabled version of its search engine, according to a slide presentation seen by The New York Times and two people familiar with the matter who were not authorized to discuss the plans.

Meanwhile, Alphabet is shrinking its workforce. On Friday, the company said it would cut about 12,000 jobs in a hiring spree amid the pandemic and concerns over an economic slowdown. The job cuts are aimed at “ensuring that our people and roles are aligned with our highest priorities as a company,” Mr. Pichai wrote in a letter to employees.

“This is a very vulnerable moment for Google,” said D. Sivakumar, a former Google research chief who helped found a startup called Tonita, which provides search technology to e-commerce companies. “ChatGPT stands there and says, ‘This is what an engaging new search experience looks like.'” Sivakumar added that Google has overcome previous challenges and can deploy its AI arsenal to stay competitive.

Since quitting my day job, Mr. Page and Mr. Brin, two people familiar with the matter, said Brin took a hands-off approach to Google. They let mr. Pichai runs the company and its parent company, Alphabet, while they work on other projects such as flying car startups and disaster relief efforts.

The main purpose of their visits to the company’s Silicon Valley offices over the past few years was to examine what Alphabet called its “other bets,” a so-called moonshot, a person familiar with the matter said. Until recently, they were not deeply involved with search engines.

But they have long been keen on bringing AI into Google’s products. Former Google senior vice president Vic Gundotra recalls that he gave Mr. Page shows off new Gmail features circa 2008. But sir. Page shrugged off the effort, asking, “Why couldn’t it automatically write that email for you?” In 2014, Google also acquired DeepMind, a leading artificial intelligence research lab in London.

Less than two weeks after ChatGPT debuted, a panel of executives from Google’s senior technical review board, which includes Jeff Dean, the company’s senior vice president for research and artificial intelligence, and Kent Walker, Google’s president of global affairs and chief legal officer, met to discuss According to the slide presentation, their company’s initiatives.

They reviewed plans for products expected to debut at Google’s corporate conference in May, including Image Generation Studio, which creates and edits images, and the third version of AI Test Kitchen, an experimental app for testing product prototypes.

Other image and video projects in the works include a feature called Shop and Try, a YouTube green screen feature for creating backgrounds; a wallpaper maker for Pixel smartphones; an app called Maya, A three-dimensional shoe can be visualized; there is also a tool that can summarize a video by generating a new one, according to the slideshow.

Google has a list of artificial intelligence programs it plans to make available to software developers and other companies, including image-creation technology, which could boost revenue for Google’s cloud unit. According to the introduction, there are also tools that can help other businesses create their own AI prototypes in the Internet browser, called MakerSuite, which will have two “Pro” versions.

According to reports, Google also hopes to release a tool called Colab + Android Studio in May to make it easier to build applications for Android smartphones, which will generate, complete and fix code. Another code generation and completion tool called PaLM-Coder 2 is also in use.

Google executives want to reaffirm their company’s status as an AI pioneer. The company has been aggressively working on AI for the past decade and has already offered a handful of chatbots that rival ChatGPT, called LaMDA, or Language Model for Conversational Applications.

“We continue to test our AI technology internally to make sure it’s useful and safe, and we look forward to sharing more of our experience externally soon,” Google spokeswoman Lily Lin said in a statement. She added that AI will benefit individuals, businesses and communities, and that Google is considering the technology’s wider societal impact.

Google, OpenAI and others develop their artificial intelligence using so-called large language models that rely on information online, so they sometimes share false statements and display racist, sexist and other biased attitudes.

That’s enough to make the company wary of making the technology available to the public. But some new companies, including and, are already offering online search engines that let you ask questions via live chatbots, much like ChatGPT. Microsoft is also working on a new version of its Bing search engine that will include similar technology, The Information reports.

gentlemen. Pichai sought to expedite product approval reviews, according to a presentation seen by The Times. The company established a fast-track review process called the “Green Lane” initiative to push for faster approval of its upcoming artificial intelligence technology by employee groups trying to ensure the technology is fair and ethical.

The company will also look for ways for teams developing AI to conduct their own reviews and will “recalibrate” the level of risk they are willing to take when releasing the technology, according to the report.

The consequences of Google’s leaner approach are unclear. According to an analysis compiled by Google, its technology lags behind OpenAI’s self-reported metrics when it comes to identifying hateful, toxic, sexual or violent content. In every category, OpenAI beat Google’s tools, which were also less accurate than humans at evaluating content.

In a slide presentation, Google cited copyright, privacy and antitrust as the main risks of the technology. Actions, such as filtering answers to weed out copyrighted material and preventing AI from sharing personally identifiable information, are needed to reduce these risks, it said.

For the chatbot search demo Google is planning for this year, getting the facts right, keeping it safe and fighting misinformation are top priorities. For other upcoming services and products, the company has a lower bar and will work to curb issues related to hate and toxicity, danger and misinformation, not stop them, according to the presentation.

For example, the company intends to block certain words to avoid hate speech and will minimize other potential problems.

Google wants governments to scrutinize its AI products for signs of these problems. The company has recently been the subject of numerous government investigations and lawsuits alleging anticompetitive business practices. According to the presentation, it expects “increased pressure across all regulatory efforts due to growing concerns about misinformation, harmful content, bias and copyright.”

trip mickle and Cade Metz Contribution report.

Source link