Shulou(Shulou.com)11/24 Report--

Beijing time on March 9 morning news, According to reports, The latest news shows, Google has made an internal request, In a few months time "Generative artificial intelligence" Technology integrated into all major products.

Artificial intelligence was supposed to be Google's strength. Google has been making long-term investments in technology for years, and much of the research underpinning the previous generation of AI chatbots was done in Google-owned labs. However, startup OpenAI launched ChatGPT last November, becoming an early leader in the field of "generative AI." Generative AI software can generate text, images, or videos on its own. Its sudden success has forced Google's parent company Alphabet to speed up and try to catch up in this space. Alphabet CEO Sundar Pichai also stressed that the technology could "go further than fire and electricity."

ChatGPT is seen by some as the ultimate challenger to the traditional Google search engine. And given OpenAI's close relationship with Microsoft, ChatGPT seems to pose a double threat. According to current and former Google employees and others close to the company, Google's current sense that it is lagging behind in a key area of strength has caused considerable anxiety within the company. As one current employee put it: "There are both unusually high expectations and a great deal of insecurity associated with any AI-related project. "

The work also took Pichai back to his days as a product manager, delving into details of product features that would normally have been done by employees far below him, a former Google employee said. Google co-founders Larry Page and Sergey Brin are also getting more involved with the company than they were a few years ago, and Brin has even submitted code changes to Google's equivalent Bard. Google senior management has announced a "red directive" that requires all of Google's important products, including those with more than 1 billion users, to integrate generative AI technology in the coming months, according to a person familiar with the matter. Google, for example, announced in March that YouTube video platform creators would soon be able to use the technology to virtually dress up.

It also reminded some Googlers of the company's last such move: Google+, the failed Social networks that began in 2011. The analogy isn't entirely appropriate: Google has never been seen as a leader in Social networks, but its dominance in AI technology is indisputable. However, this kind of action gives employees a similar feeling. For example, Google employee bonuses were once linked to Google + success, while current and former Google employees say that at least some employee performance is now linked to the productization of generative AI technology. Red Order has already driven dozens of projects integrating generative AI technology. "We tried everything, but it was a far cry from what we needed to do to transform the company and stay competitive," said one Google employee. "

In the end, the all-out effort around Google + failed, and the Social networks did not appeal to users. Google finally announced in 2018 that it would shut down Google+ for ordinary users. A former Google executive said the failure was a cautionary tale for the company. Larry page's requirement was that every product had to have a social element, and the end result was terrible. "

A Google spokesman disputed the analogy between "red command" and Google +. Google + projects involve all products, he said, but current AI projects mainly encourage Google employees to test the company's AI tools internally, which is common in the tech industry. He also said that most Google employees did not spend extra time on artificial intelligence, except for those working on related projects.

Google isn't the only company that believes AI is everything now. Silicon Valley is entering a full-blown hype cycle. Many venture capitalists and entrepreneurs are suddenly calling themselves AI visionaries, while moving away from previously hot tech topics like blockchain. At the same time, some companies 'share prices soared after announcing the integration of artificial intelligence technology. Meta CEO Mark Zuckerberg is also beginning to shift his focus from the meta-universe to artificial intelligence, sources said. This contradicts his earlier claim that metacentric technology is the basis of Meta, and even renamed the company for it.

For some Google employees, the new requirements are good news. They know that Google has done speculative research in the past, but eventually ran into problems in the commercialization process. One person familiar with the matter said team members already working on generative AI projects thought they could now "deliver more and have a greater impact than just doing research work."

In the long run, it may not matter that OpenAI has stolen the limelight in recent months, because Google has a very deep accumulation in this area. Back in 2016, Pichai positioned Google as an "AI first" company. For years, the company has used machine learning to drive its advertising business, while also integrating AI into key consumer products like Gmail and Google Photos, using AI to help users compose emails and organize photos.

In a recent study, industry research firm Zeta Alpha analyzed the 100 most cited AI research papers between 2020 and 2022 and found that Google dominates the field. Amin Ahmad, a former AI researcher at Google and co-founder of startup Vectara, said: "It now seems that Google is a sleeping giant, although it is temporarily behind but catching up. I think Google has done a great job of applying this technology to core products, way ahead of the rest of the industry. "Vectara also offers conversational search tools for businesses.

At the same time, Google is struggling to balance business with responsible new technology development. Automated tools based on AI techniques often have problems due to biases in training datasets. Many also worry that some tools may not be ready before they are open for public testing. This aspect of generative AI is particularly risky, leading Google to be reluctant to rush these technologies to market. In search, for example, chatbots seem to be able to provide answers directly from developers, and ChatGPT seems to have become a sounding board for OpenAI. This is inherently riskier than replying to a list of links to other websites.

Google's "red mandate" also appears to affect the usual calculation of risk reward, which worries some industry experts. Emily Bender, a professor of computational linguistics at the University of Washington, said Google and other companies following the trend of generative AI technology may not be able to guide their AI products to avoid bias. However, a Google spokesperson said Google's work in this area is guided by the company's artificial intelligence development principles, while the company also takes a prudent approach to technology development. Google published a set of guidelines in 2018 on how to develop AI technology in a responsible way.

Other companies have indicated they will continue to move forward whether Google does so or not. One of the most important contributions of Google researchers in this field was a landmark paper,"Attention is all you need." The author introduces the concept of "transformer" in his paper. Such systems help AI models focus on the most important information in the data being analyzed. At present, the converter has become a key component of the large language model, and the large language model is the basic technology of the current chat robot, and the "T" in ChatGPT also represents the converter. Five years after this paper was published, all but one of the authors have left Google. Some say they mainly want to escape the yoke of a big, slow-moving company.

They are among dozens of AI researchers who have jumped ship to OpenAI, as well as to smaller startups such as Characterist.AI, Anthropic and Adet. Several companies founded by former Google employees, including Neeva, Perplexity AI, Tonita, and Vectara, are reimagining how to use large language models for search. Sara Hooker, a former Google Brain employee who currently works for AI startup Cohere, says that only a handful of research institutions have the knowledge and ability to build these models, so competition for talent is "much fiercer than in areas where other model training methods don't require much expertise."

Some people or organizations contribute significantly to the development of breakthrough technologies, but the ultimate economic benefits are taken away by others. This has happened before. Keval Desai, a former Google employee and managing director of venture capital firm Shakti, cited Xerox Parc as an example. The lab's work laid much of the groundwork for the PC era, but eventually saw Apple and Microsoft swoop in and use its technology to build trillion-dollar empires. Google wants to make sure it doesn't end up like xerox Parc,"desai says." All innovation happens there, but there is no commercial execution. "

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.