Sina science and technology news Beijing time May 18 morning news, it is reported that the United States technology giant Google is integrating the latest generative artificial intelligence technology into its product array, and internal documents show that Google network advertising business becomes the latest integration of the department.
Google internally approved the use of generative AI technology based on "large language models" for its automated advertising business and ad-related customer service, according to the documents.
At its developer conference last week, Google unveiled PaLM 2, a new and more powerful language model that has been trained on massive amounts of text data and is as smart as a human being, able to answer questions and follow instructions.
Internal documents show that several product teams within Google are preparing to adopt the language model for uses ranging from allowing advertisers to quickly create their media content to, on Youtube, advising video writers on what content to create.
On Youtube, the Google team has been testing PaLM 2 models to create titles and introductory text for young videos. For video creators, Google will use the model to find the latest popular topics and suggest five possible videos.
In recent days, AI conversation tools have become the focus of the tech industry and the obsession of Wall Street. Giants like Google, Microsoft, Meta, and Amazon are racing to build generative AI into their products.
Google is under particular pressure in this wave. ChatGPT's blockbuster debut came from OpenAI, a company backed by Google's rival Microsoft, which is also rapidly integrating its web search service. Some say that the Web search market, which has been dominated by Google, may be reshuffled.
After two decades of steady, high growth, Google is showing signs of fatigue, with several quarters of disappointing revenue growth.
Since the beginning of last year, many advertisers, fearing a U.S. recession, have cut their advertising budgets, hurting companies like Google and Meta. This year, advertisers across many industries have seen Google's search AD conversion rates decline.
In addition to web search, email and online spreadsheet software, Google plans to incorporate generative AI into many other areas of its business, with the ultimate goal of expanding revenue and profit margins, according to internal documents.
Google plans to develop an AI-based customer service system that will service hundreds of its products, including the Play app store, email, Android search and maps.
For example, robo-dialogue tools can communicate with advertisers in plain language and ultimately recommend the best campaign for them.
A Google spokesperson declined to comment on broader AI integration plans.
Recently, Google launched new services such as Google Duet, which allows users to ask questions about cloud computing in their everyday language, and the system will provide answers. For example, they can ask how to use a cloud computing service or feature, or get a detailed implementation plan for their own project.
Google is also working on an AI Diffusion product, similar to its well-known Stable Diffusion product, according to internal documents. Like OpenAI's DALL-E, Stable Diffusion allows you to quickly create different styles of images by providing just a few keywords.
It's no surprise that Google is integrating AI into its advertising business. Just last week, Google's advertising rival Meta launched an "AI sandbox," which is essentially a test bed for advertisers to try out Meta's various tools based on generative AI technology. Meta also recently updated its advertiser Arsenal to help advertisers optimize their marketing plans.
Google's annual advertising event, Google Marketing Live, is expected to introduce new tools and technologies for advertisers. Google has given few details about what it will unveil, but it has made it clear that artificial intelligence will be a central issue.
At the conference, advertisers will see Google's AI-powered advertising solutions, which will enrich the advertiser experience and lead to better marketing returns in today's economy, Google said.
Responsible Editor: Zheng Zhuo
User comments