Recently, traditional Internet search giant Google announced that its “Gemini” artificial intelligence (AI) model will integrate Google search functions. Subsequently, the U.S. Open Artificial Intelligence Research Center (OpenAI) issued an announcement to officially launch the real-time search function of ChatGPT. With the rapid development of AI technology, AI search engines have gradually become an emerging force in the field of information acquisition, providing users with a more intelligent and personalized search experience.
Why are various companies deploying AI search? How is AI search different from regular search? Can domestic search engines be transformed into AI? These are issues that people care about.
Not an upgrade of traditional search engines
With the breakthrough development of generative AI in 2023, AI tools are rapidly deployed in many fields, and search is one of them. AI search is not just a simple upgrade of traditional search engines, but a way of reshaping information acquisition through technologies such as deep learning and natural language processing. Aravind Srinivas, the founder of Puzzle AI, describes it as an "answer engine" - when users ask questions, it will give a concise and clear answer, and all answers will be marked with the source. , and gives a series of association questions to form further interactions with users and further expand and extend related topics.
Fu Xiaolan, director of the Technology and Management Development Research Center of the University of Oxford and an academician of the British Academy of Social Sciences, said in an interview with reporters that AI search has reshaped the traditional keyword matching model through semantic understanding and deep learning technology. This major change in the way we obtain information will allow AI search to play an important role in medical, legal, financial, start-up valuation and other fields.
Zhou Xiaopeng, vice president of Alibaba Intelligent Information Business Group, said in an interview that the core value of search engines is to connect users with the information they seek. AI search has reshaped the way of obtaining information, bringing the distance between users and information to zero. It is the core foundation for AI search to have broad prospects in the future.
In addition, the characteristics of AI search technology determine that it can handle more complex queries, not limited to text, but also understand and index diverse content such as videos, pictures, and voices. This multi-modal processing capability enables AI search to extract information from a wider range of data sources and provide users with richer and more accurate search results. At the same time, AI has also reshaped the product form of search, enabling AI writing, AI document summarization, etc., turning search from a tool into an all-round AI assistant.
Traditional search introduces agents
Facing the wave of AI search, traditional search engines are also upgrading to "AI" by introducing agents.
Rong Wei, head of Baidu search brand operations, said in an interview recently that agents are a widely optimistic AI application direction. According to the common definition in the industry, an intelligent agent is a computer program that can interact with the environment, collect data, and use the data to make independent decisions to complete preset tasks. Its greatest feature is autonomy—human beings only need to give it a final goal, and it will It can dismantle targets, obtain relevant data resources, and feedback results by itself, so it can complete more complex and ambiguous tasks, thus greatly improving the practical value of generative AI.
To a large extent, intelligent agents can be regarded as extensions and upgrades of search engines. Before the birth of generative AI, in addition to referring to information on websites and communities, people mainly found answers manually through search engines - trying to search for a large number of information sources using different keywords. In the era of generative AI, search engines have become the role of “intelligent central connectors”.
Regarding the information acquisition scenario, Zhou Xiaopeng said that search is the entry point for users to initiate demands. When faced with complex problems, AI search can disassemble and reason about user intentions, retrieve, analyze and summarize information across the entire network, and generate and aggregate multi-modal answers. Search results are not the purpose, solving problems is the core. Therefore, AI search can provide richer content forms and interaction methods, and to a certain extent, evolve into an AI assistant in various user scenarios.
Privacy protection becomes the focus of future development
Although AI search engines show broad market prospects, their development still faces many challenges. Data privacy and security issues are the focus of users' attention. How to protect user privacy while providing high-quality services will be an important issue that AI search engines need to solve.
Fu Xiaolan believes that privacy protection, technical compliance and ethical issues have put forward higher requirements for the development of AI search engines. In the future, with the continuous advancement of technology and changes in market demand, AI search engines are expected to achieve breakthroughs in more fields and bring users a more intelligent and convenient search experience.
Zhou Xiaopeng said that security and compliance are the cornerstones for every product to seek greater development. AI products will definitely encounter many new problems in the development process, such as strengthening technical support, clarifying privacy policies, limiting the scope of data use, and data compliance. Use is the responsibility of every practitioner. He said: "In this process, multiple roles including industry, academia, institutions, and government need to continuously communicate to continuously narrow the cognitive differences and reach consensus."