AI-native maps are about to debut, and Baidu Maps will comprehensively apply large model technology.

Baidu Maps is about to launch a brand new AI-native map product, which will be a complete reconstruction of the way maps are interacted with. The new version of the map will fully apply Wenxin large model technology, bringing users a brand new experience.

Through natural language dialogue, users can easily activate various functions and services in the map, greatly improving travel efficiency and decision-making efficiency. This interaction method is more intuitive and convenient, and is expected to change people's habits of using maps.

It is reported that the reconstructed Baidu Map V19 version will launch a new "AI Guide" feature. This feature supports multi-turn natural language interaction, making the experience closer to a conversation with a real person. In addition to the system's default avatar, users can also upload photos to create a personalized digital avatar, further enhancing the personalized experience.

In addition, Baidu Maps plans to cover lane-level navigation features in over 100 cities within the year. This will provide drivers with more precise route guidance, effectively enhancing driving safety and convenience.

The deep integration of AI technology and maps signifies that map products are entering a new stage of intelligence. In the future, maps will not just be a navigation tool, but will also become a travel assistant that integrates various intelligent services. This transformation is expected to provide users with a more convenient, intelligent, and personalized travel experience.

Overview of Artificial Intelligence Industry Trends

Domestic AI Development Continues to Advance

ZTE Corporation has launched the industry's first "Smart Governance" SMS anti-fraud system based on large models, applying large model technology to the anti-fraud field. This system improves the identification and interception of fraudulent SMS by constructing an anti-fraud large model and utilizing generalized feature neural networks to govern from dimensions such as the intent and semantics of fraudulent messages.

Tencent's QQ Browser has launched a PDF reading assistant powered by the Hunyuan large model, featuring functions such as intelligent summarization and Q&A. This function is currently open for testing applications.

The ZhiYuan Research Institute announced a comprehensive upgrade of the Wudao ・ Tianying Aquila large language model series to Aquila2, with the addition of the Aquila2-34B model featuring 34 billion parameters. At the same time, it released an open-source package that includes semantic vector models, training frameworks, and more.

NetEase Youdao has launched Hi Echo, a virtual spoken language tutor based on the Ziyue Education large model, providing users with round-the-clock speaking practice services.

International AI Company Dynamics

OpenAI plans to launch a major upgrade for developers next month, aimed at reducing the cost and difficulty of building applications based on its models. The company is trying to attract more businesses to use its technology.

DeepMind's 2022 financial report shows that employee costs and profits have both decreased. To improve efficiency, Alphabet merged Google's AI department with DeepMind's in April this year.

EY and IBM have partnered to launch the AI solution EY.ai Workforce, which helps businesses apply AI technology to human resource processes and improve efficiency.

AI Education Applications

Australia has announced that starting in 2024, it will allow the use of generative AI technologies, including ChatGPT, in all schools. The country will release a relevant framework to guide schools in the safe and reasonable use of AI, and it has committed to reviewing regulatory measures every 12 months.

AI Startup Financing

AI startup Anysphere has secured an $8 million investment from OpenAI, planning to build an AI-driven integrated development environment. Former GitHub CEO and other investors also participated in this funding round.