According to reports from Titanium Media, Xiaohongshu is set to integrate the DeepSeek-R1 open-source model, and its AI search product 'DianDian' App will launch a 'Deep Thinking' feature. This new feature is currently in beta testing. With the rise of ChatGPT, Xiaohongshu is taking a deeper part in the development of generative AI, introducing an independent AI search application 'DianDian'. 'DianDian' is positioned as a life search assistant, covering various scenarios such as transportation, food, and travel, aiming to provide users with timely information and genuine feedback.

Today, Weibo is testing the integration of its Smart Search function with the DeepSeek large model. Weibo CEO Wang Gaofei has showcased the testing effects of related features multiple times on the platform, including intelligent analyses of hot topics such as the impact of increased social security fees on rider incomes. The tests demonstrate that the DeepSeek R1 model can effectively integrate existing discussion content on the Weibo platform, providing users with clear topic summaries and analytical conclusions. The introduction of this feature will not only enhance the quality of Weibo's search and user experience but also indicates the platform's move toward a more comprehensive intelligent service.

DingTalk AI Corporate Search, launched in 2024, focuses on addressing the pain points of enterprise users related to information fragmentation and low retrieval efficiency. This tool leverages the understanding, reasoning, and generating capabilities of large models to integrate unstructured data such as chat logs, documents, knowledge bases, schedules, and logs into a structured knowledge network, improving search efficiency by 300% over traditional methods. By offering this free access, DingTalk aims to further assist enterprise users in building dynamic knowledge networks and optimizing workflows.

Recently, the Evo2 bio AI model, jointly developed by Arc Institute and Nvidia, was officially released. This foundational model is based on DNA data from over 100,000 species and aims to deeply decode various complex phenomena in biology. Evo2 can identify patterns in the genetic sequences of different organisms that researchers would typically take years to discover, significantly enhancing the ability to identify disease-related mutations, and can design entirely new genomes comparable to simple bacteria.

Google has officially announced the launch of its AI research assistant system based on Gemini 2.0 - the AI co-scientist, aimed at providing virtual collaboration for scientists to generate novel research hypotheses and suggestions. This system not only offers traditional literature review and summarization capabilities but also breakthrough original knowledge discovery abilities. The AI co-scientist can simulate the reasoning process of scientific methods and provide researchers with innovative research hypotheses and proposal plans based on existing evidence and specific research objectives.

Recently, the Arc Institute partnered with NVIDIA, along with researchers from Stanford University, the University of California, Berkeley, and the University of California, San Francisco, to jointly release the world's largest biological artificial intelligence model — Evo2. This model is based on data from over 128,000 genomes, training on 93 trillion nucleotides, making its scale comparable to the most powerful generative AI language models. Evo2's deep learning capabilities enable it to rapidly identify sequences of genes across different organisms.

Recently, Aomni announced the completion of a $4 million seed round funding led by Decibel, with participation from Sancus Ventures and Ride Home Fund. Aomni is dedicated to assisting sales teams in conducting in-depth customer research using a human-centered approach aimed at improving sales efficiency rather than replacing sales personnel. Image source note: The image is AI-generated and licensed by Midjourney. Aomni's founder, David Z.

Recently, a study led by Meta's Chief AI Scientist Yann LeCun revealed how artificial intelligence can develop a basic understanding of physics by watching videos. This research, conducted by scientists from Meta FAIR, Paris University, and EHESS, shows that AI systems can acquire intuitive physical knowledge through self-supervised learning without predefined rules. The research team employed a new method called Video Joint Embedding Predictive Architecture (V-JEPA).

Today, well-known popular science blogger 'Three Dads Laboratory' announced via Weibo that Tencent's Yuanbao app has integrated DeepSeek technology and embedded advertising content in its answers. Screenshots shared by the blogger show that in the suggestions and reminders provided by the Yuanbao app to avoid pitfalls, users are explicitly recommended to compare prices and make appointments through legitimate platforms like 58 Daojia, as well as to inquire prices through Kuaihuishou or local legitimate recyclers. These suggestions have been criticized for potentially containing embedded advertisements.

Recently, Tencent launched an AI search function on its social platform WeChat, which became unusable due to server overload less than a day into its grayscale testing. According to Tencent, the feature utilizes the Hunyuan large model for searches and integrates the DeepSeek-R1 model to provide deeper thinking services. However, despite the very small user count participating in the testing, the call volume was extraordinarily high, resulting in frequent prompts stating 'Service is busy, please try again later.' In this situation, Tencent had to quickly take action.

Recently, Meeta Technology announced that its AI search platform has launched a new research model - 'Think First, Search Later'. The core of this innovative model is to allow the AI model to first propose a framework and pathway for thinking, thereby laying the foundation for subsequent information integration and analysis. This approach enables users to obtain the information they need more efficiently, significantly enhancing research productivity. It is understood that the 'Think First, Search Later' model adopts a collaborative architecture of 'small model + large model'. Specifically, the framework for deep reasoning and the task of breaking down steps are entrusted to...

A recent study from Microsoft and Carnegie Mellon University indicates that excessive reliance on artificial intelligence (AI) tools may be eroding people's critical thinking abilities. The research team surveyed 319 knowledge workers and collected 936 real-world examples of generative AI usage across various industries, including IT, design, administration, and finance. The study analyzed six dimensions of critical thinking: knowledge, understanding, application, analysis, synthesis, and evaluation. The findings reveal that people's thinking patterns in problem-solving change significantly after using AI tools.

On February 18th, Shanghai Mita Network Technology Co., Ltd. announced the launch of its latest research mode - 'Shallow Research'. This feature aims to quickly gather information and conduct preliminary analysis using AI technology, helping users obtain valuable research results in a short amount of time.

On February 17, Baidu announced the official launch of the Ernie Bot free rights card, which will be fully open for free use starting April 1. This news marks another significant move by Baidu in the field of artificial intelligence, aimed at further lowering the barrier for users to access AI technology and promoting the popularization and application of artificial intelligence.

On February 16, Tencent's QQ Browser integrated the full version of the DeepSeek-R1 model, supporting deep thinking, online search, multi-turn dialogue, and historical record retrieval. Users can search for 'DeepSeek' in the QQ Browser app or PC, click on the dedicated version of the DeepSeek model, and log in to use it for free. The app also supports searching, translating, note-taking, and exporting content to PDF or Word within the same interface, allowing users to ask questions, process information, and store it.