Search AI Products and News

Google DeepMind recently published a blog post, officially introducing the new Gemini Robotics On-Device localized robot AI model. The model uses a vision-language-action (VLA) architecture, enabling precise control of physical robots without cloud support. The biggest highlight of the new model is that it runs entirely locally on the robot device, achieving low-latency response capabilities. This feature makes it particularly suitable for environments with unstable network connections, such as critical application scenarios like medical facilities.

Recently, the DeepMind team of Google collaborated with Brown University to develop a new technology called "force prompting". This technology can generate realistic motion effects without 3D models and physics engines, marking a significant breakthrough in the field of AI video generation. With this technology, users only need to specify the direction and intensity of the force to manipulate the content generated by AI videos. The force prompting technology can be applied to global forces (e.g., overall wind) and local forces.

A new study has raised concerns that OpenAI may have used copyrighted O'Reilly Media books to train its latest AI models without permission. The study was released by the AI Disclosures Project, a non-profit organization founded in 2024 by media mogul Tim O'Reilly and economist Ilan Strauss. AI models can be viewed as sophisticated prediction engines that...

On March 27, 2025, Microsoft co-founder Bill Gates, in a lecture at the New York Economic Club, shared the creative process behind his autobiography, "The Source Code," and delved into the development of artificial intelligence (AI) and its impact on global society. Regarding the timing of his memoir, Gates stated: "While I usually focus more on the future, with my 70th birthday approaching, Microsoft's 50th anniversary, and the Gates Foundation's 25th anniversary, I felt it was time to look back and share some important reflections." He specifically mentioned that this book...

In the field of deep learning, normalization layers are considered an indispensable component of modern neural networks. Recently, research led by Meta FAIR research scientist Zhuang Liu, titled "Transformer without Normalization Layers", has garnered significant attention. This research not only introduces a new technique called Dynamic Tanh (DyT), but also demonstrates the effectiveness of Transformer architectures even without traditional normalization layers.

Researchers from the Tübingen University's ELLIS Unit, the University of Maryland, and Lawrence Livermore National Laboratory have developed Huginn, a novel language model with a recursive architecture that significantly enhances reasoning capabilities. Unlike traditional models, Huginn doesn't require specialized 'reasoning chain' training; it can autonomously reason within the neural network's 'latent space' before outputting results. The research team developed Huginn, a novel language model with a recursive architecture that significantly enhances reasoning abilities. Unlike...

In the field of artificial intelligence (AI), while large language models (LLMs) excel at processing natural language, they often struggle with complex reasoning tasks. These tasks typically require multi-step reasoning, domain-specific knowledge, or effective integration of external tools. To overcome these limitations, researchers have been exploring how to enhance LLM capabilities through the use of external tools. Traditional enhancement methods usually necessitate fine-tuning or additional training of the models, which limits their adaptability and flexibility for tasks.

Early this morning, OpenAI announced an important update regarding the ChatGPT search feature. Users can now directly access the ChatGPT search engine without logging into an account. This change not only broadens the accessibility of ChatGPT but also positions it in direct competition with traditional search engines like Google and Bing.

Recently, Shanghai Jumpsky Star Intelligent Technology Co., Ltd.'s Jumpsky App launched a new feature called 'AI Creative Board', providing users with an innovative experience to develop applications without coding skills. In the digital era, many creative ideas struggle to materialize due to a lack of programming skills. The 'AI Creative Board' feature aims to break this barrier, lowering the threshold for application development to its minimum. Users simply need to upgrade the Jumpsky App to the latest version, click on the 'Creative Board' at the bottom left, and use simple text descriptions (Prompt

Recently, a startup named Astral has attracted widespread attention for its proposal to use AI technology to flood Reddit with advertisements. Founder Savannah Feder demonstrated the rapid application abilities of this technology, which can quickly generate content similar to user discussions, quietly promoting clients' products. In a demonstration, Feder used AI-generated text to respond to a post about improving customer feedback capabilities, making the content appear authentic.

Recently, a Xiaohongshu user named 'Summer Little Camisole' posted a video interacting with Moflin, which instantly attracted a large number of viewers. The comment section turned into a live showcase for 'cloud petting', with many users expressing that 'Moflin is so cute' and eagerly inquiring about its price. So far, this user has released three videos interacting with Moflin, receiving significantly more likes and comments than her other content. Moflin officially went on sale in Japan on November 7 this year, priced at 2832 RMB, and was sold out almost immediately upon release. Its appearance is similar to