Elon Musk's xAI company has announced the official launch of the voice function for its AI chatbot Grok3. This highly anticipated feature is now available to users through the Grok app, currently supporting English interactions. According to the latest news on the X platform, users simply need to update the Grok app to find the new voice icon next to the 'New Conversation' button in the upper right corner of the interface, where they can click to experience this new feature. xAI revealed that Grok3's voice function offers up to 10 different modes, including two marked as

On the first day of DeepSeek Open Source Week, we officially released our latest technological achievement, FlashMLA, an efficient Multi-Layer Attention decoding kernel specifically designed for NVIDIA Hopper architecture GPUs. This technology is optimized for variable-length sequence scenarios and can significantly enhance the inference performance of large models. The core technical features of FlashMLA include comprehensive support for BF16 precision and a paged key-value cache with a block size of 64.

On February 24, LiblibAI announced it has completed four rounds of financing within just one year, setting a new record for the speed of financing in the domestic AI application sector. Following hundreds of millions in financing disclosed in July 2024, LiblibAI has successfully secured hundreds of millions again. The latest two rounds of financing were led by Yongce Capital and Shunwei Capital, with existing shareholders like Ming Shi Venture Capital also participating heavily.

Recently, the Meta AI team launched the video joint embedding prediction architecture (V-JEPA) model, an innovative initiative aimed at advancing machine intelligence. Humans can naturally process information from visual signals and recognize surrounding objects and motion patterns. An important goal of machine learning is to reveal the fundamental principles that drive unsupervised learning in humans. Researchers proposed a key hypothesis—the predictive feature principle—arguing that the representations of continuous sensory inputs should be able to predict each other. Early research methods utilized slow feature analysis.

Recently, Tencent's Hunyuan Video Model has officially begun recruiting testing partners on Platform X, marking a critical testing stage for this cutting-edge AI video generation technology. According to official sources, there is a high probability that the model will be open-sourced after the testing concludes, contributing its technological achievements to the global AI community. The Hunyuan Video Model is an important innovation by Tencent in the field of AI video generation, boasting over 13B parameters, making it one of the largest video generation models among open-source models.

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.

According to multiple media reports, the National Institute of Standards and Technology (NIST) plans to lay off up to 500 employees, a decision that further threatens the newly established U.S. Artificial Intelligence Security Institute (AISI). Axios reported that AISI and the 'U.S. Chip Plan' will face 'significant' layoffs targeting probationary employees, who are typically new hires with one to two years of service. Bloomberg noted that some employees have already received verbal notifications of impending layoffs.

In a shocking statement, a senior defense official revealed to Defense One that the Pentagon's interest in autonomous killer robots is accelerating. The official, speaking on condition of anonymity, stated that the Pentagon plans to reduce funding for artificial intelligence research in favor of actual AI weapon systems. He noted, 'We will no longer invest in 'artificial intelligence' because I don't know what that means. We will invest in autonomous killer robots.' The official added that the current government is more focused on weapon systems.

Recently, OpenAI researchers admitted in a newly published paper that despite the advanced nature of current AI technologies, these models still cannot compete with human programmers. OpenAI CEO Sam Altman previously stated that by the end of this year, AI is expected to surpass 'junior' software engineers, but research findings show that these AI models still face significant challenges. Image credit: Image generated by AI, image licensed from service provider Midjourney in the study.

Users can now activate DeepSeek-R1 by saying "Open Deep Thinking" to Super Xiao Ai, enjoying its capabilities such as coding, writing articles, and logical reasoning. Although Xiaomi has not made an official announcement on this matter, the coverage of Super Xiao Ai is already clear, encompassing over 25 Xiaomi and Redmi models.

On February 20th, Elon Musk announced an important update on the social platform X, stating that the Grok application has successfully launched the early testing version of Grok 3 voice mode. This news immediately attracted widespread attention.

Today, there is new news from the free APP download rankings of the Apple App Store. Tencent Yuanbao has successfully surpassed Doubao, rising to the second position on the China regional chart, second only to the consistently top-ranked DeepSeek. This achievement is attributed to a major update that Tencent Yuanbao underwent recently. On February 13th, Tencent Yuanbao officially integrated the full-version Deepseek - R1 model, which provides users with unprecedented convenience upon opening Tencent Yuanbao and entering the chat interface.