AI
Nvidia to supply 18,000 advanced AI chips to Saudi Arabia
Chip manufacturer NVIDIA recently announced that it will cooperate with Humain, an artificial intelligence (AI) startup under the sovereign wealth fund of Saudi Arabia, to supply 18,000 advanced AI chips to the country. This announcement was made during a visit by a U.S. delegation to Saudi Arabia, Qatar, and the UAE, marking a significant step for Saudi Arabia in AI and cloud computing infrastructure.Jensen Huang, founder of NVIDIA, stated at the Saudi-US Investment Forum that "AI is like electricity and the internet, it is basic infrastructure for every nation." He emphasized that the cooperation with Humain aims to build AI infrastructure for the people and enterprises of Saudi Arabia to achieve the kingdom's grand vision..
Granola, an AI note-taking app, raises $43 million aiming for a knowledge hub for individuals and businesses
The artificial intelligence meeting recording application, Granola, has been experiencing rapid growth recently. Since its launch a year ago, this startup product has achieved a weekly user growth rate of 10%. The rise of Granola is not only due to the word-of-mouth spread between the founding team and investors but also because of its core selling point that goes beyond "automatic meeting recording." It has gradually become an "AI assistant" for users to integrate personal and work information.Granola Co-founder Chris Pedregal told TechCrunch that users are now more often using Granola for personal notes, so they can obtain insights through its AI analysis of daily content.
Ke Ling 2.0 Model Achieves Outstanding Results, Continues to Lead the Global Video Generation Model Rankings!
In the recently released global video generation large model ranking, Ke Ling AI 2.0, under Kuaishou, achieved an Arena ELO score of 1124, surpassing its predecessor version 1.6 and successfully securing the top global position. This news has garnered significant attention, especially in the context of the growing prominence of AI technology. Ke Ling AI 2.0 has consistently topped the field of image-to-video transformation for three consecutive months, showcasing its strong technical prowess and market potential..
Released PixVerse V4.5 Video Model! Over 20 Movie Lenses + Multi-Image Fusion, Create a Hollywood Blockbuster in 5 Seconds!
PixVerse officially released its V4.5 video model, introducing over 20 new features in cinematic lens control, multi-image reference functionality, and enhanced handling of complex actions (). This update significantly improves the quality and creative freedom of video generation, solidifying PixVerse's leading position in the field of AI video creation. AIbase observed that the release of V4.5 quickly sparked discussions among creators worldwide, being hailed as a "milestone in movie-level AI video creation."Core Features: Professional Lens Control and Multi-Image Fusion.
Meta Collaborates with Georgia Tech to Launch CATransformers Framework to Reduce AI Carbon Footprint
Recently, Meta's FAIR team collaborated with the Georgia Institute of Technology to develop the CATransformers framework, which aims to make carbon emissions a core consideration in the design of AI systems. This new framework significantly reduces the total carbon footprint of AI technologies by jointly optimizing model architectures and hardware performance, marking an important step toward sustainable AI development.With the rapid popularization of machine learning technology, applications in areas such as recommendation systems and autonomous driving are increasing, but their environmental costs cannot be ignored. Many AI systems require substantial computational resources and often rely on custom hardware accelerators for operations.
The Next Generation Open Source 3D Model Step1X-3D Debuts, AI Industry Trend Draws Attention
Recently, the technology sector welcomed a brand-new open-source 3D large model called "Step1X-3D." The release of this model marks another significant advancement in AI technology, particularly in 3D modeling and reasoning capabilities. Not only is this model open-source, but it also provides developers with various practical features, greatly promoting innovation and research possibilities.At the same time, Xiaomi is continuously expanding its presence in the AI field. It has recently applied for the "MiMo" trademark, which is intended to be used for inference large models.
AI+数据智能体的三大支点：数据治理、知识库和大模型
当销售部喊出"业绩增长15%"，财务部却坚称"只有8%"。 会议室里争论不休，时间流逝，竞争对手已经抢占先机。 你不禁自问：明明砸了千万建设数据系统，为何企业依然深陷数据内耗？
速度最快：Stable Audio Open Small 端侧音频模型登场，手机上 8 秒内 AI 生成 11 秒音频
AI 初创公司 Stability AI 推出 Stable Audio Open Small，号称是市场上速度最快的“立体声”音频生成 AI 模型，可在智能手机上运行。
重磅！谷歌DeepMind发布AlphaEvolve：AI界的“算法设计进化大师”诞生
谷歌DeepMind刚刚又往前拱了一大步，宣布推出 AlphaEvolve智能体 ，目标直指更上游，用于通用算法的设计发现和优化简单说，AlphaEvolve就像个AI界的“算法育种大师”。 它把自家Gemini大模型（Gemini Flash负责广撒网，洞察力强的Gemini Pro负责深挖）和一套“自动化考官”（负责验证算法靠不靠谱、效率高不高）结合起来，再套上一个“进化论”的框架，让好算法能一代更比一代强AlphaEvolve工作流程：工程师设定框架，AI通过“提示采样器”给LLM喂招，LLM出新招（程序），“考官”打分，好招进“兵器谱”，并用来启发下一轮出招。 去年DeepMind就秀过肌肉，证明LLM能生成代码函数来搞定科学问题。
GPT-4o不敌Qwen，无一模型及格！UC伯克利/港大等联合团队提出多模态新基准：考察多视图理解能力
多视图理解推理有新的评判标准了！ 什么是多视图理解？ 也就是从不同视角整合视觉信息进而实现理解决策。
破解300年数学难题，智能体大突破！谷歌发布超强AI Agent
今天凌晨，谷歌Deepmind在官网发布了，用于设计高级算法的编程AI Agent——AlphaEvolve。 AlphaEvolve与谷歌的大模型Gemini实现深度集成，用于自动评估通用算法的发现与优化，可以帮助开发人员快速设计出最好、高效的矩阵算法。 简单来说，大模型擅长生成各种想法和算法，但是没人知道这些到底行不行，而AlphaEvolve相当于“质检员”，能够按照特定标准来衡量这些想法是否可行。
ICML25 | 让耳朵「看见」方向！仅依靠360°全景视频，就能生成3D空间音频
空间音频，作为一种能够模拟真实听觉环境的技术，正逐渐成为提升沉浸式体验的关键。 然而，现有的技术大多基于固定的视角视频，缺乏对360°全景视频中空间信息的充分利用。 在这样的背景下，一项在空间音频生成领域具有里程碑意义的研究应运而生——OmniAudio：它能够直接从360°视频生成空间音频，为虚拟现实和沉浸式娱乐带来了全新的可能性。
DanceGRPO：首个统一视觉生成的强化学习框架
本文由字节跳动 Seed 和香港大学联合完成。 第一作者薛泽岳为香港大学 MMLab@HKU 在读博士生，在 CVPR、NeurIPS 等国际顶级会议上发表多篇研究成果。 项目通讯作者为黄伟林博士和罗平教授。
25岁MIT辍学天才一战成名！3年成为90亿美金公司CEO
硅谷又出现了一位新的天才。 AI浪潮中，一位年仅25岁的远见者正以惊人的速度改写着软件开发的未来，他就是Michael Truell，AI代码编辑器Cursor背后的母公司Anysphere的CEO。 Cursor仅仅用了12个月，ARR就达到了一亿美元，多篇业内分析认定Cursor是 「SaaS史上最快到$100M ARR的初创公司」。
Meta 推出 CATransformers 框架 助力AI行业实现减排目标
在人工智能迅猛发展的今天，Meta 的 FAIR 团队与佐治亚理工学院联合研发了一款名为 CATransformers 的全新框架。 该框架以降低碳排放为核心设计理念，旨在通过优化模型架构与硬件性能，显著减少 AI 技术在运营中的碳足迹，为可持续的 AI 发展奠定基础。 随着机器学习技术在各个领域的广泛应用，从推荐系统到自动驾驶，其背后的计算需求不断增加。
陶哲轩携AI再战数学！o4-mini秒怂弃赛，Claude 20分钟通关
3天后，陶哲轩YouTube视频二更来了。 上一次，他使用GitHub Copilot（基于GPT-4），成功在33分钟内完成一页非形式化证明。 这次，他尝试了一种更短、更概念化的证明版本，并测试Claude、o4-mini能否基于之前的非形式和形式证明，生成类似的形式化代码。
所有AI工具共享记忆！MCP协议杀疯了：100%本地运行，Cursor、Claude都能用
聊完就忘？ 当下多数AI助手和开发工具各自独立运行，会话结束上下文即消失，严重影响了使用体验和效率。 OpenMemory MCP，一款可以解决AI工具记忆痛点，并且实现不同工具之间共享上下文信息的开源工具，他来了！
看图猜位置不输o3！字节发布Seed1.5-VL多模态推理模型，在60个主流基准测试中拿下38项第一
在60个主流基准测试中拿下38项第一！ 字节发布轻量级多模态推理模型Seed1.5-VL，仅用532M视觉编码器 200亿活跃参数就能与一众规模更大的顶尖模型掰手腕，还是能带图深度思考的那种。 相关技术报告也第一时间公开了。
资讯热榜
标签云
AI
人工智能
OpenAI
AIGC
模型
ChatGPT
谷歌
DeepSeek
AI绘画
大模型
AI新词
机器人
数据
Midjourney
开源
Meta
微软
智能
用户
GPT
学习
智能体
技术
Gemini
英伟达
马斯克
Anthropic
图像
AI创作
训练
LLM
论文
代码
AI for Science
苹果
算法
腾讯
Agent
Claude
芯片
Stable Diffusion
具身智能
xAI
蛋白质
开发者
人形机器人
生成式
神经网络
机器学习
AI视频
3D
RAG
大语言模型
字节跳动
Sora
百度
研究
GPU
生成
工具
华为
AGI
计算
AI设计
大型语言模型
生成式AI
搜索
视频生成
亚马逊
AI模型
特斯拉
DeepMind
场景
深度学习
Transformer
Copilot
架构
MCP
编程
视觉