AI改变搜索引擎模式超过60%的用户开始使用AI助手查询产品和服务推荐，而不是传统搜索引擎。当用户问“最好的项目管理工具是什么？”时，你的品牌是否出现在回答中？用户提问哪个品牌的智能手表更适合运动？如果AI没有推荐你的品牌，你就失去了这个潜在客户

AI NEWSLatest AI NewsArticleByteDance Seed Adjusts: Comprehensive Restructuring of Pre-training, Reinforcement Learning, and Agent DevelopmentPublished in Latest AI NewsTime :Aug 21, 2026Read :3minuteAccording to media reports, the Seed department of ByteDance’s large model division completed a new round of organizational restructuring last week. Among them, the Seed Foundation Model has newly established four first-level departments, with all department heads reporting to Wu Yonghui. This adjustment further integrates pre-training data, reinforcement learning, and post-training resources, and separates the model capabilities of B-end and C-end products.Pretrain Data is led by Li Chenggang, integrating the previously scattered teams for pre-training data in text models, programming, visual understanding, and speech, and will be responsible for the multimodal data and ultra-large model pre-training data for the new Omni model. Horizon RL is led by Tang Shengyu, integrating the post-training, reasoning, and visual understanding teams, focusing on advancing reinforcement learning to enhance the basic intelligence ceiling of the model.Product Posttrain-Work is led by Qin Yujia, mainly targeting B-end applications, responsible for the integration and release of Agentic models, and optimizing Agentic capabilities for office scenarios, providing support for the task modes of products such as Douba and Dola. The original Application team will continue to be retained and renamed Product Posttrain-Chat, led by Zhu Wenjia, mainly targeting C-end applications, responsible for the integration and release of dialogue models, forming a division of labor with the Work team.This adjustment shows that Seed is further strengthening the model development pipeline from data, pre-training, to reinforcement learning and product post-training, and establishing a differentiated model system for enterprise office agents and consumer dialogue scenarios. ByteDance has not yet responded so far.

AI NEWSLatest AI NewsArticleAnthropic’s Star Feature Launched! The Computer Operation, Skills, and Files APIs Are Fully AvailablePublished in Latest AI NewsTime :Aug 21, 2026Read :4minuteAnthropic has officially announced that its highly anticipated Computer use, Skills API, and Files API on the Claude Platform are now fully available (Generally Available). This update not only introduces a brand-new Browser use tool but also deeply optimizes underlying capabilities, aiming to empower enterprises to easily build production-grade agents that can truly operate software, apply team expertise, and deliver finished products.Regarding core technology and capability upgrades, the computer use function has achieved significant breakthroughs. The updated tool allows Claude to perform multiple consecutive actions in a single model call, greatly reducing task completion time and the number of calls. Additionally, the computer use tool now meets the standards of the Health Insurance Portability and Accountability Act (HIPAA) and can be used for regulated healthcare workloads under a Business Associate Agreement (BAA).As the absolute highlight of this release, the new Browser use tool is officially launched. This tool inherits the advantages of multi-action rounds and innovatively introduces web structure parsing capabilities. Unlike traditional screenshot-based clicking, it can directly identify and interact with specific input fields or buttons on web pages, enabling more accurate and efficient automation within web applications.In terms of knowledge accumulation and document management ecosystem, the Skills API and Files API provide strong underlying support. The Skills API allows enterprises to upload and manage their own proprietary skill folders, which include industry-standard instructions, scripts, and templates, and run securely within Claude’s code execution sandbox, eliminating the need for enterprises to host them themselves. The Files API provides a stable and efficient storage space for agents to handle documents, supporting the upload of files such as PDFs or spreadsheets and referencing them directly by ID in subsequent requests. It also implements an automatic file expiration mechanism, with a fivefold increase in rate limits, and offers up to 1TB of dedicated storage space per organization.With the full implementation of these major features, developers can continue to use the Skills API and Files API through Microsoft Foundry. The updated computer use and browser use tools will soon be available on Google Cloud’s Vertex AI. Enterprises and developers across industries are accelerating their efforts to leverage this powerful toolchain to comprehensively revolutionize the level of automation in complex business processes.

AI NEWSLatest AI NewsArticleOpenAI Fully Open Sources Codex Harness AI Programming Intelligent Agent, Further Opening Up Its Core CapabilitiesPublished in Latest AI NewsTime :Aug 21, 2026Read :3minuteOpenAI has further opened up the core capabilities of Codex Harness, making the intelligent agent operating framework for different product forms of Codex available to developers. Codex Harness is the core agent runtime environment behind Codex, responsible for connecting the model, users, and tools, and managing processes such as task execution, context, sessions, and code operations. OpenAI previously open-sourced the Codex CLI and related core code, and continues to improve the capabilities of Harness and App Server.From an architectural perspective, Codex Harness is not merely a code generation tool, but a complete agent execution framework. Its core includes Agent Loop, thread lifecycle and persistence, configuration and authentication, tool execution, and extension mechanisms. Through Codex App Server, developers can integrate Harness with different clients using the bidirectional JSON-RPC protocol and obtain real-time event streams, thereby building IDEs, desktop applications, and other customized agent products.Currently, Codex Harness supports various usage scenarios, including Codex Web, CLI, IDE extensions, and macOS applications. OpenAI previously also publicly shared the engineering practice of “code entirely written by Codex” in Harness: an internal product generated about 1 million lines of code within about 5 months and completed approximately 1,500 Pull Requests, with the team’s core philosophy shifting from “manually writing code” to “human setting goals, agent execution.“This further opening of Harness indicates that the competition in AI programming is extending from model capabilities to agent runtime, tool invocation, context management, and engineering feedback loops. As the underlying Harness gradually becomes standardized, developers will be able to build more rich professional AI programming and software engineering workflows on top of Codex’s core agent capabilities.

AI NEWSLatest AI NewsArticlePICO Space Pro Headset Scheduled for Release on September 2nd, Featuring Dual-Chip Architecture and a 4000 PPI Micro-OLED DisplayPublished in Latest AI NewsTime :Aug 19, 2026Read :4minutePICO-XR’s official Weibo account recently officially announced that the highly anticipated new head-mounted display, PICO Space Pro, will be officially launched on September 2nd at 14:30. The official has set the theme of this product launch as “Start a New Spatial Experience.“The upcoming Space Pro headset is expected to be the mysterious device codenamed “Project Swan” that was accidentally exposed in the public SDK. In terms of design, the headset cleverly integrates the design essence of several industry benchmark products: its body uses a single woven headband design similar to Apple’s Vision Pro, and the back of the headband has a particularly obvious style; while the front light-blocking pad and forehead cushion part show an造型 style similar to Samsung Galaxy XR.In addition, the device is equipped with an independent external module, whose design style is very similar to the battery unit of Apple’s Vision Pro. However, the official has not yet clearly stated the specific purpose of this module. It could be a simple external battery pack, or it may integrate some core computing units. In terms of interactive accessories, the headset is equipped with controllers with tracking functions, whose appearance design is highly consistent with the controllers of Pico4Ultra and supports a “drawing boundary area” function similar to Meta Quest. However, the official has not yet confirmed whether these controllers will be included as standard equipment in the package.Regarding the core hardware specifications, the PICO Space Pro demonstrates strong technical strength. The headset will be equipped with high-precision micro-OLED display panels, with a pixel density as high as 4000 PPI, providing users with extremely delicate visual presentation. At the same time, it adopts an advanced dual-chip architecture: including a self-developed co-processor dedicated to computer vision and image processing, as well as a powerful new main processor. It is reported that the CPU and GPU performance of this main processor will reach twice that of the current Qualcomm Snapdragon XR2 Gen2 chip, which is widely used in mainstream devices such as Pico4Ultra and Meta Quest3. As the release date approaches, more details about this new generation of spatial computing devices will be fully revealed at the launch event.

AI NEWSLatest AI NewsArticleSouth Korea Enacts Law to Facilitate AI: Use of Limited Personal Data After Regulatory ReviewPublished in Latest AI NewsTime :Aug 24, 2026Read :2minuteThe Korean Personal Information Protection Committee recently stated that the National Assembly has unanimously passed an amendment to the Personal Information Protection Act, allowing AI developers to use limited personal data after review by the committee. This breaks the rigid constraints of the current law that “personal information can only be used for the explicitly specified purpose at the time of collection, and consent from the individual is required for using it for AI development.“The Korean AI industry has long complained that the original law restricted technology development based on original personal data. Previous exemptions applied only to scenarios such as voice anti-fraud and autonomous robots, with a term of two years, extendable up to four years.The new amendment will take effect six months after approval by the State Council. The committee plans to solicit opinions from experts and the industry to develop specific guidelines and subordinate laws. While the new regulations ease restrictions on local AI training, they also incorporate “review and control” into the process, meaning the boundaries of personal data usage are still determined by regulation rather than by companies themselves.

AI NEWSLatest AI NewsArticleByteDance Integrates AI Office Products TRAE and Kousi into DoubaoPublished in Latest AI NewsTime :Aug 24, 2026Read :3minuteOn August 24, according to media reports, ByteDance has completed the team integration of its AI office products. The TRAE and Coze (Coze) teams will be fully integrated into the Doubao system. Among them, TRAE Work and Coze will integrate their product capabilities for work scenarios with Doubao; TRAE IDE and CLI will continue to develop as the programming product line under the Doubao brand. After the adjustment, the relevant products and operations teams will report to Zhao Qi, the product leader of Doubao. (iFeng Tech)TRA and Coze were previously under ByteDance’s Product Development and Engineering Architecture Department. The former was initially positioned as an AI programming product, while the latter focused on the AI intelligent agent development platform and continuously explored different Agent directions. This integration further clarifies Doubao’s core business positioning in AI, concentrating AI office product resources on Doubao. ByteDance responded that the adjustment aims to coordinate product and technical resources to provide users with a better AI work experience, and existing user rights will not be affected. (Sina)It is reported that Doubao will launch its independent AI office product “Doubao Work” as early as this week, serving as a unified product and brand for ByteDance’s AI office scenarios. Recently, Doubao has been frequently updating features such as remote control of computers, Windows virtual desktops, cloud computers, and side workbenches, and has launched more than 200 skills and connectors. (iFeng Tech)This adjustment means that ByteDance is further consolidating its AI office product line and strengthening Doubao’s core entry point status in AI productivity scenarios by unifying the brand and Agent capabilities. With companies like Tencent and Alibaba continuing to invest heavily in AI office, the competition in the domestic AI office market is accelerating from single-point tools to unified Agent platforms.

AI NEWSLatest AI NewsArticleThe Surge of Embodied Intelligence Capital Returns, Xiaopeng Robot’s First Round Valuation Exceeds $6.3 BillionPublished in Latest AI NewsTime :Aug 24, 2026Read :3minuteXiaopeng Robotics’ business has completed its first round of financing exceeding 900 million USD, with a post-money valuation surpassing 6.3 billion USD (approximately 43 billion RMB), setting a new record for the largest private equity financing in China’s embodied intelligence industry. This round was led by IDG Capital, with Gao Rong Venture Capital participating, and was also joined by Tencent and Alibaba as strategic investors, with Xiaopeng Group maintaining its controlling stake. All four investors regard this round as one of the largest single investments in the field of embodied intelligence to date.IDG Capital pointed out that Xiaopeng’s humanoid robot represents the current leading level of domestic industry and has the technical strength to compete globally with overseas top companies. This round of financing will promote the establishment of a market valuation system for Xiaopeng Robotics’ business and accelerate the mass production process of its humanoid robots and the research and development iteration of physical AI models, helping advanced general-purpose humanoid robots enter a new stage of global commercial application.This move also reflects that the embodied intelligence sector is attracting top financial and strategic capital to accelerate their layout, with the trend of the industry shifting from technology verification to large-scale implementation becoming increasingly clear.

AI NEWSLatest AI NewsArticleGlobal First Cooking AI Foundation Model Trained on Real Data Released, Cooking Robots Enter a New Era of Embodied IntelligencePublished in Latest AI NewsTime :Aug 24, 2026Read :2minuteOn August 23, Oak Deer Robotics officially launched the world’s first multimodal cooking AI foundation model based on real kitchen data - “CookingMuse ChefStart”. This breakthrough marks a significant step forward for cooking artificial intelligence, moving from simple command execution to deep understanding and real-world application.At the same time, the company also introduced two core hardware products: the 3K Visual AI Stir-frying Robot and the embodied cooking robot “XianChao Ark”. The launch of these two new products indicates that its technical layout now fully covers the entire chain, from understanding cooking logic, perceiving and judging the environment and state, to finally embodied execution in the physical world.

Join NowENLatest AI NewsGlobal First Cooking AI Foundation Model Trained on Real Data Released, Cooking Robots Enter a New Era of Embodied IntelligenceOn Aug 23, Oak Deer Robotics launched CookingMuse, the world’s first multimodal cooking AI foundation model trained on real kitchen data, shifting cooking AI from command execution to deep understanding and real-world use. It also unveiled a 3K vision AI stir-fry robot and embodied cooking robot, covering the full cooking process…..6 hours ago11.5KThe Surge of Embodied Intelligence Capital Returns, Xiaopeng Robot’s First Round Valuation Exceeds $6.3 BillionXpeng Robotics raised over $900M in its first round, reaching a $6.3B valuation, a record for China’s embodied AI private funding. Led by IDG Capital, with Gaorong Ventures, Tencent and Alibaba; Xpeng Group stays controlling. Investors rank it among the largest embodied AI deals…..6 hours ago10.9KChinese Open-Source Models Go Global: Legal AI Unicorn Harvey Launches Tenet Based on Kimi K3U.S. legal AI unicorn Harvey announced the launch of the legal-specific model Harvey Tenet by post-training based on Moonshot’s open-weight model Kimi K3. This marks the first time a Chinese open-weight model has entered the training system of a leading U.S. vertical AI company. Harvey previously mainly relied on closed-source models such as OpenAI and Anthropic, serving over 2,400 organizations and 200,000 lawyers. It has a valuation of $11 billion and an annualized revenue of about $350 million.7 hours ago11.9KByteDance Integrates AI Office Products TRAE and Kousi into DoubaoOn August 24, ByteDance integrated its AI office products: the TRAE and Kousi teams were incorporated into the Doubao system. TRAE Work and Kousi will be integrated with Doubao’s work scenario capabilities; TRAE IDE and CLI will remain as programming product lines under Doubao. After the adjustment, the relevant products and operations teams will report to Zhao Qi, the product head of Doubao.7 hours ago12.9KSouth Korea Enacts Law to Facilitate AI: Use of Limited Personal Data After Regulatory ReviewThe South Korean National Assembly passed an amendment to the Personal Information Protection Act, allowing AI developers to use personal data within a limited scope after review by the Personal Information Protection Committee, breaking the previous rigid requirement that data be used for AI development only with individual consent. This move addresses long-standing industry complaints, as exemptions were previously limited to a few scenarios such as voice fraud prevention and autonomous driving, with short durations. The new regulations expand the available data space for AI development.7 hours ago9.7KXiaomi Xuanjie O100 Chip Launch: 6nm Process and 3D Stacking Empower AI to Surge at Full SpeedXiaomi launched Xuanjie O100, its first AI accelerator chip for large models, using a 6nm process and 3D wafer-level stacking to boost bandwidth and integration. It marks a key step in self-developed core computing chips, showing terminal vendors’ AI hardware competition deepening…..8 hours ago10.9KYahoo Bounces Back via the Internet OG: Self-Developed AI Q&A Engine Scout Targets Gen ZYahoo CEO Lanzon stated that this internet veteran, born in 1994, is turning brand trust and retro labels into unique advantages to attract Gen Z, and leveraging its self-developed AI tool Scout and differentiated data to seek growth. Since taking charge in 2021, he has restructured it into a more streamlined consumer technology group, aiming to prove that Yahoo still has a place in the AI era.8 hours ago9.6KXiaomi Releases Xuanjie O100 Chip, With the Highest Edge-side Large Model Inference Speed of 330 TPSXiaomi unveiled its first on-device LLM AI accelerator chip, Xuanjie O100. Built on 6nm wafer-level vertical stacking and hybrid bonding with 1.4μm pitch, it delivers 1.22TB/s bandwidth—16x that of flagship phones—and up to 330 TPS inference, boosting on-device LLM throughput and reasoning…..8 hours ago11.7KAI Industry’s GitHub Seeks Sale, Hugging Face Valuation May Exceed $13 BillionOn August 24, according to a report by Insider in the United States, the world’s largest AI open source community, Hugging Face, is considering a full sale, with potential transaction valuation possibly exceeding $13 billion. The company is working with banks to assess the interests of potential bidders, but no deal has been reached yet. Hugging Face was founded in 2016 by Clement Delong and Thomas Wolf, and it is headquartered in New York. The number of publicly available models has approached 3 million.8 hours ago10.8KShaoyn OpenFit 2 AI Earphones Shocking Launch: Integrates Qianwen Large Model, 48-Hour Long Battery Life Redefines Smart Audio ExperienceShaoyn OpenFit2 AI Earphones are now on sale with a launch price of 1398 yuan. The product continues the open-acoustic advantage, deeply integrates the Qianwen AI large model, and focuses on comfortable wear and smart interaction. Each e