Latest news with #NLWeb
Business Mayor
26-05-2025
- Business
- Business Mayor
The battle to AI-enable the web: NLweb and what enterprises need to know
Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More In the first generation of the web, back in the late 1990s, search was okay but not great, and it wasn't easy to find things. That led to the rise of syndication protocols in the early 2000s, with Atom and RSS (Really Simple Syndication) providing a simplified way for website owners to make headlines and other content easily available and searchable. In the modern era of AI, a new group of protocols is emerging to serve the same basic purpose. This time, instead of making sites easier for humans to find, it's all about making websites easier for AI. Anthropic's Model Control Protocol (MCP), Google's Agent2Agent and large language models/ are among the existing efforts. The newest protocol is Microsoft's open-source NLWeb (natural language web) effort, which was announced during the Build 2025 conference. NLWeb is also directly linked to the first generation of web syndication standards, as it was conceived and created by RV Guha, who helped create RSS, RDF (Resource Description Framework) and NLWeb enables websites to easily add AI-powered conversational interfaces, effectively turning any website into an AI app where users can query content using natural language. NLWeb isn't necessarily about competing with other protocols; rather, it builds on top of them. The new protocol uses existing structured data formats like RSS, and each NLWeb instance functions as an MCP server. 'The idea behind NLWeb is it is a way for anyone who has a website or an API already to very easily make their website or their API an agentic application,' Microsoft CTO Kevin Scott said during his Build 2025 keynote. 'You really can think about it a little bit like HTML for the agentic web.' NLWeb transforms websites into AI-powered experiences through a straightforward process that builds on existing web infrastructure while leveraging modern AI technologies. Building on existing data: The system begins by leveraging structured data that websites already publish, including markup, RSS feeds and other semi-structured formats that are commonly embedded in web pages. This means publishers don't need to rebuild their content infrastructure completely. Data processing and storage: NLWeb includes tools for adding this structured data to vector databases, which enable efficient semantic search and retrieval. The system supports all major vector database options, allowing developers to choose the solution that best fits their technical requirements and scale. Read More Building a 'virtual Vegas' in honor of CES AI enhancement layer: LLMs then enhance this stored data with external knowledge and context. For instance, when a user queries about restaurants, the system automatically layers on geographic insights, reviews and related information by combining the vectorized content with LLM capabilities to provide comprehensive, intelligent responses rather than simple data retrieval. Universal interface creation: The result is a natural language interface that serves both human users and AI agents. Visitors can ask questions in plain English and receive conversational responses, while AI systems can programmatically access and query the site's information through the MCP framework. This approach allows any website to participate in the emerging agentic web without requiring extensive technical overhauls. It makes AI-powered search and interaction as accessible as creating a basic webpage was in the early days of the internet. There are a lot of different protocols emerging in the AI space; not all do the same thing. Google's Agent2Agent , for example, is all about enabling agents to talk to each other. It's about orchestrating and communicating agentic AI and is not particularly focused on AI-enabling existing websites or AI content. Maria Gorskikh, founder and CEO of AIA and a contributor to the Project NANDA team at MIT, explained to VentureBeat that Google's A2A enables structured task passing between agents using defined schemas and lifecycle models. 'While the protocol is open-source and model-agnostic by design, its current implementations and tooling are closely tied to Google's Gemini stack — making it more of a backend orchestration framework than a general-purpose interface for web-based services,' she said. Another emerging effort is Its goal is to help LLMs better access web content. While on the surface, it might sound somewhat like NLWeb, it's not the same thing. 'NLWeb doesn't compete with it is more comparable to web scraping tools that try to deduce intent from a website,' Michael Ni, VP and Principal Analyst at Constellation Research told VentureBeat. Krish Arvapally, co-founder and CTO of Dappier, explained to VentureBeat that provides a markdown-style format with training permissions that helps LLM crawlers ingest content appropriately. NLWeb focuses on enabling real-time interactions directly on a publisher's website. Dap pier has its own platform that automatically ingests RSS feeds and other structured data, then delivers branded, embeddable conversational interfaces. Publishers can syndicate their content to their data marketplace. MCP is the other big protocol, and it is increasingly becoming a de facto standard and a foundational element of NLWeb. Fundamentally, MCP is an open standard for connecting AI systems with data sources. Ni explained that in Microsoft's view, MCP is the transport layer, where, together, MCP and NLWeb provide the HTML and TCP/IP of the open agentic web. Forrester Senior Analyst Will McKeon-White sees a number of advantages for NLWeb over other options. 'The main advantage of NLWeb is better control over how AI systems 'see' the pieces that make up websites, allowing for better navigation and more complete understanding of the tooling,' McKeon-White told VentureBeat. 'This could reduce both errors from systems misunderstanding what they're seeing on websites, as well as reduce interface rework.' Microsoft didn't just throw NLWeb over the proverbial wall and hope someone would use it. Microsoft already has multiple organizations engaged and using NLWeb, including Chicago Public Media, Allrecipes, Eventbrite, Hearst (Delish), O'Reilly Media, Tripadvisor and Shopify. Andrew Odewahn, Chief Technology Officer at O'Reilly Media is among the early adopters and sees real promise for NLWeb. 'NLWeb leverages the best practices and standards developed over the past decade on the open web and makes them available to LLMs,' Odewahn told VentureBeat. 'Companies have long spent time optimizing this kind of metadata for SEO and other marketing purposes, but now they can take advantage of this wealth of data to make their own internal AI smarter and more capable with NLWeb.' In his view, NLWeb is valuable for enterprises both as consumers of public information and publishers of private information. He noted that nearly every company has sales and marketing efforts where they might need to ask, 'What does this company do?' or 'What is this product about?' 'NLWeb provides a great way to open this information to your internal LLMs so that you don't have to go hunting and pecking to find it,' Odewahn said. 'As a publisher, you can add your own metadata using standard and use NLWeb internally as an MCP server to make it available for internal use.' Using NLWeb isn't necessarily a heavy lift, either. Odewahn noted that many organizations are probably already using many of the standards NLWeb relies on. 'There's no downside in trying it out now since NLWeb can run entirely within your infrastructure,' he said. 'It's open source software meeting the best in open source data, so you have nothing to lose and a lot to gain from trying it now.' Constellation Research Analyst Michael Ni has a somewhat positive viewpoint on NLWeb. However, that doesn't mean enterprises need to adopt it immediately. Ni noted that NLWeb is in the very early stages of maturity and enterprises should expect 2-3 years for any substantial adoption. He suggests that leading-edge companies with specific needs, such as active marketplaces, can look to pilot with the ability to engage and help shape the standard. 'It's a visionary specification with clear potential, but it needs ecosystem validation, implementation tooling, and reference integrations before it can reach mainstream enterprise pilots,' Ni said. Others have a somewhat more aggressive viewpoint on adoption. Gorskikh suggests taking an accelerated approach to ensure your enterprise doesn't fall behind. 'If you're an enterprise with a large content surface, internal knowledge base, or structured data, piloting NLWeb now is a smart and necessary step to stay ahead,' she said. 'This isn't a wait-and-see moment — it's more like the early adoption of APIs or mobile apps.' That said, she noted that regulated industries need to tread carefully. Sectors like insurance, banking and healthcare should hold off on production use until there's a neutral, decentralized verification and discovery system in place. There are already early-stage efforts addressing this — such as the NANDA project at MIT that Gorskikh participates in, which is building an open, decentralized registry and reputation system for agentic services. For enterprise AI leaders, NLWeb is a watershed moment and a technology that should not be ignored. AI is going to interact with your site, and you need to AI enable it. NLWeb is one way that will be particularly attractive to publishers, much like RSS became a must-have for all websites in the early 2000s. In a few years, users will just expect it to be there; they will expect to be able to search and find things, while agentic AI systems will need to be able to access the content as well. That's the promise of NLWeb.
Forbes
21-05-2025
- Business
- Forbes
Microsoft Launches NLWeb To Simplify Website-Agent Interactions
Web pixabay Microsoft unveiled NLWeb at Build 2025, offering developers a way to integrate conversational AI capabilities into websites with minimal code. This open-source project enables websites to communicate with both human users and AI agents through natural language interfaces, potentially transforming how users and AI systems interact with web content. Unlike previous web interface innovations that focused on visual elements, NLWeb addresses a fundamental shift toward conversational interactions. As AI agents gain prominence across business and consumer applications, websites need standardized methods to communicate with these systems. NLWeb provides this bridge, allowing websites to serve as both human conversation endpoints and machine-readable data sources for emerging AI ecosystems. Essentially, NLWeb offers a standard schema and a protocol to add conversational user experience to almost any website. NLWeb functions as a lightweight framework that leverages existing web standards like and RSS to build conversational capabilities. The system processes user queries through language models, performs semantic searches against website content and generates natural responses. A retailer might implement NLWeb to help customers find specific products through conversational queries like 'show me business casual clothes under $50' rather than navigating traditional category filters. The technical architecture is designed to be provider agnostic, supporting multiple AI models and vector databases. This flexibility allows developers to choose components that match their needs without vendor lock-in. The framework includes connectors to popular language models from providers like OpenAI, Anthropic and Google and various vector database options, including Qdrant, Milvus and Snowflake. Each NLWeb implementation functions as a Model Context Protocol server, adhering to Anthropic's emerging standard for connecting AI models with data sources. This dual functionality means websites can simultaneously serve human visitors through chat interfaces while making their content available to external AI systems that support MCP. For instance, a recipe website using NLWeb could provide ingredient substitution recommendations to visitors while also enabling external AI assistants to access its recipe database when users ask cooking questions. From a development perspective, NLWeb significantly reduces implementation complexity. The GitHub repository contains core service code, model connectors, data ingestion tools and a web server frontend. Beyond the technical components, NLWeb includes tools for processing structured data from various formats including JSON-LD, RSS feeds and XML sitemaps. Several organizations have already implemented NLWeb during early testing phases. The initial cohort includes media companies like Chicago Public Media and Hearst, e-commerce platforms like Shopify and travel sites like Tripadvisor. These early implementations demonstrate the framework's versatility across different types of web content and business models. NLWeb enters a competitive landscape of agent frameworks and standards. Companies like Google, Anthropic and numerous startups offer tools for building AI agents and enabling cross-agent communication. Microsoft positions NLWeb to serve as a foundational standard similar to how HTML standardized document sharing on the web. However, the proliferation of competing standards creates challenges for adoption, as developers face potential framework fatigue amid rapidly evolving technologies. The business value proposition centers on making website content more accessible to both humans and AI systems. For website owners, implementing NLWeb potentially increases content discoverability as AI assistants become more prominent in user search behaviors. E-commerce sites could expand customer reach by making product catalogs accessible to shopping assistants, while content publishers might gain additional distribution channels through AI recommendation systems. For technology decision makers, NLWeb represents both opportunity and complexity. The framework offers a standardized approach to conversational interfaces without requiring expertise in prompt engineering or complex language model integration. However, implementation still requires technical resources for data preparation, model selection and ongoing maintenance. Organizations must also consider data privacy implications when exposing content to AI systems. The headless agent aspect proves particularly significant for enterprise architecture. NLWeb-enabled sites can function as endpoints for autonomous AI systems, enabling machine-to-machine workflows without human intervention. This capability supports emerging multi-agent systems where specialized AI agents collaborate to complete complex tasks. For example, a procurement system might use multiple agents to research products, compare specifications and place orders, with NLWeb-enabled supplier websites providing structured product information. Security and governance concerns remain important considerations. Exposing website content to AI agents through standardized interfaces creates new attack surfaces and data leakage risks. Organizations implementing NLWeb should develop clear policies regarding which content becomes available to external systems and how user interactions with conversational interfaces are monitored. Looking forward, NLWeb's impact depends on adoption rates across the web ecosystem. Microsoft brings considerable influence through its developer platforms and AI partnerships, but widespread implementation requires demonstration of tangible business value. The technology appears most immediately valuable for information-rich websites where users benefit from natural language navigation of complex content. As AI assistants become more integrated into daily workflows, NLWeb-enabled sites may gain competitive advantages through improved machine discoverability and interaction capabilities. For technology leaders, NLWeb represents another step toward an increasingly agentic computing environment where humans and AI systems collaborate through natural language interfaces. Organizations should evaluate NLWeb as part of broader AI integration strategies, considering both customer-facing applications and backend integration with emerging agent ecosystems.
Mint
20-05-2025
- Business
- Mint
Mint Explainer: Microsoft envisions a web driven by AI agents. What will it look like?
Next Story Shelley Singh The 'open agentic web' represents a significant evolution of the internet, moving from a system primarily designed for human consumption of information to one in which AI agents can autonomously understand, interact with, and act upon web content. Microsoft chairman and chief executive Satya Nadella at Microsoft Build 2025 in Seattle on 19 May. Photo: Jason Redmond/AFP Gift this article At Microsoft Build, the company's annual conclave of web developers and engineers (19-22 May), CEO Satya Nadella introduced a vision of a next-generation web powered by AI agents. At Microsoft Build, the company's annual conclave of web developers and engineers (19-22 May), CEO Satya Nadella introduced a vision of a next-generation web powered by AI agents. Microsoft's 'open agentic web' represents a significant evolution of the internet, moving from a system primarily designed for human consumption of information to one in which AI agents can autonomously understand, interact with, and act upon web content. Unlike AI chatbots, which simply respond to commands, AI agents work autonomously, collaborating with each other across different platforms and services to get things done. This shift will have profound impacts on both users and enterprises. In a post on X, Nadella elaborated, 'we are building the open agentic web. It is reshaping every layer of the stack, and our goal is to help every developer build apps and agents that empower people and organisations everywhere." For instance, coding agents are being built into GitHub, Microsoft's platform for software developers, to allow AI to autonomously fix bugs, add new features or simply maintain code. Let's take a closer look at the changes and their implications. What's the 'open agentic web'? Championed by Microsoft and others, the open agentic web is a vision of a future web in which AI agents seamlessly interact across platforms, services and devices to help users and organisations complete tasks. It's a vision of the web in which AI agents autonomously retrieve, process and act on information. On the open agentic web, AI agents will communicate with each other, making decisions and solving problems without constant human input. Enterprises will be able to create custom AI agents that understand their unique processes and workflows. This marks a step towards making AI more useful and independent, allowing people and businesses to automate tasks, improve efficiency, and create smarter and more intuitive digital experiences. All of this will be enabled by a standardised AI communication protocol developed by Microsoft called NLWeb, in which NL stands for natural language. Can you tell me more about NLWeb? NLWeb is an open project that aims to make it easy for any website to integrate natural language interfaces and AI-powered search capabilities. It allows users to engage with web content using conversational language, just like chatting with an AI assistant. NLWeb is similar to Hyper Text Markup Language (HTML), the foundation of the web as we know it. HTML is used to structure and organise web pages, and text, images and multimedia content on the web). NLWeb will do the same, but with AI agents instead of humans performing the tasks. The difference is that while HTML structures content for humans to read and interact with directly, NLWeb aims to structure content so that AI agents can understand, process, and act upon it autonomously or to better serve human queries. What difference will NLWeb make? Microsoft envisions NLWeb as a cornerstone of the open agentic web. Just like HTML revolutionised web accessibility, NLWeb aims to make AI-powered interactions standard. As AI agents become more prevalent, NLWeb will ensure websites remain discoverable and interactive. Developers will be able to tailor NLWeb to their needs, integrating custom AI models and data sources. How will different AI agents collaborate? AI agents will collaborate through multi-agent systems, in which specialised AI agents work together to complete complex tasks. For example, people's experiences with customer service chatbots can be frustrating because these bots cannot comprehend various nuances. In the open agentic web, one AI agent may handle query retrieval while another focuses on sentiment analysis and response generation, ensuring a much better experience. Also read: Why AI is central to the new browser wars AI agents will also be able to delegate tasks to other specialised agents, ensuring each focuses on its area of expertise. Agents will communicate using Agent2Agent (A2A) protocols, allowing seamless interaction across platforms. How will the open agentic web impact users and enterprises? For starters, it will create more personalised and context-aware AI. AI agents will be able to recall past interactions and provide more relevant responses. AI agents will also be able to process and act on web information autonomously. Eventually, websites will become interactive through NLWeb, allowing users to engage with them through natural language. For instance, the open agentic web will be able to create personalised experiences for users. AI agents will understand individual preferences and proactively assist with tasks such as research, planning trips, or summarising content, without users needing to hunt for information themselves. A user interested in, say, FIFA World Cup 2026, to be hosted jointly by the US, Canada and Mexico, will get information and updates on it once AI agents know of his interest. For routine tasks, AI-driven tools could generate content, suggest ideas, or automate tasks—freeing up users to focus on more creative and meaningful work. For businesses, it will be easier to deploy AI agents to handle customer service inquiries, manage supply chains, streamline workflows and more. It will also allow for better interoperability and quicker innovation, with AI agents collaborating to deliver results. What are the potential downsides? While AI agents acting autonomously sounds exciting, there are potential risks as well. That's because they are not passive, like AI chatbots, but can make decisions on their own. There could therefore be issues spanning lack of transparency, trust, ethics, biases, security, accountability to regulators, and other challenges. AI agents may make decisions in ways that are difficult for users or developers to understand, leading to trust issues. AI systems can also unintentionally reinforce biases present in their training data, leading to unfair or discriminatory outcomes. For instance, if all the CEOs of a 100-year-old company have been male, it might overlook female candidates when asked to select potential replacements. Autonomous AI agents could also be exploited by malicious actors, leading to data breaches, misinformation, or unauthorised actions. Besides, over-reliance on AI is in itself a risk. Users and businesses may become too dependent on AI agents, reducing human oversight and critical thinking in decision-making. While the risks of autonomous AI systems are significant, they can be mitigated (though perhaps not completely eliminated) through strong governance, ethical AI development, and user education. Topics You May Be Interested In Catch all the Business News, Market News, Breaking News Events and Latest News Updates on Live Mint. Download The Mint News App to get Daily Market Updates.
Time of India
20-05-2025
- Business
- Time of India
Microsoft Build 2025: From Copilot to Windows upgrades, everything that Microsoft announced at its biggest developer event of the year
Microsoft 's yearly Build conference kicked off yesterday, in Seattle, with CEO Satya Nadella taking the stage to unveil the company's latest innovations. The four-day event showcased Microsoft's continued push toward AI integration across its ecosystem, with significant upgrades to developer tools, Windows features, and cloud capabilities. Here's a rundown of the most important announcements from this year's conference. GitHub Copilot transforms into a full-fledged AI coding agent GitHub's AI assistant can now independently fix bugs and build features while developers work on other tasks. The company unveiled a major evolution of GitHub Copilot that turns it from a mere code suggestion tool into an autonomous coding agent. This new version can automatically boot a virtual machine, clone repositories, analyze codebases, and make improvements without constant developer supervision. The agent can fix bugs, add features, improve documentation, and save changes as it works, providing detailed session logs of its reasoning. When finished, it tags developers for review and can automatically address any feedback comments. Windows AI Foundry brings advanced AI capabilities to local development Microsoft introduced the Windows AI Foundry, a comprehensive toolkit empowering developers to create AI applications for Windows. The platform supports both pre-built and open-source models through Foundry Local, allowing developers to fine-tune and deploy projects with minimal friction. This announcement comes alongside native support for Model Context Protocol (MCP), often described as the "USB-C of AI apps," which enables AI applications to seamlessly communicate with other apps, web services, and Windows components. Microsoft 365 gets copilot tuning and multi-agent orchestration Businesses can now train personalized AI assistants that understand company-specific data and workflows. For Microsoft 365 users, the company unveiled Copilot Tuning, allowing businesses to train AI models on their own data, workflows, and processes. This enables companies to generate content that matches their specific style and language. Additionally, multi-agent orchestration in Copilot Studio combines the specialized skills of different AI agents to tackle complex business tasks with greater efficiency. NLWeb protocol reimagines web interactions through natural language New open protocol allows websites to behave like AI agents, responding to natural language queries. Microsoft introduced NLWeb, an open protocol that transforms how users interact with websites. Rather than navigating through traditional interfaces, users can simply prompt websites for information or to perform specific tasks using natural language. NLWeb supports the Model Context Protocol (MCP), making content more discoverable and accessible to AI agents. Technical Fellow Ramanathan V. Guha described this as part of the "fourth revolution" in personal computing—communicating with applications through free-form language. Azure AI Foundry expands with 1,900+ models including Elon Musk's Grok 3 For cloud developers, Microsoft announced Azure AI Foundry, providing tools to select and test appropriate AI models for various applications. The platform now includes access to xAI's Grok 3 and Grok 3 mini models from Elon Musk's company, alongside approximately 1,900 other AI models. Microsoft confirmed these models will be hosted and billed directly by the company, with standard Azure service level agreements. Windows Subsystem for Linux goes open-source After nine years, Microsoft open-sources WSL, allowing the developer community to contribute directly to the project. In a significant move for the open-source community, Microsoft announced that its Windows Subsystem for Linux (WSL) is now open-source. Windows chief Pavan Davuluri explained that this long-requested feature required significant operating system refactoring to allow WSL to function independently, enabling developers to make contributions that Microsoft can integrate into the Windows pipeline at scale. AI Masterclass for Students. Upskill Young Ones Today!– Join Now
Yahoo
20-05-2025
- Business
- Yahoo
2025 Microsoft Build 開發者大會：智慧代理崛起，開啟開放式 Agentic Web 新時代
AI 智慧代理（AI agents）的時代正式登場。隨著推理與知識管理技術突飛猛進，AI 模型不只變得更強大，效率也更高，展現出能用嶄新方式協助解決問題的潛力。像是 GitHub Copilot，已吸引 1,500 萬名開發者使用，其內建的代理模式和程式碼審查功能，正逐步改變開發者撰寫、檢查、部署及除錯程式碼的流程。 在企業領域，數十萬名使用者已開始透過 Microsoft 365 Copilot 進行資料研究、激盪想法與開發解方。目前已有超過 23 萬家企業組織——包含九成財星 500 大企業——使用 Copilot Studio 打造 AI agents 和自動化工作流程。 例如，富士通與 NTT Data 運用 Azure AI Foundry 協助潛在客戶排序，加快提案進度，並提供市場洞察；Stanford Health Care 則透過 healthcare agent orchestrator 建構 AI agents，進一步簡化行政作業及醫療討論會議的準備流程。 微軟在今年的 Build 開發者大會中展示了一系列針對平台、產品與基礎架構的創新，目的是讓 AI 智慧代理不只存在於概念，更實際落地在開發、研究與產業應用中。這些變革圍繞著「Agentic Web」——一種讓 AI agents 成為日常任務與決策助手的網路架構，並且是一個開放式、生態系整合的未來網路藍圖。 這樣的願景背後，有一個清楚的方向：協助開發者與組織發揮創意，驅動未來的數位轉型與創新。 在技術方面，微軟透過 GitHub、Azure AI Foundry 與 Windows 等平台，帶來多項全新工具與功能。GitHub Copilot 升級為具備智慧代理能力的程式助手，加入非同步代理功能，還新增提示管理、評估工具與控管機制；此外，GitHub Copilot Chat 在 VS Code 中也正式開放原始碼，促進更多開放式協作。 Windows AI Foundry 則成為一個從模型訓練到推論都能涵蓋的完整開發平台，支援在裝置端或雲端部署大型語言模型。Azure AI Foundry 也進一步引入模型排行榜與智能模型路由器，可根據需求自動選擇合適的模型，強化使用彈性。 在 AI agents 的建置與部署方面，微軟也推出一系列安全又容易擴展的方案，包括預設代理、客製化模組與多代理協作等功能。新登場的 Azure AI Foundry Agent Service 整合 Semantic Kernel 與 AutoGen，提供開發者友善的 SDK，同時支援模型上下文協定（Model Context Protocol, MCP）與代理互動機制（A2A）。 企業在部署 AI agents 時的資料安全與身份管理，也是本次更新的重點。Microsoft Entra Agent ID 為每個代理自動產生唯一身分識別，結合 Microsoft Purview 的資料治理機制與報告工具，有助於防堵無序擴張與潛在風險。 在 Microsoft 365 的應用層面，Copilot Tuning 功能讓企業能簡易訓練模型、建構專屬代理；同時多代理協作功能也已正式支援，可以讓不同專長的代理協同完成複雜任務，例如律師事務所就可設計專門生成法律文件的代理。 為了推動整個 AI 生態更進一步走向開放，微軟也全面支援 Model Context Protocol，並與 GitHub 共同加入其指導委員會。此外，全新開放專案 NLWeb 正式發表，未來網站可以透過 NLWeb 建立互動式對話介面，同時成為 AI agents 可探索與應用的資料端點。 最後，在 AI 加速科研應用上，微軟發表 Microsoft Discovery 平台，讓研究人員能以智慧代理技術重新設計研發流程，從藥物研發到永續發展議題都能加快探索與解決的腳步。 更多3C資訊Motorola razr 60 Ultra 實測vivo X200 Ultra 比 X200 Pro 強很多？
