内容来源：https://aibusiness.com/nlp/most-read-google-launches-ai-guided-learning-tool-to-teach-users-dell-tackles-unstructured-data-for-generative-ai-applications

内容总结：

【AI商业周报】谷歌推出教育导向AI学习伴侣，戴尔升级数据平台助力企业AI转型

谷歌推出"引导式学习"AI工具
谷歌云旗下Gemini推出"引导式学习"（Guided Learning）新功能，该工具突破传统问答模式，通过分步引导培养用户批判性思维。谷歌学习与可持续发展副总裁Maureen Heymans表示，该功能基于与教育专家合作开发的LearnLM模型，能动态调整解释方式，"旨在帮助用户建立深层理解而非简单获取答案"。

戴尔升级AI数据平台
戴尔科技集团发布AI数据平台升级方案，重点解决企业非结构化数据利用率低的痛点。新推出的"非结构化数据引擎"与Elastic合作开发，具备高级向量搜索和语义检索能力，可将杂乱数据转化为生成式AI可用资源。戴尔指出，当前企业数据中约80%因缺乏有效索引而无法被AI利用。

NTT DATA成立微软云全球事业部
为应对企业AI上云需求激增，NTT DATA宣布成立覆盖50国的微软云专项部门，由微软实践全球负责人Aishwarya Singh领导。该部门拥有超2.4万名认证专家，将聚焦云原生开发、安全合规等领域，助力客户应对AI时代数据中心现代化挑战。

PTC与英伟达深化合作
工业软件巨头PTC宣布将英伟达Omniverse技术集成至Creo和Windchill平台，加速AI硬件及数据中心设备的设计仿真流程。公司CEO Neil Barua强调，加入OpenUSD联盟彰显其对AI开发开放标准的承诺，新方案可让客户在实时仿真环境中处理复杂工程设计。

英伟达发布新一代机器人模型
在温哥华Siggraph 2025大会上，英伟达推出包括Cosmos Reason模型在内的多款代理AI与物理机器人新品。其中7B参数的视觉语言模型赋予机器"类人推理能力"，而RTX Pro 6000 Blackwell服务器GPU将重塑数字孪生与企业高性能计算生态。

（本期内容由谷歌云赞助呈现）

中文翻译：

由谷歌云赞助
选择您的首批生成式AI应用场景
要开启生成式AI之旅，首先应聚焦于能优化人类信息交互体验的领域。

本期内容还包括：NTT DATA成立全球微软云部门加速AI布局；PTC与英伟达合作推进AI产品设计等
以下是本周AI Business最受关注的报道。

谷歌推出了一款被称作"个人学习伙伴"的新AI工具。这款名为"引导式学习"的功能内置于Gemini中，其设计理念不仅是回答问题，更通过逐步引导帮助用户培养批判性思维。"目标是促进深度理解而非直接获取答案，"谷歌学习与可持续发展副总裁莫琳·海曼斯在博客中解释，"该功能会分解问题步骤，并动态调整讲解方式，最终帮助用户构建知识体系。"该功能是与教育工作者、学生及教学专家共同开发的专用模型LearnLM的成果。

戴尔科技升级其AI数据平台，通过增强索引和向量检索能力，将杂乱的企业数据转化为生成式AI可用的资源。更新核心是与AI搜索公司Elastic合作开发的"非结构化数据引擎"，为需要处理复杂内容的应用提供语义检索支持。

NTT DATA新成立的全球微软云业务单元横跨50余国，由24,000余名持证专家组成，专注于云原生开发与合规解决方案，响应企业对AI云服务的激增需求。

PTC宣布与英伟达深化合作，将Omniverse技术整合至Creo和Windchill平台，加速AI硬件从设计到上市的进程。该公司同时加入OpenUSD联盟，推动AI开发的开放标准。

英伟达在温哥华Siggraph 2025大会上发布了新一代自主AI与机器人模型，包括可定制的70亿参数视觉语言模型Cosmos Reason，使机器能像人类一样基于常识进行决策。

您可能还喜欢

英文来源：

Sponsored by Google Cloud
Choosing Your First Generative AI Use Cases
To get started with generative AI, first focus on areas that can improve human experiences with information.
Also inside, NTT DATA launches global Microsoft Cloud unit to accelerate AI; PTC, Nvidia partner to accelerate AI product design, and more
Here are the most-read stories on AI Business this week.
Google has launched a new AI tool that the company said acts as a “personal learning companion” to users.
The tool, Guided Learning, is available within Gemini and is designed to teach users rather than simply answering their questions, leading them through problems and helping them develop critical thinking skills.
“The aim is to help you build a deep understanding instead of just getting answers,” Maureen Heymans, Google’s vice president of learning and sustainability explained in a blog post. “Guided Learning breaks down problems step-by-step and adapts explanations to your needs — all to help you build knowledge and skills.”
The new capabilities were developed in collaboration with educators, students and “pedagogical experts,” Heymans said. Using this information, Google created a family of models designed specifically for educational teaching, called LearnLM, which were integrated into Gemini to create Guided Learning.
Dell Technologies has updated its AI Data Platform to make unstructured enterprise data more accessible for generative AI applications. The enhancements focus on transforming the vast amounts of unusable enterprise data into actionable intelligence through improved indexing and vector retrieval.
Generative AI requires high-quality, accessible data, but according to Dell, most enterprise data remains unusable without continuous indexing and robust vector retrieval engines. The updated platform directly addresses the challenge through enhanced data preparation and access capabilities.
At the core of the update is a new "unstructured data engine" developed in collaboration with AI search company Elastic. This engine provides advanced vector search and semantic retrieval capabilities essential for generative AI applications that need to process and understand unstructured content.
NTT DATA has unveiled a new global business unit dedicated to Microsoft Cloud, designed to meet “rising demand” for secure, AI-powered enterprise cloud solutions.
Led by Aishwarya Singh, the global head of Microsoft practice, cloud and agentic AI for NTT DATA, the unit spans more than 50 countries and includes a network of Microsoft-certified specialists with more than 24,000 certifications.
The move comes as organizations across sectors look to modernize IT and data center operations to support generative and agentic AI applications.
The partners said the unit will focus on cloud-native development, security and observability, while helping customers navigate complex compliance and sovereignty challenges.
PTC has announced a major expansion of its collaboration with Nvidia, integrating the computing company’s Omniverse technologies into its flagship Creo and Windchill platforms.
The move, PTC said, will “transform” how AI and data management products are designed, simulated and brought to market. These next-gen products include AI foundation infrastructure, high-performance PCBs, advanced cooling systems and large-scale data center equipment.
PTC has also joined the Alliance for OpenUSD, which it said reinforces its “commitment to open standards and data interoperability for AI development.”
“Today’s most advanced products — from AI hardware to industrial machinery — are more complex, integrated and engineering-intensive than ever before,” said Neil Barua, PTC’s CEO. “By deepening our collaboration with Nvidia and joining the Alliance for OpenUSD, we’re giving our customers the ability to incorporate design and configuration data in a real-time, immersive simulation environment.”
Nvidia has launched a series of next-generation agentic AI and physical robotics models at the Siggraph 2025 conference this week in Vancouver.
Among the announcements are new Omniverse libraries, the Cosmos Reason model and the RTX Pro 6000 Blackwell Server Edition GPU, advancements that Nvidia said would reshape digital twins, physical A, and high-performance enterprise computing.
Nvidia's new Cosmos Reason model is a customizable 7-billion-parameter vision-language model designed for physical AI, meaning machines that reason, plan and act in the physical world.
The model allows vision agents and robots to “reason like humans” using prior knowledge and common sense to make decisions.
You May Also Like