内容来源：https://aibusiness.com/agentic-ai/prompt-agentic-ai-outpacing-enterprise-readiness

内容总结：

企业加速部署AI智能体，但组织准备度明显滞后

一项最新调查显示，美国约75%的企业领导者预期AI智能体将在未来四年内重塑其半数业务流程，但仅有20%认为公司目前具备相应能力来支持这一变革。德勤发布的报告指出，尽管AI智能体部署速度不断加快，但企业在流程梳理、数据整合、成本控制及监管机制等方面的准备严重不足。

Salesforce的研究数据显示，企业平均拥有的AI智能体数量在15个月内增长了近两倍，创建并激活一个智能体的时间缩短了53%，目前不到两天。与此同时，每个账户的智能体操作数量以每月31%的复合增长率攀升。这表明，企业部署AI的速度前所未有，但如何确保其有效运作却成为更大的挑战。

业内专家指出，主要障碍集中在企业内部：业务流程不清晰、数据和系统彼此割裂，以及对改变既有工作模式的抵触情绪。随着智能体承担更多关键任务，数据获取的可靠性直接影响输出质量。此外，Gartner警告称，由于复杂推理和规划推高推理成本，AI智能体可能无法享受传统意义上的规模经济效应。HFS Research与TCS的联合调查也显示，仅35%的企业高管认为AI能持续带来业务成果、获得监管机构信任并提供充足控制。

分析人士认为，当前正出现一个明显的“剪刀差”：智能体部署越快、任务越重，支撑其运转的流程、数据、成本和控制机制却远远跟不上。AI智能体时代正快速到来，但企业的整体准备度却难以同步提速。

其他AI领域要闻速览：

• OpenAI放缓模型迭代，给CIO的启示： 该实验室决定放慢最先进模型的开发速度，提醒企业技术负责人需制定灵活的AI路线图，以应对模型能力、可用性和供应商时间表的变化。
• 美光投资100亿美元建设AI内存研发中心： 该芯片制造商将在爱达荷州新建研发中心，聚焦于满足日益复杂的AI系统对内存、计算和封装技术的需求。
• 英伟达发布SONIC模型，教人形机器人“动起来”： 该基础模型使人形机器人能够从人类示范中学习全身运动技能，推动其物理AI战略落地。
• Starling银行升级客服聊天机器人： 该银行为其AI客服助手增添新技能，以处理更复杂的银行业务查询，提供更个性化的支持。
• AI提升生产力，但未减轻IT人员工作负担： 尽管AI每周为IT专业人士节省数小时，但多数人表示总工作量未减反增，新职责抵消了效率收益。
• 美CISA与FBI警告：AI攻击瞄准西门子S7设备： 两机构警告称，黑客正利用AI辅助技术，针对能源、水务等关键基础设施领域的西门子工业控制系统发起攻击。
• Pony.AI计划在海外部署4000辆机器人出租车： 这家中国自动驾驶企业正加速全球化布局，计划在海外市场投放4000辆无人驾驶出租车。
• 企业如何跟上AI快速迭代步伐： 专家建议，企业需清晰理解自身业务流程，并保持模型和智能体切换的灵活性，以应对能力与成本的动态变化。

中文翻译：

由谷歌云赞助
选择你的首批生成式AI用例
要开始使用生成式AI，首先应聚焦于能够改善人类与信息互动体验的领域。
随着智能体部署加速，许多企业在规模化支持这些智能体所需流程、数据、成本和控制方面仍面临困境。
编者按：欢迎阅读Prompt，这是为您带来的每周AI格局动态简报。我们以分析视角解读本周重大进展，并精选重要新闻报道。
企业对AI智能体有着宏大规划。但它们是否真的准备好了，则是另一个问题。
德勤最新调查发现，约75%的美国企业领导者预计AI智能体将在四年内重塑其组织中约半数流程。但仅有20%的人认为其公司目前具备为自主智能体重新设计工作流程的能力。
这一发现恰逢一个有趣的节点。企业正在部署更多智能体，供应商让智能体更易于上线，AI能力也在持续进步。但支撑这些智能体所需的组织基础建设并未同步推进。
其结果是，智能体AI的能力与企业实际准备就绪的管理水平之间差距日益扩大。
即便如此，部署仍在加速推进。Salesforce研究显示，各组织平均拥有的AI智能体数量在15个月内增长了近两倍。与此同时，创建并激活一个智能体所需时间下降了53%，缩短至不到两天。
智能体也在承担更多工作，每个账户的平均操作数量在15个月内以31%的月复合增长率持续增长。
换言之，企业能以比以往更快的速度部署智能体。但让组织做好有效使用它们的准备，则被证明要复杂得多。
但最大的障碍在于企业内部。德勤发现，许多组织根本没有做好准备，业务流程不清晰、数据与系统相互割裂，以及员工对改变既有工作方式的抵触情绪，都是横亘在前的拦路虎。
随着智能体承担更大责任，这些障碍的影响也愈发显著。如果没有可靠的数据访问渠道，智能体将难以产出可信结果。成本是另一个担忧。Gartner发现，由于更复杂的推理和规划推高了推理成本，智能体AI可能无法享有传统的规模经济效应。
此外，可靠性问题依然存在。在HFS Research和TCS调查的高管中，仅35%表示AI能持续交付业务成果、赢得监管机构信任并提供充分的控制能力。
综合来看，这些进展指向一个日益明显的脱节。组织正在部署更多智能体并赋予其更多任务，而支撑这些智能体所需的流程、数据、成本经济性和控制手段仍在追赶之中。
智能体AI或许正在快速到来。但企业准备就绪的速度，则被证明难以同步加快。
本周其他AI新闻：
OpenAI模型减速为CIO提供AI规划一课：这家AI实验室决定放慢其最先进模型的开发速度，凸显了CIO为何需要灵活的AI路线图，以便随着模型能力、可用性和供应商时间表的变化而调整。
美光斥资100亿美元投入美国AI内存研究：这家芯片制造商正投资100亿美元在爱达荷州新建研究基地，专注于推进面向日益严苛的AI系统的内存、计算和封装技术。
英伟达SONIC教人形机器人行动：该基础模型使人形机器人能够从人类演示中学习各种全身动作，推进了该公司在物理AI领域的更广泛布局。
Starling Bank为面向客户的聊天机器人添加AI驱动技能：该银行正在扩展其面向客户的AI助手，新增技能旨在处理更复杂的银行查询并提供更个性化的支持。
AI生产力提升并未转化为IT工作负担减轻：AI每周为IT专业人员节省数小时时间，但大多数人表示其工作量保持不变甚至增加，因为新职责抵消了这些生产力收益。
CISA和FBI警告：AI辅助攻击瞄准易受攻击的西门子S7设备：两家机构警告称，黑客正使用AI辅助技术攻击关键基础设施领域（包括能源和供水）中易受攻击的西门子工业控制系统。
小马智行计划在海外部署4000辆机器人出租车：这家中国供应商计划在全球部署4000辆机器人出租车，加速其自动驾驶业务的扩张。
企业如何跟上AI快速发展的步伐：希望跟上AI飞速发展的企业，需要清晰了解自身业务流程，并具备在能力和成本变化时灵活切换模型和智能体的能力。

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.
As agent deployments accelerate, many enterprises are still struggling with the processes, data, costs and controls needed to support them at scale.
Editor’s Note: Welcome to Prompt, your weekly briefing on the shifting AI landscape. We provide an analytical look at the week’s biggest developments, paired with a curated roundup of the stories that matter.
Enterprises have big plans for AI agents. Whether they're ready for them is another question.
A new Deloitte survey found that roughly 75% of U.S. business leaders expect AI agents to reshape about half of their organizations' processes within four years. But just 20% believe their companies are currently equipped to redesign workflows for autonomous agents.
That finding lands at an interesting moment. Enterprises are deploying more agents, vendors are making them easier to launch and AI capabilities continue to advance. But the organizational foundations needed to support them are not moving at the same speed.
The result is a widening gap between what agentic AI can do and what enterprises are actually prepared to manage.
Even so, deployment continues to accelerate. Salesforce research found the average number of AI agents across organizations nearly tripled over 15 months. Meanwhile, the time required to create and activate an agent fell 53% to less than two days.
Agents are also handling more work, with the average number of actions per account growing at a compound monthly rate of 31% over the 15 months.
In other words, enterprises can deploy agents faster than ever. Preparing the organization to use them effectively is proving much more complicated.
But the biggest barriers are within the enterprise. Deloitte found many organizations simply aren't ready, with unclear business processes, disconnected data and systems, and resistance to changing established ways of working standing in the way.
Those barriers become more consequential as agents take on greater responsibility. Without reliable access to data, agents can struggle to produce trustworthy results. Cost is another concern. Gartner found that agentic AI may not benefit from traditional economies of scale as more complex reasoning and planning hike inference costs.
And questions about reliability remain. Just 35% of executives surveyed by HFS Research and TCS said AI consistently delivers business outcomes, earns regulator confidence and provides sufficient control.
Taken together, the developments point to a growing disconnect. Organizations are deploying more agents and asking them to do more, while the processes, data, economics and controls needed to support them are still catching up.
Agentic AI may be arriving quickly. Enterprise readiness is proving much harder to accelerate.
Also in AI News This Week:
OpenAI's Model Slowdown Offers CIOs a Lesson in AI Planning: The AI lab’s decision to slow development of its most advanced models highlights why CIOs need flexible AI roadmaps that can adapt as model capabilities, availability and vendor timelines change.
Micron Puts $10B Behind US AI Memory Research: The chipmaker is investing $10 billion in a new Idaho research hub focused on advancing memory, compute and packaging technologies for increasingly demanding AI systems.
Nvidia’s SONIC Teaches Humanoids to Move: The foundation model enables humanoid robots to learn a wide range of whole-body movements from human demonstrations, advancing the company's broader push into physical AI.
Starling Bank Adds AI-Powered Skills to Customer-Facing Chatbot: The bank is expanding its customer-facing AI assistant with new skills designed to handle more complex banking queries and provide more personalized support.
AI Productivity Gains Aren’t Translating to IT Workload Relief: AI is saving IT professionals hours each week, but most say their workloads have stayed the same or increased as new responsibilities offset those productivity gains.
AI-Backed Campaign Targeting Vulnerable Siemens S7 Devices, CISA and FBI Warn: The agencies are warning that hackers are using AI-assisted techniques to target vulnerable Siemens industrial control systems across critical infrastructure sectors, including energy and water.
Pony.AI Has Plans for 4,000 Robotaxis Outside China: The Chinese vendor plans to deploy 4,000 robotaxis globally as it accelerates the expansion of its autonomous driving business.
How Enterprises Can Catch Up with the Rapid Pace of AI Advances: Enterprises looking to keep pace with rapid AI advances need a clear understanding of their business processes and the flexibility to switch between models and agents as capabilities and costs change.