Code Velocity
企业级AI

企业级AI:OpenAI创新的下一阶段

·8 分钟阅读·OpenAI·原始来源
分享
Denise Dresser探讨OpenAI企业级AI战略和平台发展的下一阶段

企业级AI成为焦点:超越实验阶段

OpenAI首席营收官Denise Dresser最近公布了企业级AI下一阶段的引人注目的愿景,强调了企业采用人工智能方式的迅速而深刻的转变。在与数百家客户进行了90天的互动后,Dresser注意到各行各业都表现出前所未有的紧迫感和信念。领导者们现在认识到AI是他们一生中最重要的技术变革,这促使他们围绕其能力从根本上重新思考和重塑公司。

这种信念清楚地反映在OpenAI的财务表现中。企业部门目前占公司收入的40%以上,并有望在2026年底前与消费者收入持平。这种指数级增长标志着AI应用已超越实验性阶段。关键指标,例如Codex达到300万周活跃用户,OpenAI的API每分钟处理超过150亿个令牌,以及GPT-5.4在智能体工作流中推动创纪录的参与度,都表明AI正在对各种业务运营产生深远影响。需求不仅来自高盛、飞利浦和State Farm等新巨头,也通过与Cursor、DoorDash、Thermo Fisher和LY Corporation等现有合作伙伴的扩展合作而增长。

行业已经度过了最初的“实验阶段”。AI现在正在执行切实的、关键的工作,这使得公司需要解决两个关键问题:如何在整个业务而非仅仅孤立的应用中部署最强大的AI,以及如何将AI无缝集成到员工的日常工作中以充分发挥他们的潜力。这些问题正在塑造OpenAI的企业战略,该战略专注于将“前沿平台”作为统一的智能层,并将“AI超级应用”作为核心员工体验。

OpenAI企业级AI的里程碑与影响

OpenAI企业级产品在战略上的进步,得到了重要的运营指标和广泛采纳的有力支撑。这些数据凸显了AI对企业产生的切实影响以及企业对OpenAI基础AI技术日益增长的依赖。

指标/重点领域详情影响
企业收入占总收入的40%以上,有望在2026年底前与消费者收入持平表明强大的B2B采纳率和市场成熟度的加速
Codex周活跃用户300万表明对开发者和智能体任务自动化的高实用性
API令牌处理量每分钟超过150亿个令牌凸显了在全球AI基础设施中的大规模和基础作用
GPT-5.4参与度在智能体工作流中推动创纪录的参与度展示了先进模型能力和复杂任务自动化程度的提升
前沿平台愿景实现全公司范围内的互联AI智能体从点解决方案转向统一的、上下文感知的AI操作
AI超级应用愿景在一个界面中统一AI智能体、浏览和高级功能简化日常工作流,倍增个人和团队生产力

通过前沿平台赋能全公司范围的AI智能体

OpenAI认识到存在显著的“能力过剩”,即当前AI模型拥有远超大多数企业目前利用的潜力。它的承诺是通过使前沿智能变得既可用又可信,并将其深度嵌入到业务运营的结构中来弥合这一差距。公司一直以来的痛点是孤立的AI点解决方案的激增,这些解决方案无法相互通信,导致混乱而非协同。企业现在正在寻求为其业务提供一个统一的AI操作层。

这正是OpenAI通过其“前沿平台”所提供的。Frontier旨在帮助Oracle、State Farm和Uber等客户在其整个组织中无缝构建、部署和管理AI智能体。与将智能体限制在单一产品或环境中的解决方案不同,Frontier允许智能体遍历公司的系统和数据,在各种工具中工作并随着时间推移持续改进。这种方法确保AI同事植根于公司上下文,连接到内部系统和外部数据源,并受强大的权限和控制管理。

OpenAI不仅仅是一个研究强手;它也是一家部署公司。通过与数百家大型企业的直接接触,OpenAI积累了经验并开发了可扩展的基础。通过与麦肯锡公司(McKinsey & Company)、波士顿咨询集团(BCG)、埃森哲(Accenture)和凯捷(Capgemini)等领先咨询公司建立的“前沿联盟”,以及与亚马逊网络服务(AWS)、Databricks和Snowflake等技术合作伙伴的合作,OpenAI将其智能集成到现有的企业基础设施和数据生态系统中。一个典型的例子是与AWS共同开发的有状态运行时环境,它允许AI智能体保留上下文、记住过去的行动,并在企业的工具和数据之间有效地操作,以应对复杂的实际应用。

通过统一AI超级应用赋能个人和团队

随着AI在组织内部的规模化采纳,它必须无缝集成到每个个人和团队的日常工作流程中。这推动了OpenAI向“统一AI超级应用”的战略转型——这是一个单一平台,员工可以在其中轻松地与AI智能体协作,以完成任务并在其现有工具中采取行动。这个超级应用有望结合ChatGPT、Codex、智能体浏览以及其他高级功能的最佳特性,显著提高个人和小型团队的生产力。

在领先的AI用户中,已观察到一个显著的转变:他们正从使用AI协助任务转向主动管理整个智能体团队来代表他们执行任务。这一演变始于Codex等智能体工具,仅今年就实现了超过5倍的增长。GitHub、Nextdoor、Notion和Wonderful等公司已在构建复杂的多种智能体系统,能够执行端到端的工程工作。这种智能体驱动的方法也正在各种业务职能中得到采纳;例如,OpenAI自己的销售团队使用一个智能体来研究潜在客户、评估销售线索、向合格联系人发送个性化电子邮件,并自动更新CRM系统。

OpenAI致力于使这些先进的智能体解决方案对所有企业更易于访问。一个显著的优势在于它能够连接个人和专业用例。凭借9亿周活跃的ChatGPT用户,庞大的员工群体已经熟悉与OpenAI的AI进行交互。这种熟悉度大大减少了企业的推广阻力,加速了每个员工可以委派繁琐任务并专注于更宏大、更具战略意义项目的时刻。

在AI转型中建立信任和信心

OpenAI的第一季度再次强化了AI转型正在以超乎大多数人想象的速度展开的信念。企业正在积极寻找一个合作伙伴,这个合作伙伴不仅理解这一转变的巨大性,而且能够自信地引导他们完成这一转型。这包括在企业现有系统中与其对接,提供从初步实验到全面部署的实用路线图,并通过灵活的定价和打包简化采纳过程。

至关重要的是,企业需要相信其AI合作伙伴已深度投入到他们的成功中,并积极构建适合其不断发展需求的解决方案。OpenAI强调其在各个层面和职能部门的承诺,以持续赢得这份信任。公司全心全意致力于帮助企业以及推动企业发展的人们,以清晰、信心和坚定不移的信任重塑其组织,迎接AGI的未来。这既是无与伦比的机遇,也是OpenAI与客户和合作伙伴共同承担的深远责任。

常见问题

What is OpenAI's vision for the next phase of enterprise AI?
OpenAI envisions a future where enterprise AI moves beyond isolated copilots and assistants to become a unified, foundational intelligence layer across entire businesses. This involves two core tenets: the 'Frontier' platform serving as the underlying intelligence governing all of a company's AI agents, and a 'unified AI superapp' acting as the primary interface where employees seamlessly interact with these agents to accomplish tasks. The goal is to fully embed capable AI into daily work, empowering employees to unlock their full potential and drive significant business transformation, leveraging OpenAI's full-stack approach from infrastructure to user interfaces.
How has OpenAI's enterprise business grown, and what metrics demonstrate this?
OpenAI's enterprise segment has shown remarkable growth, now accounting for over 40% of its total revenue, and is projected to reach parity with consumer revenue by the end of 2026. Key metrics highlighting this expansion include Codex achieving 3 million weekly active users, indicating widespread adoption for agentic task automation. Furthermore, OpenAI's APIs are processing an astonishing 15 billion tokens per minute, underscoring its role as a critical infrastructure provider. The advanced GPT-5.4 model is also driving record engagement across diverse agentic workflows, demonstrating increasing sophistication and utility in business applications.
What is the significance of the 'Frontier' platform in OpenAI's enterprise strategy?
The 'Frontier' platform is central to OpenAI's enterprise strategy, addressing the challenge of fragmented AI point solutions. It's designed to be a unified operating layer for businesses, enabling AI agents to operate across a company's diverse systems and data. Unlike solutions that confine agents to single products, Frontier allows agents to move fluidly between tools, maintain context, and continuously improve, grounded in a company's specific context. This platform, developed with partners like AWS, Oracle, and State Farm, provides the necessary permissions, controls, and external data source connectivity to deploy and manage AI agents company-wide effectively and securely.
Explain the concept of a 'unified AI superapp' for enterprise use.
A 'unified AI superapp' represents OpenAI's vision for a singular, seamless interface where employees can interact with AI agents throughout their workday. This superapp will integrate the best functionalities of existing OpenAI products like ChatGPT and Codex, combined with advanced agentic browsing and broader capabilities. Its purpose is to multiply individual and team productivity by providing one centralized place for employees to delegate tasks to AI agents, which can then take action across various tools they already use. This approach aims to make AI effortlessly accessible and deeply embedded in daily workflows, moving from AI assisting with tasks to AI managing teams of agents to complete tasks.
How are companies utilizing AI agents to transform their operations, as observed by OpenAI?
Companies are rapidly moving beyond basic AI experimentation to leverage AI agents for substantial operational transformation. The shift is from using AI for task assistance to managing 'teams of agents' that can execute complex, end-to-end tasks. For instance, customers like GitHub and Notion are building multi-agent systems for engineering work. OpenAI's own sales team uses an agent to research prospects, score leads, send personalized emails, and update CRM, showcasing practical, function-specific agent deployment. This agent-driven approach enables delegation of tedious tasks, freeing employees to focus on more ambitious and strategic projects, fundamentally changing how work gets done across various business functions.
What role do partnerships play in accelerating OpenAI's enterprise AI deployment?
Strategic partnerships are crucial to OpenAI's enterprise AI deployment strategy, helping to bridge the gap between frontier AI research and practical business integration. Through 'Frontier Alliances' with leading consulting firms like McKinsey & Company, Boston Consulting Group (BCG), Accenture, and Capgemini, OpenAI helps enterprises effectively integrate its intelligence. Additionally, collaborations with technology partners such as Amazon Web Services (AWS), Databricks, and Snowflake ensure that OpenAI's solutions are compatible with the existing infrastructure and data ecosystems of large enterprises. For example, the Stateful Runtime Environment, co-developed with AWS, allows AI agents to maintain context and operate seamlessly across diverse business tools and data, accelerating real-world application.

保持更新

将最新AI新闻发送到您的收件箱。

分享