Tibo 访谈：极速模式、重置机制与 OpenAI 的 Agent 路线

简要说明Brief Description

在这段对话中，Tibo 分享了从 DeepMind 和 OpenAI 获得的经验、ChatGPT 和 Codex 背后的产品发展方向，以及更快的模型为何会改变人们与智能体协作的方式。他描绘了这样一种未来：个人 AI 会适应每个个体，而不是要求个体去适应软件；同时，他也谈到了模型效率、安全性、广泛普及，以及 OpenAI 在产品体验未达预期时向用户提供补偿的做法。In this conversation, Tibo discusses lessons from DeepMind and OpenAI, the product direction behind ChatGPT and Codex, and why faster models change the way people work with agents. He describes a future in which a personal AI adapts to an individual rather than asking the individual to adapt to software, while also addressing model efficiency, safety, broad access, and OpenAI's practice of compensating users when a product experience falls short.

目录Table of Contents

• 从 DeepMind 研究到偏向快速交付
  From DeepMind research to a bias for shipping
• 打造赋能型产品文化
  Building an empowered product culture
• 个人智能体、运行框架与云规模工作
  Personal agents, harnesses, and cloud-scale work
• 用更快的智能体管理注意力
  Managing attention with faster agents
• ChatGPT 和 Codex：一个自适应产品
  ChatGPT and Codex as one adaptive product
• 自然且融入环境的交互方式
  Natural and ambient interaction
• 竞争、社区与使用量重置
  Competition, community, and usage resets
• 计算能力与递归改进
  Compute capacity and recursive improvement
• 安全暂停与操作判断
  Safety pauses and operational judgment
• 超快速模式会带来哪些变化
  What ultrafast mode changes
• 效率、可及性，以及尝试人工智能的理由
  Efficiency, access, and the case for trying AI

从 DeepMind 研究到偏向快速交付From DeepMind research to a bias for shipping

马修：在 ChatGPT 改变公众对话方式之前，谷歌就有一款内部语言模型聊天产品。参与开发那些似乎预示着这一时刻即将到来的产品，是怎样的体验？Matthew: Before ChatGPT changed the public conversation, Google had an internal language-model chat product. What was it like to work on products that seemed to anticipate that moment?

Tibo：DeepMind 是一个极具创造力的地方。我的工作主要集中在基础设施和能够加速研究的产品上，而另一个团队则在推进语言模型并扩大其规模。当成果足够出色之后，自然就会有人问，这些模型能否成为人们可以聊天、并用于完成许多任务的工具。一个类似 LM Chat 的内部产品由此诞生，而且早在 ChatGPT 发布大约一年前，就已经有了将其公开推出的雄心。Tibo: DeepMind was an exceptionally creative place. My work was largely on infrastructure and products that accelerated research, while another group was advancing language models and scaling them. Once the results became strong enough, it was natural to ask whether the models could become something people could chat with and use for many tasks. An internal product resembling LM Chat emerged, and there was an ambition to make it public roughly a year before ChatGPT.

DeepMind 的组织方式并不像 OpenAI 那样以推出产品为导向。在 OpenAI，研究与产品团队紧密合作：一起构思、共同设计，并且非常注重把有用的东西交到人们手中。使命、团队、人才密度，以及让成果触达用户的能力，这些因素共同构成了我加入 OpenAI 的重要原因。DeepMind was not organized to ship products in the way OpenAI is. At OpenAI, research and product work closely together: they ideate together, co-design, and have a strong bias toward getting useful things into people's hands. That combination of mission, people, talent density, and availability to users was a major reason to join.

早期的模型让人觉得与众不同。它们能生成连贯的文本，起初，输出内容带来的乐趣多于实际帮助。随着时间推移，它们逐渐变得更有用。这段经历至今仍是一个有益的参照，让我们思考一种文化中哪些部分值得保留，又该避免哪些失败。The early models felt special. They produced coherent text, and at first the output was more amusing than helpful. Over time, it steadily became more useful. That experience remains a useful reference point for thinking about what parts of a culture to preserve and what failures to avoid.

打造赋能型产品文化Building an empowered product culture

Matthew：那段经历中的哪些教训塑造了你希望在 OpenAI 建立的文化？你会对试图打造类似文化的创始人说些什么？Matthew: What lessons from that period shape the culture you want at OpenAI, and what would you tell a founder trying to build a similar culture?

Tibo：OpenAI 是一种自下而上、赋能员工的组织。大家可以提出想法、组建团队，并迅速推出产品，而新产品创意所受的阻力相对较小。这令人振奋，因为其目标是带来积极影响；但赋能不能演变成一堆彼此无关的功能。它必须与简洁性、统一连贯的方向，以及对产品质量的自豪感相平衡。Tibo: OpenAI is bottom-up and empowering. People can bring forward ideas, form a group, and ship something quickly with relatively little stopping energy around new product ideas. That is exciting because it is aimed at positive impact, but empowerment cannot become a hodgepodge of unrelated features. It has to be balanced by simplicity, a coherent direction, and pride in product quality.

组织致力于打造令人愉悦的体验，并重视性能、效率与简洁。产品可以快速发展，同时不背离这些原则。对创始人而言，坚定的信念至关重要；尽早找到用户，并根据反馈快速迭代也同样重要。公司还需要有意愿主动颠覆自己：新的研究和想法值得投入，即使这意味着要从成熟的主营业务中调配资源。The organization invests in delight, performance, efficiency, and simplicity. A product can move quickly without abandoning those principles. For founders, conviction matters, as does finding users early and iterating quickly from feedback. Companies also need the willingness to disrupt themselves: new research and new ideas can deserve investment even when that means reallocating resources away from the established main business.

随着公司的发展成熟，这一点可能会变得更加困难。AI 的未来不会等到下个季度既定产品推出后才到来。团队需要对技术浪潮的走向保持敏锐，直接试用模型，而非只依赖基准测试；当模型能力展现出帮助人们的新方式时，也要调整产品思路。As companies mature, that can become harder. The future of AI does not wait for the next quarter's established product. Teams need to remain open-eyed about where the wave is going, experiment directly with models rather than relying only on benchmarks, and adjust product thinking when capabilities reveal new ways to help people.

个人智能体、运行框架与云规模工作Personal agents, harnesses, and cloud-scale work

Matthew：随着模型不断进步，智能体运行框架的哪些部分仍有创新空间？Matthew: What parts of the agent harness are still open to innovation as models improve?

Tibo：仍有很大的改进空间。熟练使用编码智能体的用户已经学会了忍受一些笨拙的操作：维护技能文件、应对不完善的记忆，以及显式管理子智能体。这些机制可能很有用，但也会破坏智能体作为无缝协作伙伴的感觉。Tibo: There is still a great deal to improve. Sophisticated users of coding agents have learned to live with clunky constructs: maintaining skill files, dealing with imperfect memory, and explicitly managing subagents. These mechanisms can be useful, but they also break the feeling that the agent is a seamless partner.

目标是打造一个智能体，让它深入了解一个人的目标、日常工作、知识和团队，能够做出恰当回应，也能主动采取行动。界面应保留这种伙伴关系的感觉。语音已经体现了这一方向：自然的语音交互和强大的听写功能改变了人们使用模型的方式，也让人们更容易向基于电话的智能体布置一系列可借助已连接工具完成的任务。The goal is an agent that deeply understands a person's goals, daily work, knowledge, and team, reacts appropriately, and can be proactive. The interface should preserve that sense of partnership. Voice already points in this direction: natural voice interaction and strong dictation change how people use a model, including by making it easier to give a phone-based agent a set of tasks that can use connected tools.

能力更强的模型也凸显了笔记本电脑作为主要工作环境的局限。笔记本电脑是围绕人的局限设计的：一个人打字、思考和处理工作的速度，以及一个人能够管理多少个应用程序。而模型不必受这些限制。它可以同时跨多个应用程序工作，因此云端资源、并行工作以及一种新型的运行框架变得越来越重要。More capable models also expose the limits of the laptop as the primary work environment. Laptops were designed around human limits: how fast a person types, thinks, and consumes work, and how many applications a person can manage. A model need not share those limits. It could work across many applications at once, which makes cloud resources, parallel work, and a new kind of harness increasingly important.

使用更快的智能体管理注意力Managing attention with faster agents

Matthew：如果令牌速度大幅提升，独立开发者的工作流程会发生怎样的变化？Matthew: If token speeds become dramatically faster, how does a solo developer's workflow change?

Tibo：随着模型速度越来越快，工具调用、网络操作和其他堆栈开销可能会成为瓶颈。一种应对方式是并发：探索、编写测试、编译和验证假设都可以同时进行。但让人们并行处理更多工作并不一定更好。启动十个或十五个代理会增加认知负担，还会让人们在等待结果时不得不频繁切换上下文。Tibo: As models get faster, tool calls, network work, and other stack overheads can become the bottleneck. One response is concurrency: exploration, test writing, compilation, and testing a hypothesis can happen at the same time. But giving people more parallel work is not automatically better. Launching ten or fifteen agents creates cognitive overhead and forces context switching while the person waits for results.

产品应当尊重用户的注意力。它应该判断此刻的打扰是否有用，还是应该稍后再说；同时，让用户保持专注，在构想、原型和简短报告实时呈现的过程中不被打断。技术应适应用户，而不是要求用户去适应技术。The product should be friendly to attention. It should understand whether an interruption is useful now or should wait, and it should let the person remain in flow while ideas, prototypes, and small reports appear in real time. The technology should adapt to the user instead of requiring the user to adapt to the technology.

智能体问题大致分为两类。一类是个人智能体：融入某人的工作流程、为其量身定制，能够提出有用的想法，并处理技术工作、研究或提供建议。另一类是完全自动化：系统能够承担复杂流程，例如读取生产日志、进行性能优化、识别回归问题，或帮助缩短漏洞未修复的时间。在这类系统中，人类可能只需批准高风险操作，而不必全程参与。There are two broad categories of agent problems. One is the personal agent: an agent in a person's flow, tailored to that person, able to surface useful ideas and handle technical work, research, or advice. The other is full automation: systems that can take on complex processes such as reading production logs, performing performance optimizations, identifying regressions, or helping reduce the time a vulnerability remains open. In those systems, the human may only need to approve high-risk actions rather than remain in every loop.

将 ChatGPT 和 Codex 打造成一个自适应产品ChatGPT and Codex as one adaptive product

Matthew：ChatGPT 和 Codex 的融合受到怎样的反响？预期的最终形态是什么？Matthew: How has the convergence of ChatGPT and Codex been received, and what is the intended end state?

Tibo：最初有人担心将它们合并，但未来的模型需要一个统一的产品。目标是打造一个能力强大的个人智能体，无论用户是否在编写代码，都使用相同的底层技术和运行框架。它应该具备多模态能力，以语音为先，高效且能胜任各种任务。Tibo: There was initial concern about merging them, but the future models call for a unified product. The aim is a highly capable personal agent using the same underlying technology and harness regardless of whether someone is coding. It should be multimodal, voice-first, efficient, and capable across tasks.

人们不应该被迫在面向编程人员的界面和面向非技术人员的界面之间做出选择。软件工程师或设计师之类的标签是有用的抽象概念，但每个人都处于一个连续谱上，各自有不同的目标、工具和工作方式。同一个个人 AGI 既可以服务于技术能力很强的用户，也可以服务于这位用户的父母，同时根据不同的任务、已连接的工具和偏好进行调整。People should not need to choose between a coder interface and a nontechnical interface. Labels such as software engineer or designer are useful abstractions, but individuals sit on a spectrum and have their own goals, tools, and ways of working. The same personal AGI could serve a technically sophisticated user and that user's parent, while adapting to different tasks, connected tools, and preferences.

理想的交互应植根于人与人之间的沟通。自然语言之所以奏效，是因为人们早已懂得如何使用它，也能理解的不只是字面意思，还有情绪与微妙之处。优秀的 AI 应减少人们因它未能领会请求背后的意图或语气而产生的被误解感。它应让人觉得，是人们在现实世界中一贯行事方式的自然延伸。The desired interaction is rooted in human communication. Natural language succeeded because people already know how to use it and can understand not only literal words but emotion and nuance. A good AI should reduce the experience of feeling misunderstood because it missed the intent or tone behind a request. It should feel like a natural extension of how people already act in the world.

自然与环境式交互Natural and ambient interaction

Matthew：人类交流包括手势、面部表情和共享语境。这对人工智能有多重要？Matthew: Human communication includes gesture, facial expression, and shared context. How important is that for AI?

Tibo：未来很可能是无处不在、自然流畅的。有人在白板上写下想法、在办公室里口头梳理问题，或想通过语音进行自然对话时，智能体都应该能够参与其中。新版 ChatGPT 语音功能推出后，语音交互的增长印证了一个简单的道理：当技术变得更加自然、使用门槛降低时，人们往往会选择它。Tibo: The future is likely to be ambient and natural. An agent should be able to participate when someone writes an idea on a whiteboard, speaks through a problem in an office, or wants to have a natural conversation over voice. The growth of voice interaction after the launch of newer ChatGPT voice capabilities reinforces a simple lesson: when technology becomes more natural and lowers friction, people tend to choose it.

对一些用户来说，在小文本框中输入很自然，但并非人人如此。语音、视觉上下文以及其他非文本交互方式，都能让使用体验更轻松、功能更强大。发展方向并非为了增加模态而增加模态，而是让技术更善于理解人们想要做什么。Typing into a small text box can be natural for some users but not for everyone. Voice, visual context, and other non-text interactions can make the experience easier and more capable. The direction is not merely adding modalities for their own sake; it is making the technology better at understanding what a person is trying to do.

竞争、社区和使用量重置Competition, community, and usage resets

Matthew：你如何看待竞争市场，尤其是与 Anthropic 的比较？Matthew: How do you view the competitive market, especially comparisons with Anthropic?

Tibo：重点是打造能力最强、效率最高的模型，并以开发出人人都能使用的产品为荣。合并 ChatGPT 和 Codex，是让强大的技术在各种岗位上更安全、更易用的一部分：包括产品管理、设计、销售、营销、传播等。广泛的现有用户群有助于快速推广这些能力。Tibo: The focus is on building the most capable and efficient models, then taking pride in products that can reach everyone. Merging ChatGPT and Codex was part of making powerful technology safer and easier to use across roles: product management, design, sales, marketing, communications, and beyond. A broad existing user base can help distribute those capabilities quickly.

与其密切关注竞争对手，不如思考 OpenAI 有哪些独特优势、秉持什么价值观，以及能以多快的速度朝着这些目标迈进。社区至关重要：倾听用户的想法、保持透明，并为全世界而非仅为公司构建技术，这些都能激发活力，带来脚踏实地的使命感。Rather than watching competitors closely, the more useful question is what OpenAI can do uniquely well, what its values are, and how quickly it can move toward them. Community matters: taking ideas from users, being transparent, and building technology for the world rather than only for the company creates energy and a grounded sense of purpose.

广为人知的使用限额重置，最初是对产品故障或体验不佳的直接回应。团队在迭代过程中，如果弄坏了某些功能，或错误配置了用户体验，就会额外提供一些使用额度，以此承认用户依赖我们的产品，并感谢他们愿意尝试早期版本。重置按钮正是源于这一原则，而非营销或财务流程。如果体验不够好，我们的本意就是作出补偿。The well-known usage-limit resets began as a direct response to product failures or a suboptimal experience. When the team was iterating, broke something, or misconfigured an experience, giving users additional usage was a way to acknowledge that users rely on the product and to thank them for trying an early system. The reset button grew out of that principle, not from a marketing or finance process. If the experience is not good enough, the intent is to make up for it.

算力与递归式改进Compute capacity and recursive improvement

Matthew：这种慷慨是否需要进行大量的容量规划？你们如何看待加速模型进步？Matthew: Does that generosity require substantial capacity planning, and how do you think about accelerating model progress?

Tibo：计算规划很重要，但有效利用算力同样重要。更强大的模型和效率更高的模型相辅相成。先进模型可以帮助人们理解系统、改进代码，并推进研究和工程流程中的各个环节。这是一种切实可行的递归式自我改进：不一定是模型独立创造出下一代模型，而是由能力强大的模型协助改进开发未来能力所依赖的关键路径。Tibo: Compute planning matters, but so does using capacity well. Better models and more efficient models reinforce one another. Advanced models can help people understand systems, improve code, and work through parts of the research and engineering process. That is a practical form of recursive self-improvement: not necessarily a model independently creating the next model, but capable models helping improve the critical path used to develop future capabilities.

这也改变了产品的演进方式。团队可以使用这些系统，了解它们在哪些方面有帮助、在哪些方面会失效，然后将这些经验应用到下一次迭代中。组织需要具备足够的能力来服务用户，同时也要让内部团队能够理解并强化系统。This also changes how products evolve. Teams can use the systems, learn where they help and where they fail, then apply those lessons to the next iteration. The organization needs enough capacity to serve users while still giving internal teams the ability to understand and harden the system.

安全暂停与操作判断力Safety pauses and operational judgment

Matthew：当出现安全问题时，人们应如何看待培训或开发工作的暂停？Matthew: How should people interpret a pause in training or development when safety questions arise?

Tibo：当有必要暂停时，这能让团队和个人有时间理解并加固系统的各个部分。这样做的目的是，只有在全面掌握情况后才重新启动。OpenAI 应能够在需要时高效地做出这类决定。Tibo: When a pause is necessary, it gives teams and individuals time to understand and harden all parts of the system. The purpose is to restart only with full command of the situation. OpenAI should be able to make that kind of decision efficiently when it is needed.

这一决策不能简化为单一的机械性门槛。它应由安全团队负责，并随着证据不断浮现，通过讨论和探索来作出。同时，团队可以确立明确的原则，界定何时具备良好条件恢复工作。必须将能力与安全结合起来考量，而不能把二者视为彼此独立的问题。The decision is not reducible to a single mechanical threshold. It belongs with the safety team and involves debate and discovery as evidence develops. At the same time, the team can establish clear principles that define when it is in a good position to resume. Capabilities and safety have to be considered together rather than as independent concerns.

超快模式会改变什么What ultrafast mode changes

Matthew：当令牌生成速度大幅提升后，可以实现哪些用例？Matthew: What use cases become possible when token generation is much faster?

Tibo：在事关重大的时刻，超快速访问尤其有价值。例如，发生服务中断时，事故指挥人员和响应团队都能从中受益，因为每一秒都至关重要。正在处理关键事项的团队，或需要在近期决定是否将某个想法纳入重要版本的团队，也能从即时迭代中受益。Tibo: Ultrafast access is especially valuable in high-stakes moments. During an outage, for example, incident commanders and response teams benefit because seconds matter. Teams working on something critical, or facing a near-term decision about whether to include an idea in an important release, can also benefit from immediate iteration.

这种优势取决于任务和个人的工作风格。偏好专注于一件事的人，在系统能够跟上思考速度做出响应时，可以始终保持心流状态；另一些人则擅长切换上下文、做出许多细小的决策。超高速生成在需要大量生成、但工具调用相对较少的任务中最为突出，例如快速制作网站或游戏原型。当智能体的执行过程包含大量网络或工具操作时，这些外部开销会限制端到端的提速效果。The advantage depends on the task and on personal working style. People who prefer to stay focused on one thing can remain in flow when the system responds at the pace of thought. Others thrive on context switching and making many small decisions. Ultrafast generation is most striking when a task needs substantial generation with relatively few tool calls, such as rapidly prototyping a website or game. When the agent trajectory contains many network or tool operations, those external costs limit the end-to-end speedup.

OpenAI 并不会简单地为每位员工提供无限的超高速算力。内部使用有助于理解和改进产品，也有助于实现递归式改进，但大部分算力都留给了客户。除了文本之外，快速交互还可能支持共享画布、实时创作、快速探索图像，以及让人们通过语音或文本引导原型的开发。工程师有时不必花很长时间只是在脑中推演系统的需求和权衡取舍，而是可以迅速创建一个版本，观察它的表现，并在创作思路仍然流畅时不断迭代。OpenAI does not simply give every employee unlimited ultrafast capacity. Internal use is valuable for understanding and improving the product and for recursive improvement, but most capacity is reserved for customers. Beyond text, fast interaction may enable shared canvases, real-time creation, rapid image exploration, and prototypes that a person can steer through voice or text. Instead of spending a long time only reasoning about a system's requirements and trade-offs, an engineer can sometimes create a version quickly, see how it behaves, and iterate while still in the creative flow.

效率、获取机会，以及尝试 AI 的理由Efficiency, access, and the case for trying AI

马修：超高速能力会继续作为高级功能吗？对于那些对人工智能感到不安的人，你有什么想说的？Matthew: Will ultrafast capability remain a premium feature, and what would you say to people nervous about AI?

Tibo：随着时间推移，技术通常会变得更加普及。进步不仅来自推理速度的提升，也来自模型的 token 效率提高，以及硬件的持续改进。一两年后，如今的超高速可能就会接近常态，尽管始终会有一个更高端的层级，通过使用更多硬件来获得额外能力。Tibo: Technology generally becomes more broadly accessible over time. Progress comes not only from raw inference speed but also from models becoming more token-efficient and from continued improvements in hardware. In a year or two, current ultrafast speeds may be close to normal, even if there will always be a higher tier that uses more hardware for extra capability.

广泛普及有赖于高效率。Luna 这样的小型模型成本可能极低，却能提供几个月前还称得上前沿水平的能力。随着推理服务成本降低，强大的智能便能惠及更多人，并在日常生活中提供更直接的帮助。人们期待，如今的前沿模型随着时间推移，运行成本会大幅下降。Broad access depends on efficiency. A smaller model such as Luna can be remarkably cheap while delivering capability that would have looked frontier-level only months earlier. As serving costs fall, powerful intelligence can reach more people and provide more direct daily utility. The expectation is that today's frontier becomes much cheaper to run over time.

对于那些犹豫是否尝试 AI 的人来说，最实用的起点其实近在眼前。人们已经在用 ChatGPT 写作、获取个人建议，以及在就诊前了解更多信息。新的健康和金融体验或许很有用，但不能取代专业判断。了解他人如何使用这些工具，并从它们带来的实际益处中汲取灵感，是发现 AI 或许能在哪些方面帮助自己的合理方式。For people hesitant to try AI, the most useful starting point is close at hand. People already use ChatGPT for writing, personal advice, and to become better informed before a doctor's visit. New health and finance experiences can be useful, but they do not replace professional judgment. Seeing how others use the tools and drawing inspiration from their practical benefits is a reasonable way to discover where AI may help in one's own life.