微软CEO：如果所有价值都集中在少数几个模型手中，政治经济体系绝对无法容忍（中英全文）

来源：市场资讯

（来源：财经会议圈）

刚刚，微软董事长兼CEO纳德拉，发表一封长信，讲得非常深刻。

其中一段话发人深思：

我们最不希望看到的就是，

所有行业、所有公司都将价值

拱手让给少数几个攫取一切的模型。

如果所有价值都集中在少数几个模型手中，

政治经济体系绝对无法容忍。

社会绝不会允许人工智能的未来掏空整个行业。

原文如下：

我一直在思考公司在人工智能驱动的经济环境下的未来发展方向。

这次转型与以往任何平台变革都截然不同。过去，我们利用数字系统来提升人力资本。而现在，我们首次能够在人与数字系统之间建立真正的认知闭环。这令人耳目一新，因为它彻底改变了我们对企业内部工作的理解。

关键不在于某些数字工具或系统及其使用，而在于在人工智能模型可以不断吸收人类和组织的专业知识并将其商品化的世界中，组织如何继续学习、构建知识产权、实现差异化并蓬勃发展。

每家公司都必须构建我所谓的人力资本和代币资本。人力资本包括员工的知识、判断力、人脉关系、创造力和模式识别能力，而代币资本则是公司构建和拥有的人工智能能力。

重要的是，随着代币资本的增长，人力资本的价值并不会降低，只会增加！我相信人的主动性将是代币资本增长的驱动力。人类会设定远大的目标，将不同领域的信息联系起来，建立人脉关系，并识别出最重要的模式。如果没有人的引导，计算机就会原地打转。

这意味着真正的机遇不在于选择最佳模型，而在于构建一个基于模型的学习循环，使人力资本和代币资本能够复利增长。你可以外包一项任务，甚至一份工作，但你永远无法外包学习。企业的未来在于能否在人员和人工智能之间复利增长这种学习成果。

这需要一种全新的架构方法，让每个企业都能构建随着时间推移不断改进的智能系统，同时又能保持对其知识产权的控制权。企业应该能够在不丢失其学习系统中内置的“公司资深人士”专业知识的情况下，替换掉现有的“通用”模型。这将是未来时代对企业控制权和自主权的关键“考验”。

企业需要将自身的工作流程、领域知识和积累的判断转化为人工智能系统，并使其在每次使用中不断改进。私有评估应能捕捉模型是否真正针对对业务至关重要的结果（而不仅仅是外部基准！）进行了改进。私有强化学习环境应允许模型基于组织内部的真实数据不断成长。其知识库使机构记忆可查询，并提高了令牌的使用效率。

这个循环将成为公司新的知识产权。我把它比作一台爬山机器。与大多数资产不同，它具有复利效应。每一次工作流程的改进都会产生更好的训练信号，从而加速公司独有的隐性知识的积累。那些早期构建这一循环的公司将拥有难以复制的优势，无论其拥有何种新的单一模型能力。

我们最不希望看到的就是，所有行业、所有公司都将价值拱手让给少数几个攫取一切的模型。如果所有价值都集中在少数几个模型手中，政治经济体系绝对无法容忍。社会绝不会允许人工智能的未来掏空整个行业。

想想全球化第一阶段发生了什么，外包掏空了整个工业经济体。表面上看，GDP数据看起来不错，但产业转移是真实存在的，其后果至今仍在显现。我们绝不能让这种模式重演到人工智能时代，让少数人工智能系统攫取所有经济利益，而整个行业却眼睁睁地看着自己的知识被商品化，最终被彻底摧毁。

我认为，我们的首要任务必须是构建一个前沿生态系统，而不仅仅是一个前沿模式，这样价值才能广泛地流遍每家公司、每个行业和每个国家。在这个生态系统中，每个组织都能拥有编码其机构知识的学习循环，从而不断积累其人力资本和代币资本。

我从小就秉持着这样的理念：平台能够创造比平台本身所能提供的价值更大的额外价值，并且每家公司都可以不断创新，创造属于自己的价值。

当这种情况发生时，企业不仅能为自身创造价值，还能为周边经济创造价值。员工的专业知识将得到提升，他们的判断力将融入到可复制、可扩展的系统中，而企业和周边社区也将从中受益。

这就是企业如何为自身和更广泛的经济创造价值的方式。而这正是我们应该共同构建的稳定平衡。

Satya Nadella Full Original Text: A frontier without an ecosystem is not stable

Posted June 14, 2026 on X (@satyanadella)

A frontier without an ecosystem is not stable

I’ve been thinking a lot about the future of the firm in an AI-driven economy.

This transition is different than any previous platform shift. In the past, we used digital systems to enhance human capital. This is the first time we can create a real cognitive loop between people and digital systems. That is a mind-bender, because it changes how we even conceptualize work inside an enterprise.

What is at stake is not some digital tool or system and its use, but how organizations continue to learn, build IP, differentiate, and thrive in a world where AI models can continuously absorb the expertise of humans and organizations and commoditize it.

Every company is going to have to build what I think of as human capital and token capital. Human capital comprises the knowledge, judgment, relationships, ingenuity, and pattern recognition of its people, while token capital is the firm’s AI capability it builds and owns.

Importantly, human capital does not become less valuable as token capital grows. It only becomes more valuable! I believe human agency will be the driver of token capital growth. Humans will set ambitious goals, connect dots across disparate domains, and identify novel patterns that no model can surface on its own. Without human direction, compute merely spins its wheels. Without proprietary institutional knowledge embedded in its own token capital, even the most capable frontier model remains an external commodity that erodes a firm’s competitive differentiation over time.

The last thing any of us want is a world where every company across every sector is ceding value to a few models that eat everything they see. If all the economic returns accrue to only a small handful of general-purpose foundation models, the political economy will simply not tolerate it. There is no societal permission for an AI future that hollows out entire industries.

Think back to the first wave of globalization. Outsourcing lifted aggregate GDP numbers, but it hollowed out industrial ecosystems, displaced workers, and left lasting social and political scars that we still confront today. Let us not bring that same dynamic into the AI era, with a tiny number of AI systems capturing all the economic rents, while entire industries watch their hard-earned institutional knowledge get commoditized right out from underneath them.

In my view, our collective priority has to be building a frontier ecosystem, not just a frontier model — one where value flows broadly across every company, every industry, and every country. An ecosystem where every single organization can own its own proprietary learning loop that encodes its unique institutional knowledge, continuously compounding both its human capital and its token capital together.

This is the core platform ethos I have built Microsoft around over the past decade: great platforms enable far more value to be created on top of them than they capture for themselves. That means every company, from startups to multinational enterprises, gets to continuously innovate, build proprietary differentiation, and retain the economic value of its unique domain expertise — regardless of which underlying foundation model powers its internal learning loops.

Companies need to build agentic systems anchored to their own private evaluation datasets, private reinforcement learning environments, and persistent, queryable proprietary knowledge bases. These internal learning loops turn tacit, lived organizational experience into reusable, compounding token capital that cannot be replicated by any off-the-shelf general model.

These learning loops are portable, too. They should be architected to swap out underlying foundation models as the technology evolves, rather than locking an organization into a single model provider in perpetuity. True long-term resilience comes from the learning system a company owns, not temporary dependence on any single frontier model.

This model of distributed, compounding value creation is the only sustainable path forward for an AI-powered global economy. It ensures that innovation and prosperity do not concentrate into a narrow set of technology players, but instead disperse widely across every sector, every geography, and every business of every size. That is how companies create durable value for themselves and for the broader global economy. And it is the stable, inclusive equilibrium we all must work together to build.