A Thought Leader’s Perspective: Power and Responsibility | 思想领袖的洞见：权力与责任
Navigation
< Previous Chapter | Home | Next Chapter >
Author
Marco Argenti, CIO, Goldman Sachs | Marco Argenti，高盛首席信息官（CIO）
Abstract
This thought leadership chapter (from the Agentic Design Patterns series) reflects on an unprecedented AI paradigm shift: moving from 18 months of rapid Large Language Model (LLM) “engine” advancement to the upcoming era of building agentic “car” frameworks that convert raw LLM power into actionable, adaptive intelligence. Authored by Goldman Sachs CIO Marco Argenti, it addresses the reasoning model inflection point, unique value of agentic systems, high-stakes risk considerations (especially for finance), foundational infrastructure requirements, and the core mission of augmenting (not replacing) human ingenuity.
Key Takeaways
- AI Paradigm Metaphor: Past 18 months focused on LLM (engine) capability gains; the next era will center on agentic frameworks (the “car” built around the engine) to turn text generation into actionable agentic intelligence.
- Reasoning Model Inflection: “Reasoning” models marked a step-change: AI evolved from statistical next-word prediction to nascent cognition capable of generating plans, not just content.
- Agentic vs. Rule-Based Automation: Unlike static rule-based automata (which break on unexpected obstacles), reasoning-powered AI agents can observe, adapt, and navigate edge cases—shifting instruction from “what to do” to “why to do it, trusting it to figure out how”.
- High-Stakes Accountability: In sectors like finance, agent errors carry catastrophic risks; blind trust is forbidden, and transparency/accountability are mandatory.
- Foundational Infrastructure Requirement: “Messy systems + agents = disaster”; clean data, consistent metadata, well-defined APIs, and a programmable enterprise (enterprise as software) are non-negotiable prerequisites.
- Human-AI Collaboration Model: Agentic AI augments (does not replace) human ingenuity, requiring new skills: clear task articulation, wise delegation, and diligent output verification.
Evidence Snippets (Source Excerpts)
Paradigm Shift Metaphor
If the last eighteen months were about the engine -the breathtaking, almost vertical ascent of Large Language Models (LLMs)- the next era will be about the car we build around it. It will be about the frameworks that harness this raw power, transforming it from a generator of plausible text into a true agent of action. 如果说过去的十八个月是关于「引擎」的故事——即大语言模型那惊人的、近乎垂直的飞跃——那么下一个时代将是关于我们如何围绕它造出一辆「汽车」。这个时代，将关乎我们如何构建框架来驾驭这股原始的力量，把它从能生成看似合理文本的工具，打造成真正能付诸行动的智能体。
Reasoning Model Inflection
But then came the inflection point, a step-change brought about by a new class of “reasoning” models. Suddenly, we weren’t just conversing with a statistical machine that predicted the next word in a sequence; we were getting a peek into a nascent form of cognition. 然而，转折点随之而来——推理模型的出现，实现了一次质的飞跃。那一刻，我们对话的对象不再仅仅是那个预测词语的统计机器；我们所窥见的，是一种正在萌芽的全新认知。
Agentic Framework Differentiation
This is the promise of agentic frameworks. It’s the difference between a static subway map and a dynamic GPS that reroutes you in real-time. A classic rules-based automaton follows a fixed path; when it encounters an unexpected obstacle, it breaks. An AI agent, powered by a reasoning model, has the potential to observe, adapt, and find another way. 这便是智能体框架所带来的希望。它就像一张静态的地铁线路图与一个能为你实时重新规划路线的动态 GPS 之间的区别。一个经典的、基于规则的自动程序遵循固定的路径，当遇到意外障碍时，它就会崩溃。而一个由推理模型驱动的 AI 智能体，则有潜力去观察、适应并找到另一条路。
High-Stakes Risk Example
I’ve read the disclaimers and the cautionary tales: the web automation agent that, after failing a login, decided to email a member of parliament to complain about login walls. It’s a darkly humorous reminder that we are dealing with a technology we don’t fully understand. 我读过那些免责声明和警示故事：一个网络自动化智能体在登录失败后，竟然决定给一位国会议员发邮件抱怨登录墙。这是一个黑色幽默般的提醒：我们正在打交道的，是一项我们尚未完全理解的技术。
Programmable Enterprise Foundation
Messy systems plus agents are a recipe for disaster. An AI trained on “garbage” data doesn’t just produce garbage-in-garbage-out; it produces plausible, confident garbage that can poison an entire process. Therefore, our first and most critical task is to prepare the ground. We must invest in clean data, consistent metadata, and well-defined APIs. We have to build the modern “interstate system” that allows these agents to operate safely and at high velocity. It is the hard, foundational work of building a programmable enterprise, an “enterprise as software,” where our processes are as well-architected as our code. 混乱的系统加上智能体，只会酿成灾难。一个用垃圾数据训练出来的 AI，不仅会产生垃圾结果，它还会产生貌似可信、充满自信的垃圾，足以毒化整个流程。因此，我们首要且最关键的任务，是打好基础。我们必须投资于干净的数据、一致的元数据和定义良好的 API。我们必须建立起现代化的「州际高速公路系统」，让这些智能体能够安全、高速地运行。这是一项艰巨的基础性工作，其目的就是构建一个可编程的企业——即实现「企业即软件」的理念，从而让我们的业务流程也能像代码一样，拥有精良的架构。
Closing Message
The world is asking every engineer to step up. I am confident we are ready for the challenge. Enjoy the journey. 这个世界正在召唤每一位工程师挺身而出。我相信，我们已经准备好迎接挑战。愿你享受这段旅程。