---
author: AI Engineer
date: '2026-10-16'
guest: ''
layout: post.njk
source: https://www.youtube.com/watch?v=Fva8hFosqyQ
speaker: AI Engineer
tags:
- agentic-loop
- software-factory
- collective-intelligence
title: 循环工艺与最高循环：AI工程师的循环升降法则与群体智能
summary: 探讨AI工程师的核心能力——在不同抽象层级的“循环”之间上下穿梭与调试。从最底层的Token、工具调用，到自动化Cron任务与软件工厂，AI能力正扩散至医疗、金融等各垂直领域。而人类汇聚、定义下一个循环的峰会与协作，则是驱动群体智能演进的“最高循环”。
insight: ''
draft: true
series: ''
category: ai-application
area: tech-engineering
project: []
people: []
companies_orgs:
- AI Engineer
products_models: []
media_books:
- Loopcraft
status: evergreen
---
### 循环阶梯：AI工程的核心抽象与穿梭能力
最初只有**Token**，随后出现了**对话**（Chat），接着我们学会了**使用工具**（Tool Use），再后来学会了**设定目标**（Goal Setting）。而当下，我们全面进入了**自动化**（Automations）、**定时任务**（Cron Jobs）与**执行循环**（Loops）的时代。在软件工厂的构建中，存在着大量的**循环堆叠**（Loop Stacking）。正如在《**Loopcraft**》文章中所探讨的，**AI工程师**（AI Engineer: 构建和应用AI系统的工程实践者）的核心技能，本质上就是清晰界定你当前所处的**循环层级**（Loop Level），并评估其抽象程度是否恰当：当你需要提升生产力、扩展推理规模时，你需要向上一层循环跃迁；而当你遭遇可靠性瓶颈、需要修复系统缺陷时，你又必须下沉到更低层级的循环中去深挖细节。
如果用一个程序员更易理解的类比，这就像是一个由 **While 循环组成的堆栈**（Stack of While Loops）。无论是工程师、团队领袖、研究员还是创始人，都需要具备像通用智能体一样的递归穿梭能力，能够在这些多层级的循环之间自如上升与下沉。面对海量且快速迭代的技术内容，关键在于清晰定位自己当前在哪个循环层级工作，以及当前受阻于哪一个层级。
Original English Source
In the beginning, there was the token. Then there was the chat. Then we learned to use tools. Then we learned to set goals. I'm skipping a few steps. And these days, we're all about automations. We're all about cron jobs and the loops and the rest of today on this live stream. Those of you watching at home, we're going to be talking about software factories as well. So there's a lot of loop stacking.
I wrote this essay called Loopcraft. I think Joel in the audience did a workshop on it yesterday. Can you still hear me? Are we good? Sorry, I had a mic slip. And I think this is something that is basically a core skill of AI engineer is defining what loops you're working in and figuring out whether that's the right level of abstraction and when you're ready going up a loop in order to figure out the next level in productivity and scaling your inference and when you're finding issues going down a loop and going lower level in order to figure out reliability and fixing whatever bugs you find.
If you want to simplify it, I have a better analogy that will be less complicated. It's literally just a stack of while loops, right? I think all of you in the audience and programmers will understand this. And effectively, this is what we're here to do today. I expect our engineers, our leaders, our researchers, our founders to all be understanding and recursing as a sort of human general intelligence all these loops that you guys are capable of going up and down. And that is why there's so much content. I often tell people who complain about the tracks, I'm not going to apologize for good content. It's up to you to decide and route yourself as to which level you currently working at and which level you're currently blocked on.
### 从生物节律到社会协作：循环概念的跨界映射
人类自身的生命与协作历程同样呈现出循环的结构。最初是在母体中的**心跳**（Heartbeat），那是生命的印记；随后人类学会说话，学会使用工具，然后开始工作，陷入“吃饭、睡觉、写代码/写Prompt、重复”的循环。进一步发展后，人类进入团队协作，以小团队和家庭为单位运作。**微型团队**（Tiny Teams）的概念不仅在当下愈发关键，更延伸到人类文明更宏大的共同愿景——由**群体智能**（Collective Intelligence: 多智能体或人群协同涌现的整体认知能力）所驱动的文明演进。
我们在代码工程中所实践的循环范式，正深刻地映射并跨越到人类社会的组织运作中。智能体在单一垂直领域的成功经验正迅速泛化，AI技术正全面扩散至编程之外的各个领域，包括**医疗健康**（Healthcare）、**金融**（Finance）、**市场拓展**（GTM: Go-To-Market 商业化落地）以及**交付工程**（Deployed Engineering）。这种从代码到更广泛人类生产力领域的渗透，标志着AI技术扩散进入了全新阶段。
Original English Source
So the thing that's occupying most of my time recently and what I make the human analogy for is what is the human loop? And in the beginning there's the heartbeat. When you're sort of in the womb you sort of get that heartbeat that is the proof of life. Then you learn to talk. Then we learn to use tools. Then we start to work. We eat, sleep, code, repeat or prompt. And then we start working in teams. We start working in small groups and families. We've promoted the concept of tiny teams before and I think that is increasingly important with everybody.
And I think the interesting thing to think about is like the broader human projects as a whole. Like where are we all going? You know, extend out from your personal context, your family, your country, your company, whatever, to the broader project of the human civilization and what we're building with collective intelligence. I think it's very interesting to consider that. So I think that's a little bit what I want to leave you with today. Like the things that you're doing with coding are definitely crossing over to human life.
The message of my last keynote at AI Engineer was that it's very much like agents that you see are working in one domain are definitely generalizing this year. This is the year that we are adding vertical tracks like AI in healthcare, AI in finance, AI in GTM. We have a deployed engineering track today which I'm very excited about. All of this is meant to see how AI is just diffusing finally beyond coding.
### 最高循环：聚合人类智慧以定义未来循环
从2023年首届**AI Engineer**大会到2024年约2000人规模的**World's Fair**，再到如今突破7000名参与者，社群规模经历了迅猛的增长。举办峰会以及所有人齐聚一堂的根本价值，正在于构建**最高循环**（The Highest Loop）——由人类汇聚在一起，共同探讨并定义“下一个循环应该是什么”，即**创造循环的循环**（The loop that makes loops）。
为了将这种高阶探索推向全球，AI Engineer 在纽约、巴黎、新加坡、墨尔本及迈阿密等世界各地组织峰会，并与本地伙伴紧密协同，而 World's Fair 则是所有峰会之上的峰会。这种协作旨在让每位参与者探索并确立专属于自己的**循环生成机制**（Generative Process for Generating Loops），并将其自顶向下传递到团队组织、个人工作流以及日常生活之中。
Original English Source
Last world fair I gave you all an update on the stress curves of the organizers. Just to give you an idea, this is the first AI engineer in 2023. This is sort of the sales curves basically of like how people bought. Basically it was kind of linear. Then we had the first world fair 2024, about 2,000 people. 2025 was I think I skipped a few lines, but we're basically at 6,000. Sorry, we aimed for 6,000. We actually crossed and went to 7,000 today. And basically I asked you all to buy your tickets earlier. You did not. If you measure the Gini coefficient, it has gone up and not down. That's the wrong direction. So if you all can just gather yourselves and join us earlier, I think that would be great.
Anyway, why we want to work at a conference, why I work on a conference, why you all come to a conference is basically the concept of the highest loop. I will submit to you that the highest loop of all is to, if I can make this slide work, oh my god, I'm gonna refresh. Okay, that the highest loop is where humans come together to figure out what the next loop is, right? The loop that makes loops.
So we organize AI engineer summits all around the world. This year we have summits in New York, summits in Paris, summits in Singapore. And again Melbourne and I'm forgetting one already, Miami. And we organized this as partners. We want you to organize it in your home countries. We're hiring a partnerships team to do this. And this is the highest loop of all which is the world's fair which is the summit of summits.
So I'm very excited to welcome all of you today. I hope that you figure out your generative process for generating loops and propagate it down to your teams, to your personal workflows, to your lives. Thank you very much and enjoy the show.