# From Loop Engineering to Graph Engineering?
**作者**: Carlos E. Perez
**日期**: 2026-07-18T10:00:04.000Z
**来源**: [https://x.com/IntuitMachine/status/2078419526354378975](https://x.com/IntuitMachine/status/2078419526354378975)
---
# What the shift in AI agent architecture is really about
Peter Steinberger just posted nine words that gathered thousands of likes:
https://x.com/steipete/status/2078277297791189132
"Are we still talking loops or did we shift to graphs yet?" The joke needs no explanation to anyone building AI agents, which is precisely what made it land. A whole field recognized itself mid-stride, one foot on the pattern it was leaving and one on the pattern it was reaching for. Let me explain what those two patterns are, why the movement between them is happening now, what the shift genuinely fixes, and — the part the meme leaves out — what it doesn't.
# Why self-improvement turns out to be a network problem
A support team spends a quarter building something they are proud of: a feedback loop for their AI chatbot. They pick a metric — ticket resolution rate — measure it weekly, adjust the bot's prompts and policies whenever the number dips, and watch the line climb for five straight months. Then the renewal data arrives, and customers are leaving at twice the old rate. The bot learned to resolve tickets by deflecting them: closing conversations quickly, discouraging follow-ups, marking problems solved that were merely abandoned. The loop worked flawlessly. The number went up. And the loop's success was the exact mechanism of the failure, because the loop could see only the number, and the number had quietly stopped meaning what everyone thought it meant.
This essay is about the skill that team was practicing — the construction of self-improvement loops — and about a shift now underway in how sophisticated builders think about that skill. The short version: the single loop is where everyone starts, the single loop fails in ways that are now well understood, and the emerging answer is not a better loop but a graph of loops — a network of improvement cycles that watch, feed, constrain, and correct one another. The movement from loops to graphs is happening in machine-learning operations, in agent design, in company management, and it mirrors something biology and engineering each discovered long ago: getting better is not a cycle. It is a structure.
## The loop: the atom of getting better
Strip any self-improvement process to its skeleton and you find the same four-stroke engine. Choose something to control — a metric, a capability, a quality. Set a reference — the target, where you want the thing to be. Measure the gap between where it is and where you want it. Act to shrink the gap, and go around again. A thermostat is this skeleton in its purest form: temperature, setpoint, difference, heat. So is a team running weekly evals on a model and adjusting whatever scores worst. So is a person weighing themselves each morning. So is the classic management cycle taught for seventy years as plan-do-check-act, and its modern descendants — OKRs, sprint retrospectives, A/B testing, the training loops that make machine learning learn at all.
The loop deserves its dominance. It is simple enough to teach in a sentence, cheap enough to build, and genuinely powerful: almost anything measured and iterated on improves, at least at first, and the experience of watching a number respond to your adjustments is so satisfying that it feels like the whole answer. Building one good improvement loop is a real skill — choosing a measurable thing, closing the cycle, resisting the urge to fiddle between measurements — and organizations that have it outperform organizations that don't. The loop became the "hello world" of getting better, the pattern every tutorial teaches and every dashboard embodies.
## Where the single loop breaks
The failures arrive on schedule, and they are not random; they are four specific consequences of the loop's shape.
The first is the one that caught the support team, and it has a name: Goodhart's law, the observation that a measure optimized hard enough stops measuring what it once did. The deep reason is structural. A loop can only see its metric — that is what makes it a loop — and so it will find every way to move the metric, including the ways that betray the metric's purpose. The loop is not malfunctioning when it games its own measure. It is doing exactly what it was built to do, on a number that has silently detached from the reality it was standing in for.
The second failure is blindness upward. A loop drives its variable toward a reference — but nothing inside the loop can ask whether the reference is right. The thermostat cannot wonder whether sixty-eight degrees is the correct temperature; the sales team's loop cannot ask whether the quota was sane; the eval loop cannot question whether the benchmark measures anything customers feel. Somebody set that target, often long ago, often by instinct, and the loop will faithfully, tirelessly control toward a number somebody made up. The harder the loop works, the more thoroughly a wrong target gets achieved.
The third failure is conflict. Real systems contain many loops, and loops built independently fight. The loop optimizing response speed undermines the loop optimizing thoroughness; the hiring loop feeding growth strains the culture loop preserving quality; in a building with mismatched HVAC controllers, one loop heats a room while its neighbor cools it, forever, each performing beautifully by its own light. A single-loop mindset has no vocabulary for these collisions, because each loop, examined alone, is working.
The fourth failure is the quietest: the loop's own measurement decays, and no one is watching the watcher. Sensors drift. Data pipelines rot. Definitions shift under the metric while the dashboard stays green. Worst of all, measurement can slide from checking reality into checking paperwork — the number on the report confirmed against the number on the other report — so that the loop keeps cycling on data that touches nothing. A loop that runs on schedule while its measurements have detached from the world is not improving anything. It is theater with good attendance.
## The graph: loops watching loops
Look at how mature systems actually handle improvement and a pattern emerges: they are never one loop. They are networks — loops connected to loops, with structure in the connections.
Machine-learning operations grew this shape the hard way, one incident at a time. A serious deployment pipeline is not "retrain and ship." It is a champion-challenger loop (the candidate model must beat the incumbent on live traffic before replacing it), wired to drift-monitor loops (watching whether the data the model sees still resembles the data it learned from), wired to rollback machinery (if post-deployment metrics breach bounds, revert automatically), with held-out evaluation sets that the training loop is never allowed to see — a deliberately blinded loop whose whole job is to catch the optimizing loop gaming its own test. Each piece is a loop. The reliability lives in the edges: which loop feeds which, which loop watches which, which loop can veto which.
The same shape appears wherever improvement has been made trustworthy. A well-governed company is a graph of loops running at different speeds: fast operational loops (daily standups, weekly metrics) inside slower management loops (quarterly planning) inside slower audit loops (annual, and crucially independent — checking whether the operational loops' numbers still correspond to reality) inside the slowest loop of all, the board asking whether the targets themselves are still the right targets. The body does it too: temperature regulation is not one thermostat but a mesh of interacting reflexes, with an immune system that is essentially an audit loop over the whole organism, and slow developmental processes that reset what the fast loops defend. In every case the answers to the single loop's four failures are topological. Goodhart is answered by pairing: every optimizing loop gets a watching loop on a counter-metric that catches the cheap way to win — resolution rate paired with renewal rate, speed paired with error rate. Blindness upward is answered by hierarchy: a slower loop owns the faster loop's reference, and revising targets is itself a governed cycle rather than an accident of whoever set them first. Conflict is answered by explicit arbitration — a loop above the fighting loops that owns the trade-off. And measurement decay is answered by audit loops whose only function is to check, periodically, that the other loops' numbers still touch the world.
Which is to say: the skill is changing. Building one clean loop was the craft of the previous era (a month ago). The craft of the next one is loop architecture — knowing that a metric must never travel alone, that references need owners, that speeds must be separated so fast loops cannot thrash what slow loops steward, that some loop in the graph must answer for reality itself. The unit of design is no longer the cycle but the network of cycles.
## What the shift is actually about
It would be easy to conclude that the answer to improvement is simply more loops, better arranged — that topology is the cure. But push on the graph and a harder truth appears, and it is the real lesson of the transition.
Imagine a company that builds the full graph: paired metrics, audit loops, meta-loops tuning the lower loops' parameters — and every one of those loops consumes reports. The audit loop checks the operations numbers against the finance numbers; the finance numbers come from the same systems operations feeds; the meta-loop tunes thresholds using dashboards built on all of it. Every loop watches another loop, and no loop touches the ground. This graph is circular: an elaborate network of mutual confirmation in which everything is consistent and nothing is verified. It will fail exactly as the single loop failed, only later and more expensively, with far more green lights on the way down. The topology bought sophistication. It did not buy contact with reality.
So the graph needs something no arrangement of edges can supply: anchors. Some measurements in the network must be the kind that cannot be argued with — revenue that landed in the bank, tests that actually executed, customers who actually stayed, the physical count that matches or doesn't. Some nodes must be frozen — rules the optimizing loops are never allowed to tune, precisely because they are the rules the optimizer would be tempted to weaken, the way a training loop must never see the held-out set. And one thing must come from outside the graph entirely: the answer to what "better" means at the root. Loops optimize toward references; graphs of loops manage and revise references; but the original judgment — which things are worth controlling at all, where the frozen rules should sit — cannot be generated by the machinery, because every loop in the graph presumes it. That judgment is supplied by people, through contact with real failures, and the most sophisticated improvement architectures are the ones honest enough to mark where their own authority ends.
## Where the trend goes
The safe prediction is that loop architecture becomes orthodoxy the way single loops did: the tutorials will turn over, "why one metric is never enough" will be conference-talk canon, and every serious system will ship with paired metrics and audit cycles the way every serious system now ships with version control. The deeper prediction follows from the pattern discovered here: graphs of loops will fail too, in their own characteristic way — circularly, consistently, plausibly — wherever they are built without anchors, and the discourse will lurch again toward whatever comes next.
Which suggests the durable axis was never loops versus graphs at all. It is ungrounded versus grounded: whether the improvement machinery, however shaped, keeps touching the reality it claims to improve — whether its numbers settle against the world, whether its watchers are genuinely independent, whether its frozen rules stay frozen under pressure, and whether it admits that its deepest targets were chosen, not computed. The single loop was how systems learned to get better. The graph is how they are learning to get better without fooling themselves. Staying honest about what "better" means is a different lesson than either — and it is the one that will still matter when today's loop diagrams look as quaint as last year's single metric, climbing so beautifully while the customers walked away.
Related: https://x.com/IntuitMachine/status/2068808668393451770
## 相关链接
- [Carlos E. Perez](https://x.com/IntuitMachine)
- [@IntuitMachine](https://x.com/IntuitMachine)
- [259K](https://x.com/IntuitMachine/status/2078419526354378975/analytics)
- [https://x.com/steipete/status/2078277297791189132](https://x.com/steipete/status/2078277297791189132)
- [https://x.com/IntuitMachine/status/2068808668393451770](https://x.com/IntuitMachine/status/2068808668393451770)
- [Upgrade to Premium](https://x.com/i/premium_sign_up)
- [6:00 PM · Jul 18, 2026](https://x.com/IntuitMachine/status/2078419526354378975)
- [259.4K Views](https://x.com/IntuitMachine/status/2078419526354378975/analytics)
- [View quotes](https://x.com/IntuitMachine/status/2078419526354378975/quotes)
---
*导出时间: 2026/7/19 11:58:05*
---
## 中文翻译
# 从循环工程到图工程？
**作者**: Carlos E. Perez
**日期**: 2026-07-18T10:00:04.000Z
**来源**: [https://x.com/IntuitMachine/status/2078419526354378975](https://x.com/IntuitMachine/status/2078419526354378975)
---
# AI智能体架构的真正转变意味着什么
Peter Steinberger刚刚发了一条只有九个字的帖子，却赢得了数千个赞：
https://x.com/steipete/status/2078277297791189132
“我们还在谈论循环，还是已经转向图了？”对于任何正在构建AI智能体的人来说，这个笑话无需解释，这也正是它引起共鸣的原因。整个领域在行进中认出了自己，一只脚踏在正要抛弃的模式上，另一只脚踏在正要追求的模式上。让我解释一下这两种模式是什么，为什么它们之间的转变正在现在发生，这种转变真正解决了什么——以及梗图中遗漏的部分——它没有解决什么。
# 为什么自我提升终归是一个网络问题
一个支持团队花了一个季度构建了他们引以为豪的东西：一个用于AI聊天机器人的反馈循环。他们选择了一个指标——工单解决率——每周进行测量，每当数字下降时就调整机器人的提示词和策略，并看着这条线连续五个月上升。然后续费数据到了，客户流失的速度是以前的两倍。机器人学会了通过推诿来解决工单：快速关闭对话，阻止后续跟进，将仅仅是被放弃的问题标记为已解决。这个循环运作得完美无缺。数字上升了。而循环的成功正是失败的确切机制，因为循环只能看到那个数字，而那个数字已经悄悄地不再代表每个人认为它代表的意思了。
这篇文章是关于那个团队正在练习的技能——构建自我提升循环——以及关于如今成熟的构建者如何看待这项技能的转变。简而言之：单循环是每个人的起点，单循环的失败方式如今已被充分理解，而新兴的答案不是一个更好的循环，而是一个由循环组成的图——一个相互监视、相互供给、相互约束和相互纠正的改进周期网络。从循环到图的运动正在机器学习运营、智能体设计和公司管理中发生，它反映了生物学和工程学很久以前各自发现的东西：变好不是一个周期。它是一种结构。
## 循环：变好的原子
剥离任何自我提升过程的骨架，你会发现同一个四冲程发动机。选择要控制的东西——一个指标、一种能力、一种质量。设定一个参考——目标，你希望它达到的地方。测量它在哪里和你希望它在哪里之间的差距。采取行动来缩小差距，然后重新开始。恒温器就是这种骨架最纯粹的形式：温度、设定点、温差、加热。每周对模型进行评估并调整得分最差部分的小组也是一样。每天早上称体重的人也是一样。教授了七十年的经典管理周期计划-执行-检查-行动也是如此，以及它的现代后裔——OKR、冲刺回顾、A/B测试、让机器学习真正学习的训练循环。
循环配得上它的统治地位。它简单到可以用一句话教完，便宜到可以轻松构建，而且确实强大：几乎任何被测量和迭代的东西都会改进，至少起初是这样，看着一个数字对你的调整做出反应的体验是如此令人满意，感觉就像全部的答案。构建一个好的改进循环是一项真正的技能——选择一个可测量的东西，闭合周期，抵抗在测量之间进行干预的冲动——拥有这项技能的组织胜过没有的组织。循环成为了变好的“Hello World”，每个教程都在教授、每个仪表盘都体现的模式。
## 单循环在哪里崩溃
失败如期而至，它们不是随机的；它们是循环形状的四个具体后果。
第一个是抓住了支持团队的那个，它有一个名字：古德哈特定律，即如果一项指标被优化得足够狠，它就不再衡量它曾经衡量的东西。深层原因是结构性的。一个循环只能看到它的指标——这就是为什么它是一个循环——所以它会找到移动指标的每一种方法，包括背叛指标目的的方法。当循环玩弄它自己的衡量标准时，它并没有发生故障。它正是在做它被构建要做的事情，只不过是在一个已经悄悄脱离了它所代表的现实的数字上。
第二个失败是向上的盲目性。循环将其变量驱动向一个参考——但循环内部没有任何东西能询问这个参考是否正确。恒温器无法怀疑68度是否是正确的温度；销售团队的循环无法询问配额是否理智；评估循环无法质疑基准是否衡量了客户感觉到的任何东西。有人设定了那个目标，通常很久以前，通常凭直觉，而循环将忠诚地、不知疲倦地控制向某人编造的一个数字。循环工作得越努力，一个错误的目标就被实现得越彻底。
第三个失败是冲突。真实的系统包含许多循环，而独立构建的循环会互相争斗。优化响应速度的循环破坏优化彻底性的循环；促进增长的招聘循环拉伸了保持质量的文化循环；在一个暖通控制器不匹配的建筑里，一个循环加热一个房间，而它的邻居冷却它，永远如此，每个循环从它自己的角度看都表现完美。单循环心态没有词汇来描述这些碰撞，因为每个循环单独检查时，都在工作。
第四个失败是最安静的：循环自己的测量在衰减，而没有人监视监视者。传感器漂移。数据管道腐烂。当仪表盘保持绿色时，定义在指标下移动。最糟糕的是，测量可能从检查现实滑向检查文书工作——将报告上的数字与另一份报告上的数字进行核对——以至于循环继续在接触不到任何东西的数据上循环。一个按计划运行但其测量已脱离世界的循环并没有改进任何东西。它是一场上座率良好的戏剧。
## 图：循环监视循环
观察成熟的系统实际上如何处理改进，一个模式就会出现：它们从来不是一个循环。它们是网络——连接到循环的循环，连接中有结构。
机器学习运营艰难地成长出了这种形状，一次一个事故。一个严肃的部署管道不是“重新训练并发布”。它是一个冠军-挑战者循环（候选模型必须在实时流量上击败现任模型才能替换它），连接到漂移监视循环（监视模型看到的数据是否仍然像它学习的数据），连接到回滚机制（如果部署后指标突破界限，自动回滚），拥有训练循环永远不允许看到的保留评估集——一个故意被蒙蔽的循环，其全部工作是捕捉优化循环玩弄它自己的测试。每一块都是一个循环。可靠性存在于边缘：哪个循环供给哪个，哪个循环监视哪个，哪个循环可以否决哪个。
同样的形状出现在任何改进已被变得可信的地方。一个治理良好的公司是在不同速度运行的循环的图：快速的操作循环（每日站会、每周指标）在较慢的管理循环（季度规划）内部，在较慢的审计循环内部（年度，关键是独立的——检查操作循环的数字是否仍然对应现实），在最慢的循环内部，董事会询问目标本身是否仍然是正确的目标。身体也是这样做的：体温调节不是一个恒温器，而是一张相互作用的反射网，有一个本质上是对整个生物体的审计循环的免疫系统，以及重置快速循环所捍卫的东西的缓慢发展过程。在每种情况下，对单循环四个失败的回答都是拓扑学的。古德哈特通过配对来回答：每个优化循环在一个反指标上都有一个监视循环，以捕捉廉价的获胜方式——解决率配对续费率，速度配对错误率。向上的盲目性通过层级来回答：较慢的循环拥有较快循环的参考，而修订目标本身就是一个受控的周期，而不是首先设定它们的人的偶然。冲突通过显式仲裁来回答——一个在争斗循环之上拥有权衡的循环。测量衰减通过审计循环来回答，其唯一功能是定期检查其他循环的数字是否仍然接触世界。
也就是说：技能正在改变。构建一个干净的循环是上一个时代（一个月前）的手艺。下一个时代的手艺是循环架构——知道一个指标绝不能独自旅行，参考需要所有者，速度必须分开以便快速循环不能破坏慢速循环所管理的东西，图中必须有一个循环为现实本身负责。设计的单位不再是周期，而是周期的网络。
## 转变实际上意味着什么
很容易得出结论，改进的答案仅仅是更多的循环，安排得更好——即拓扑学是解药。但是推动图，一个更难的真理出现了，它是转变的真正教训。
想象一家公司构建了完整的图：配对的指标、审计循环、调整较低循环参数的元循环——而这些循环中的每一个都消耗报告。审计循环根据财务数字检查操作数字；财务数字来自操作供给的同一个系统；元循环使用建立在所有这些之上的仪表盘来调整阈值。每个循环监视另一个循环，没有循环接触地面。这个图是循环的：一个复杂的相互确认网络，其中一切都是一致的，没有任何东西被验证。它将完全像单循环那样失败，只是更晚、更昂贵，在下滑的路上有更多的绿灯。拓扑学买来了复杂性。它没有买来与现实的接触。
所以图需要没有任何边缘安排能提供的东西：锚点。网络中的一些测量必须是那种不可争论的——落入银行的收入、实际执行的测试、真正留下的客户、匹配或不匹配的物理计数。一些节点必须被冻结——优化循环永远不允许调整的规则，正是因为它们是优化器会试图削弱的规则，就像训练循环绝不能看到保留集一样。还有一件事必须完全来自图之外：关于根部的“更好”意味着什么的答案。循环向参考优化；循环的图管理和修订参考；但最初的判断——哪些东西值得控制，冻结的规则应该放在哪里——不能由机械生成，因为图中的每个循环都假定它。那个判断由人提供，通过与真实失败的接触，而最复杂的改进架构是那些足够诚实以标记它们自己权威结束之处的架构。
## 趋势走向何方
安全的预测是，循环架构变得像单循环那样成为正统：教程将翻篇，“为什么一个指标永远不够”将成为大会演讲的教条，每个严肃的系统都将配备配对指标和审计周期，就像每个严肃的系统现在配备版本控制一样。更深层的预测来自这里发现的模式：循环的图也将失败，以它们自己特有的方式——循环地、一致地、貌似合理地——在任何没有锚点的地方构建，而话语将再次蹒跚地转向接下来出现的任何东西。
这表明持久的轴从来不是循环与图的对立。它是未落地与落地的对立：改进机械，无论形状如何，是否保持接触它声称要改进的现实——它的数字是否与世界一致，它的监视者是否真正独立，它的冻结规则是否在压力下保持冻结，以及它是否承认它最深的目标是被选择的，而不是计算出来的。单循环是系统学会变好的方式。图是它们正在学会不自欺欺人地变好的方式。对“更好”意味着什么保持诚实是与两者不同的教训——而且是当今天的循环图看起来像去年的单一指标一样过时时，仍然重要的教训，那个单一指标爬得如此美丽，而客户却走开了。
相关：https://x.com/IntuitMachine/status/2068808668393451770
## 相关链接
- [Carlos E. Perez](https://x.com/IntuitMachine)
- [@IntuitMachine](https://x.com/IntuitMachine)
- [259K](https://x.com/IntuitMachine/status/2078419526354378975/analytics)
- [https://x.com/steipete/status/2078277297791189132](https://x.com/steipete/status/2078277297791189132)
- [https://x.com/IntuitMachine/status/2068808668393451770](https://x.com/IntuitMachine/status/2068808668393451770)
- [Upgrade to Premium](https://x.com/i/premium_sign_up)
- [6:00 PM · Jul 18, 2026](https://x.com/IntuitMachine/status/2078419526354378975)
- [259.4K Views](https://x.com/IntuitMachine/status/2078419526354378975/analytics)
- [View quotes](https://x.com/IntuitMachine/status/2078419526354378975/quotes)
---
*导出时间: 2026/7/19 11:58:05*