Absurdism in the AI Era AI时代的荒诞主义

Essay Series / Bilingual Blog 随笔专栏 / 中英双语

Absurdism in the AI Era AI时代的荒诞主义

For people trying to stay lucid inside automation's contradictions. 写给想在自动化的矛盾里保持清醒的人。

Capability expands 能力扩张 Contradiction stays 矛盾仍在 Judgment returns 判断回归

This section does not ask only what AI can do in the abstract. It asks why, when AI can do a little of everything, work can start feeling more contradictory rather than more coherent. The absurd here is not comedy. It is mismatch: capability expands, activity multiplies, yet meaning and orientation refuse to stay aligned.

这个板块写的不是“AI 能做什么”这种抽象问题,而是当 AI 什么都能做一点时,工作为什么反而会变得更矛盾而不是更清楚。 这里的“荒诞主义”不是搞笑,也不是姿态,而是一种错位:能力在扩张,动作在增殖,但意义和方向却始终对不齐。

I am starting with two open-source skills I built myself. One is about subagent delegation discipline. One is about paper revision as structured sprints. On the surface they are workflow tools. Underneath, they deal with the same question: how to face contradiction without faking resolution, and how to push AI-era busyness back toward work that still contains judgment.

我先用两个自己做的 open-source skill 开这个头。一个写给子代理拆分,一个写给论文打磨。它们表面上属于 workflow tooling,底层上却都在处理同一个问题:怎样在矛盾没有被假装解决之前,继续保持清醒;怎样在 AI 时代把热闹 重新压回真正有判断的工作。

Two openings: delegation, writing, and the cost of meaning. 先从两个问题开始:拆分、写作、以及意义成本。

2026-03-24 · Essay 01

Defaulting to One Agent Is Not Conservative. It Defends the Cost of Meaning. 默认单代理,不是保守,而是对意义成本的防守

Built from token-efficient-subagent-decomposition. The essay is not about making multi-agent systems louder. It is about forcing delegation back under the rule of critical paths, clear boundaries, and coordination cost.

token-efficient-subagent-decomposition 说起。它讨论的不是怎么让多代理更热闹,而是怎么让拆分这件事 重新服从关键路径、边界和协调成本。

Read Essay 阅读文章 GitHub Repo

2026-03-24 · Essay 02

PaperSprint: When Writing Speeds Up, Final Responsibility Becomes Scarcer. PaperSprint:当论文写作被 AI 加速,真正稀缺的是节奏与人的最终责任

Built from PaperSprint. It turns paper revision into sprints, but its deeper point is that no submission decision should ever leave the human author behind.

PaperSprint 说起。它把论文修改转成 sprint,但更重要的是,它坚持任何提交决定都必须回到人的最终判断。

Read Essay 阅读文章 GitHub Repo

Why call the section “Absurdism in the AI Era”? 为什么叫“AI时代的荒诞主义”

01 · Capability / Meaning Gap

Capability and meaning stop moving in sync. 能力和意义不再同步增长。

In AI workflows, systems often become more capable faster than teams become more interpretive. You can add agents, prompts, drafts, and intermediate artifacts, while the underlying situation remains just as murky.

在 AI workflow 里,系统能力的增长速度,常常快过团队解释问题的速度。你可以增加 agent、prompt、draft 和各种中间产物, 但底层处境依然一样模糊。

02 · Contradiction Persists

Stronger tools do not dissolve contradiction for you. 工具更强,并不会替你把矛盾自动消掉。

The absurd condition begins when teams secretly expect output growth to remove ambiguity. It does not. The contradiction remains, only now it is wrapped in cleaner interfaces and faster cycles.

真正的荒诞,常常开始于团队偷偷期待“产出变多”就能把模糊性一起带走。但它不会。矛盾还在,只是换成了更漂亮的界面和更快的节奏。

03 · Lucid Action

Absurdism asks for lucid action, not false resolution. 荒诞主义要求的是清醒行动,而不是假装问题已经被解决。

That is why these essays stay close to judgment, sequencing, and final responsibility. The goal is not to erase contradiction. It is to work inside it without handing over the last meaningful decision.

所以这些文章会不断回到判断、节奏和最终责任。目标不是把矛盾一键抹平,而是在矛盾还存在的时候,仍然不把最后那个重要决定交出去。

If these essays touch a live problem you are facing, email me directly. 如果这些文章刚好碰到你的真实问题,可以直接联系我。

Treat this as a third entry point beside the book and the author page. If you want to talk about AI diagnosis, workflow design, writing process, or open-source skills, email is the cleanest path.

你可以把这里当作书页和作者页之外的第三个入口。如果你想围绕 AI diagnosis、workflow design、writing process 或 open-source skill 交流,直接发邮件最有效。