|
← 判断力与美学 ← Judgment & Aesthetics
SAE 判断力与美学 · 余项之美
SAE Judgment & Aesthetics · Beauty of the Remainder
2026-05-09

Soup of Life:当作者退场后,进化本身成为余项

Soup of Life: When the Author Exits, Evolution Itself Becomes the Remainder

Han Qin (秦汉)

Soup of Life 是一个持续运行的人工生命模拟器。没有目标,没有玩法,没有优化循环——打开网站,你看到的是一个自主运行的数字生态系统。生物体在里面竞争资源、繁殖、适应、灭绝,整个过程不接受任何外部干预。创造者刻意移除了所有游戏目标和玩家操控,把用户的角色从"参与者"降格为"观察者"。你能做的只是观看、检查个体、导出"生命卡片"来记录值得注意的生物。世界一旦启动,不可引导。模拟已经运行了数亿个时间刻度,由一个分布式的公众群体持续观测。

这是余项之美的最纯粹形式之一。进化本身就是一个凿构循环:突变是凿子,适应是构,存活下来的生物体是凿构循环在那个时刻的沉积物——余项。但还有更深一层:创造者的退场。设定规则之后,作者离开了。作品变成一台无作者的余项生产机器。Soup of Life里没有任何美的东西是被设计为美的。捕食者与猎物的军备竞赛、寄生策略、合作共生、资源的繁荣与崩溃、灭绝级联——这些全部是局部规则与全局约束碰撞后涌现出来的,是计算的余项,不是计算的目标。

让这个系统成为结构性活体而非仅仅聪明的关键,是智能的能量代价机制。在Soup of Life里,认知消耗能量。更聪明的生物体燃烧更多资源。这制造了一种反复出现的振荡:在稳定环境中,简单策略通过效率碾压复杂策略;当条件突变时,智能重新获得优势。这种张力产生了自己的时间节律——不是叙事弧线,不是随机噪声,而是余项在崩溃与恢复之间循环的脉搏。

现在是2026年5月,Soup of Life恰好处在结构性活力的精确位置。今年三月,国际人工生命学会(ISAL)的百科全书为它创建了词条;Hacker News上出现了讨论帖;一个Instagram社区开始分享生物卡片。但艺术界对它没有名字。它不是"游戏艺术"(没有游戏),不是"生成艺术"(输出不是图像),不是"互动艺术"(用户不互动),不是"计算生物学"(没有假设被检验)。它落在每个可用类别之间——余项恰恰就住在那里。一旦它被命名——一旦它变成"数字生态学平台"或"开放式进化工具"——余项就会硬化为构。此刻,它的逻辑还在生长。

soupof.life ↗

Soup of Life is a continuously running artificial life simulation. There are no goals, no player controls, no optimization loops. You open the site and find a digital ecosystem already in motion — organisms competing, reproducing, adapting, going extinct. The creator deliberately stripped away every mechanism of player agency. Once a world begins, it unfolds without steering. Users observe, inspect individuals, export "life cards" to document notable organisms. That's it. The simulation has run for hundreds of millions of ticks, operated by a distributed public audience.

This is one of the purest structural expressions of the beauty of the remainder. Evolution itself is a chisel-construct cycle: mutation is the chisel, adaptation is the construct, and what survives at any given moment is the sediment — the remainder. But there is a second, deeper layer: the creator's withdrawal. After setting the rules, the author exits. The work becomes an authorless remainder-production machine. Nothing beautiful in Soup of Life was designed to be beautiful. The predator-prey arms races, the parasitic strategies, the cooperative symbioses, the boom-bust resource cycles, the extinction cascades — all emerged from local rules colliding with global constraints. They are the remainder of computation, not the goal.

What makes this structurally alive rather than merely clever is the intelligence cost mechanic. Cognition in Soup of Life consumes energy. Smarter organisms burn more resources. This creates a recurring oscillation: in stable environments, simple strategies crush complex ones through sheer efficiency; when conditions shift, intelligence regains its advantage. This tension generates its own temporal rhythm — not narrative arc, not random noise, but the pulse of remainder cycling between collapse and recovery.

Right now, in May 2026, Soup of Life occupies a precise position of structural aliveness. The International Society for Artificial Life (ISAL) added an encyclopedia entry for it in March. A Hacker News thread surfaced it to a wider audience. An Instagram community has begun sharing organism cards. But the art world has no name for it. It is not "game art" (there is no game). It is not "generative art" (the output isn't images). It is not "interactive art" (users don't interact). It is not "computational biology" (there is no hypothesis being tested). It falls between every available category, which is exactly where remainder lives. Once it gets named — once it becomes "a digital ecology platform" or "an open-ended evolution tool" — the remainder will have hardened into construct. For now, its logic is still growing.

soupof.life ↗