diff --git a/README-ja.md b/README-ja.md index 8717201b8..b033a5f3b 100644 --- a/README-ja.md +++ b/README-ja.md @@ -2,17 +2,17 @@ [English](./README.md) | [中文](./README-zh.md) | [日本語](./README-ja.md) -## モデルこそが Agent である +## Agency はモデルから生まれる。Agent プロダクト = モデル + Harness -コードの話をする前に、一つだけ明確にしておく。 +コードの話をする前に、一つ明確にしておく。 -**Agent とはモデルのことだ。フレームワークではない。プロンプトチェーンではない。ドラッグ&ドロップのワークフローではない。** +**Agency -- 知覚し、推論し、行動する能力 -- はモデルの訓練から生まれる。外部コードの編成からではない。** だが実際に動く Agent プロダクトには、モデルと Harness の両方が必要だ。モデルはドライバー、Harness は車。本リポジトリは車の作り方を教える。 -### Agent とは何か +### Agency はどこから来るか -Agent とはニューラルネットワークである -- Transformer、RNN、学習された関数 -- 数十億回の勾配更新を経て、行動系列データの上で環境を知覚し、目標を推論し、行動を起こすことを学んだもの。AI における "Agent" という言葉は、始まりからずっとこの意味だった。常に。 +Agent の核心にあるのはニューラルネットワークだ -- Transformer、RNN、学習された関数 -- 数十億回の勾配更新を経て、行動系列データの上で環境を知覚し、目標を推論し、行動を起こすことを学んだもの。Agency は周囲のコードから与えられるものではない。訓練を通じてモデルが獲得するものだ。 -人間も Agent だ。数百万年の進化的訓練によって形作られた生物的ニューラルネットワーク。感覚で世界を知覚し、脳で推論し、身体で行動する。DeepMind、OpenAI、Anthropic が "Agent" と言うとき、それはこの分野が誕生以来ずっと意味してきたものと同じだ:**行動することを学んだモデル。** +人間が最もわかりやすい例だ。数百万年の進化的訓練によって形作られた生物的ニューラルネットワーク。感覚で世界を知覚し、脳で推論し、身体で行動する。DeepMind、OpenAI、Anthropic が "Agent" と言うとき、その核心は常に同じことを指している:**訓練によって行動を学んだモデルと、それを特定の環境で機能させるインフラの組み合わせ。** 歴史がその証拠を刻んでいる: @@ -26,7 +26,7 @@ Agent とはニューラルネットワークである -- Transformer、RNN、 - **2024-2025 -- LLM Agent がソフトウェアエンジニアリングを再構築。** Claude、GPT、Gemini -- 人類のコードと推論の全幅で訓練された大規模言語モデル -- がコーディング Agent として展開される。コードベースを読み、実装を書き、障害をデバッグし、チームで協調する。アーキテクチャは先行するすべての Agent と同一:訓練されたモデルが環境に配置され、知覚と行動のツールを与えられる。唯一の違いは、学んだものの規模と解くタスクの汎用性。 -すべてのマイルストーンが同じ真理を共有している:**"Agent" は決して周囲のコードではない。Agent は常にモデルそのものだ。** +すべてのマイルストーンが同じ事実を示している:**Agency -- 知覚し、推論し、行動する能力 -- は訓練によって獲得されるものであり、コードで組み立てるものではない。** しかし同時に、どの Agent も動作するための環境を必要とした:Atari エミュレータ、Dota 2 クライアント、StarCraft II エンジン、IDE とターミナル。モデルが知能を提供し、環境が行動空間を提供する。両方が揃って初めて完全な Agent となる。 ### Agent ではないもの @@ -367,6 +367,6 @@ MIT --- -**モデルが Agent だ。コードは Harness だ。優れた Harness を作れ。Agent が残りをやる。** +**Agency はモデルから生まれる。Harness が Agency を現実にする。優れた Harness を作れ。モデルが残りをやる。** **Bash is all you need. Real agents are all the universe needs.** diff --git a/README-zh.md b/README-zh.md index 843cce1f3..9ed73ef30 100644 --- a/README-zh.md +++ b/README-zh.md @@ -2,17 +2,17 @@ [English](./README.md) | [中文](./README-zh.md) | [日本語](./README-ja.md) -## 模型就是 Agent +## Agency 来自模型,Agent 产品 = 模型 + Harness -在讨论代码之前,先把一件事彻底说清楚。 +在讨论代码之前,先把一件事说清楚。 -**Agent 是模型。不是框架。不是提示词链。不是拖拽式工作流。** +**Agency -- 感知、推理、行动的能力 -- 来自模型训练,不是来自外部代码的编排。** 但一个能干活的 agent 产品,需要模型和 harness 缺一不可。模型是驾驶者,harness 是载具。本仓库教你造载具。 -### Agent 到底是什么 +### Agency 从哪来 -Agent 是一个神经网络 -- Transformer、RNN、一个被训练出来的函数 -- 经过数十亿次梯度更新,在行动序列数据上学会了感知环境、推理目标、采取行动。"Agent" 这个词在 AI 领域从诞生之日起就是这个意思。从来都是。 +Agent 的核心是一个神经网络 -- Transformer、RNN、一个被训练出来的函数 -- 经过数十亿次梯度更新,在行动序列数据上学会了感知环境、推理目标、采取行动。Agency 这个东西从来不是外面那层代码赋予的,而是模型在训练中学到的。 -人类就是 agent。一个由数百万年进化训练出来的生物神经网络,通过感官感知世界,通过大脑推理,通过身体行动。当 DeepMind、OpenAI 或 Anthropic 说 "agent" 时,他们说的和这个领域自诞生以来就一直在说的完全一样:**一个学会了行动的模型。** +人类就是最好的例子。一个由数百万年进化训练出来的生物神经网络,通过感官感知世界,通过大脑推理,通过身体行动。当 DeepMind、OpenAI 或 Anthropic 说 "agent" 时,他们说的核心都是同一件事:**一个通过训练学会了行动的模型,加上让它能在特定环境中工作的基础设施。** 历史已经写好了铁证: @@ -26,7 +26,7 @@ Agent 是一个神经网络 -- Transformer、RNN、一个被训练出来的函 - **2024-2025 -- LLM Agent 重塑软件工程。** Claude、GPT、Gemini -- 在人类全部代码和推理上训练的大语言模型 -- 被部署为编程 agent。它们阅读代码库,编写实现,调试故障,团队协作。架构与之前每一个 agent 完全相同:一个训练好的模型,放入一个环境,给予感知和行动的工具。唯一的不同是它们学到的东西的规模和解决任务的通用性。 -每一个里程碑都共享同一个真理:**"Agent" 从来都不是外面那层代码。Agent 永远是模型本身。** +每一个里程碑都指向同一个事实:**Agency -- 那个感知、推理、行动的能力 -- 是训练出来的,不是编出来的。** 但每一个 agent 同时也需要一个环境才能工作:Atari 模拟器、Dota 2 客户端、星际争霸 II 引擎、IDE 和终端。模型提供智能,环境提供行动空间。两者合在一起才是一个完整的 agent。 ### Agent 不是什么 @@ -367,6 +367,6 @@ MIT --- -**模型就是 Agent。代码是 Harness。造好 Harness,Agent 会完成剩下的。** +**Agency 来自模型。Harness 让 agency 落地。造好 Harness,模型会完成剩下的。** **Bash is all you need. Real agents are all the universe needs.** diff --git a/README.md b/README.md index 02561fef1..5d31cf7d1 100644 --- a/README.md +++ b/README.md @@ -1,17 +1,17 @@ [English](./README.md) | [中文](./README-zh.md) | [日本語](./README-ja.md) # Learn Claude Code -- Harness Engineering for Real Agents -## The Model IS the Agent +## Agency Comes from the Model. An Agent Product = Model + Harness. -Before we talk about code, let's get one thing absolutely straight. +Before we talk about code, let's get one thing straight. -**An agent is a model. Not a framework. Not a prompt chain. Not a drag-and-drop workflow.** +**Agency -- the ability to perceive, reason, and act -- comes from model training, not from external code orchestration.** But a working agent product needs both the model and the harness. The model is the driver, the harness is the vehicle. This repo teaches you how to build the vehicle. -### What an Agent IS +### Where Agency Comes From -An agent is a neural network -- a Transformer, an RNN, a learned function -- that has been trained, through billions of gradient updates on action-sequence data, to perceive an environment, reason about goals, and take actions to achieve them. The word "agent" in AI has always meant this. Always. +At the core of every agent is a neural network -- a Transformer, an RNN, a learned function -- that has been trained, through billions of gradient updates on action-sequence data, to perceive an environment, reason about goals, and take actions. Agency is never granted by the surrounding code. It is learned by the model during training. -A human is an agent. A biological neural network, shaped by millions of years of evolutionary training, perceiving the world through senses, reasoning through a brain, acting through a body. When DeepMind, OpenAI, or Anthropic say "agent," they mean the same thing the field has meant since its inception: **a model that has learned to act.** +Humans are the best example. A biological neural network shaped by millions of years of evolutionary training, perceiving the world through senses, reasoning through a brain, acting through a body. When DeepMind, OpenAI, or Anthropic say "agent," the core of what they mean is always the same thing: **a model that has learned to act, plus the infrastructure that lets it operate in a specific environment.** The proof is written in history: @@ -25,7 +25,7 @@ The proof is written in history: - **2024-2025 -- LLM agents reshape software engineering.** Claude, GPT, Gemini -- large language models trained on the entirety of human code and reasoning -- are deployed as coding agents. They read codebases, write implementations, debug failures, coordinate in teams. The architecture is identical to every agent before them: a trained model, placed in an environment, given tools to perceive and act. The only difference is the scale of what they've learned and the generality of the tasks they solve. -Every one of these milestones shares the same truth: **the "agent" is never the surrounding code. The agent is always the model.** +Every one of these milestones points to the same fact: **agency -- the ability to perceive, reason, and act -- is trained, not coded.** But every agent also needed an environment to operate in: the Atari emulator, the Dota 2 client, the StarCraft II engine, the IDE and terminal. The model provides intelligence. The environment provides the action space. Together they form a complete agent. ### What an Agent Is NOT @@ -372,6 +372,6 @@ MIT --- -**The model is the agent. The code is the harness. Build great harnesses. The agent will do the rest.** +**Agency comes from the model. The harness makes agency real. Build great harnesses. The model will do the rest.** **Bash is all you need. Real agents are all the universe needs.**