読書リサーチレーダー投資フレームワーク
ログイン / 新規登録
← 読解に戻る
ニューストピック · AI编码智能体工作流重大3 分で読めるDisconfirmAI 編集部

Django Co-Founder Simon Willison: A Single 'Make It More Fun' Prompt Sent Coding Agents Into Over 1,000 Simulated Matches

Simon Willison originally believed that coding agents could only produce things that looked like games but weren't actually fun. He intentionally tested this assumption with a vague instruction; Claude, running in Ultracode mode, conducted over 1,000 simulated matches, using the results to rebalance units and mechanics, ultimately producing what he considers a genuinely enjoyable game. This experiment reveals under what conditions vague quality instructions are effective versus when they merely result in superficial feature stacking.

2026-10-21 03:34 GMT+8

Simon Willison shared his assessment with friends at a dinner table: coding agents could build something that looked like a game, but couldn't create something people would genuinely want to keep playing. After stating this, he decided to give the models a chance to prove him wrong. He typed a single line into Claude Opus 5.5 on his phone: 'This game isn't fun enough yet. Make it a LOT more fun.'

This marked the fourth major iteration of his Battle Drones project. The previous three versions—the initial prototype, full sprites, and varied terrain—were each 'recognizable as an RTS' but 'wasn't any fun to play.' Willison admitted he had barely reviewed the code and didn't write much TypeScript himself; the project was driven entirely by coding agents, with all 321 commits completed by them.

For this iteration, he used Ultracode mode. According to the Claude Code documentation, dynamic workflows are JavaScript scripts that orchestrate numerous sub-agents. He started the process in the evening and received playable results late the next morning. Later, he asked Claude to write its own release notes, which included a section stating: In 1,200 normal-difficulty AI-vs-AI test matches, games still undecided after 25 minutes dropped from 14% to 0.9%; the share of wins by a side that was behind by over 800 resources rose from 24% to 46%.

ログイン / 新規登録
読書リサーチレーダー投資フレームワーク

Willison later explained why the agent took so long: it ran over a thousand simulated playtest sessions, using the results to rebalance existing units, add new ones, and introduce new mechanics. In other words, the vague instruction 'more fun' was internally converted by the agent into measurable metrics—match duration, comeback probability, resource gaps, and their correlation with win/loss outcomes—which were then optimized through extensive simulation iterations.

This mechanism works because game balance itself possesses a cheap and faithful internal feedback loop. Simulated matches do not require real players or external data; the agent can generate sufficient 'player behavior' to evaluate whether changes make matches more tense or offer better comeback potential. Willison chose games as the experimental field partly because his software engineering experience has less impact in this domain, and because 'games are super low-risk,' with the worst-case scenario being just a small bug.

However, Willison expressed clear discomfort. He wrote: 'I'm still not quite over the guilt of building software in this way. I don't feel like I've earned the right to create something this good, in a field in which I have so little prior experience.' He acknowledged having almost no game development experience, with design inputs limited to vague instructions like 'make it more fun' and 'add visual polish.' It wasn't until the eighth version that he provided a truly specific design specification for the first time—a map requiring infantry to be transported by Condors to capture islands.

The practical takeaway for readers is whether vague quality instructions work depends on whether the target domain contains a feedback loop that agents can simulate and measure internally. Game balance does have such loops, but concepts like 'is the user experience good?','is this solution elegant?', or 'is this report convincing?' typically lack equally cheap and faithful internal agent metrics. Without such loops, similar vague instructions are likely to yield only superficial feature stacking rather than genuine quality improvement.

ソース

  • simonwillison.netSimon Willison’s Weblog↗
トピック

AI编码智能体工作流

AI编码智能体正重塑软件开发流程,但“快”不等于“好”。支付公司Equals披露其92%代码由AI生成,英国数字银行Monzo的智能体每日处理1800项任务,显示规模化应用已成现实。然而,GitLab调研指出,编码仅占开发时间的16%,瓶颈…

5 ステップ5 件更新日 2026-10-22 15:57

経緯

  1. 2026-04-01GitLab发布基于1528名DevSecOps从业者的调查,指出AI虽提升个人生产力,但大型软件交付周期中活跃开发仅占10%-20%,瓶颈转向审查与环境等待。about.gitlab.com ↗
  2. 2026-09-24支付公司Equals首席产品官James Simcox在Okta大会表示,公司超92%代码现由AI编写,并强调需对智能体实施比人类更窄的权限及推理过程记录。diginomica.com ↗
  3. 2026-10-19程序员Jeremy Scheff发表文章区分两类开发者:享受解决逻辑谜题者因智能体感到真实丧失,而享受搭建系统者则受益。simonwillison.net ↗

未解決の問題

  • 当代码生成速度远超人工审查能力时,如何量化“有效审查”而非“橡皮图章”?
  • 对于非结构化或高创意需求(如游戏趣味性),智能体的自主迭代边界在哪里?
  • 企业如何平衡智能体的高权限需求与最小权限原则带来的安全合规成本?

このスレッド内の他の記事

  • Monzo engineers discuss in-house AI coding agent: from Slack emoji to 1,800 daily tasks2026-10-21
  • GitLab AI-Native Engineering Playbook: Coding Takes Only 16% of Working Time, Bottleneck Has Shifted to Review and Validation2026-10-21
トピック全文を読む →
リサーチを続ける →