论文是历史发明，不是知识命运。The paper is a historical invention, not the destiny of knowledge.它曾经适合印刷时代，但不应垄断 AI 时代。It suited the print era; it should not monopolize the AI era.
期刊应从印刷容器转型为认证、策展、导览和公共展示机构。Journals should move from print containers to certification, curation, guidance, and public display.
学术成果应从单一文本升级为经过学术策展的动态知识客体。Scholarly outputs should become curated, dynamic knowledge objects.文本、数据、代码、图表、视频、实验日志、失败记录、伦理材料和 AI 接口都是成果的一部分。Text, data, code, figures, video, logs, failures, ethics, and AI interfaces all belong to the work.
复现是设计之初就必须内嵌的结构。Reproducibility is designed in from the start.不能被检查的成果，不应享有过高的学术信用。Work that cannot be checked should not enjoy high credit.
失败结果、无效结果和重复实验应被记录、检索和评价。Failed, null, and replicated results should be recorded, searchable, and valued.
AI 应成为透明的学术助手，而不是匿名的影子作者。AI should be a transparent assistant, not an anonymous shadow author.
人类最重要的贡献是问题意识、价值判断、身体经验、伦理责任、异常感知和想象力。The human contribution is question-making, judgment, embodied experience, responsibility, anomaly perception, and imagination.
公众、初学者和跨学科读者是合法读者，不是外部观众。The public, students, and adjacent fields are legitimate readers.
学术成果应同时对人类友好和对机器友好。Scholarly objects should be friendly to humans and actionable for machines.
发表不是终结，而是进入公共修正过程。Publication begins public revision rather than ending inquiry.
评价应转向可信度、开放度、可复现度、解释力、公共价值和可接续性。Evaluation should turn to trustworthiness, openness, reproducibility, explanatory power, public value, and reuse.
建立“新科学”平台、新型期刊和“问道”套件，让人与 AI 有机配合。Build the platform, the journal, and Wendao so that humans and AI cooperate.
科学应当可玩、可看、可参与。Science should be playable, watchable, and joinable.AI 降低的是操作科研的门槛，而不是判断成果的标准；有人上手玩，所有人都能当观众，而每一件成果都必须有人讲得清、查得到、认得账。AI lowers the threshold of operating research, not the standard for judging it; some play, everyone may watch, and every result must be explained, checkable, and owned.
