The Learnocene: Learning Systems Design in the Age of Artificial Intelligence

When explanations, drafts, and preliminary solutions can be obtained with the aid of artificial intelligence at a comparatively low cost, the value of learning is no longer well stated as a volume of knowledge acquired or a count of tasks completed. This essay proposes a conceptual framework, the Learnocene. It takes autonomous human development as the orienting purpose of learning and states that purpose in four mutually informing dimensions: knowing oneself, understanding the world, steering one's learning, and making futures. Six long-term capabilities sit beneath the dimensions and are not combined into a total score. Collective agency is a group-level outcome, not a seventh individual capability. Developmental aims do not become learning conditions by themselves. Learning-systems design organizes knowledge, tasks, support, evidence, and operating conditions, and revises them in light of feedback. A personal learning system is the learner's own organization of purposes, a map of knowledge, a network of resources, relationships, tools, and evidence. A designed learning environment supports the formation of that system and hands design responsibility back over time. Structure consists of five relations: purpose, sequence, proportion, connection, and feedback. Endogenous growth is the sustained coupling of a human growth loop and a system growth loop. It is not an increase in the number of activities. The curriculum is smaller than the system. Person, family, school, and society use one six-step design cycle. A social resource becomes learning only after translation. Courses and projects place real-world topics, transferable skills, and cognitive lenses in the service of long-term capabilities. The mechanism is not a formula. Assessment separates knowledge, use, and long-term capability, and it separates the project, the learner, and the system. Where evidence is insufficient, no judgment is entered, and the gap is not scored as zero. The role of artificial intelligence is set by the learning aim, the risk of the task, and the learner's autonomy. It is not assigned by age. The four scales use one conceptual scheme and different rules of responsibility and data. Adults carry uncertainty in the family. The school is a hub, not the sole center of learning. At the societal scale, an understanding of learning is argued to precede the volume of resources. That claim remains a hypothesis. A student's personal growth data are not an object of trade. The essay reports no primary data. Definitions are stipulative. Claims about effects and historical scope are hypotheses. A complete structure is not evidence of an effect. Revision changes the text when the evidence changes and does not add parallel terms in order to avoid a counterexample. A minimum version must still stand when resources are scarce. Whether unequal resources are thereby amplified remains to be tested. 中文摘要 当解释、草稿和初步方案可以借助人工智能以较低成本获得时,学习的价值便难以再主要用知识获取量或任务完成量来说明。本文提出一个概念框架,称为学习世。它把学习的取向确定为人的自主发展,并陈述为四个相互推动的维度:认识自我、理解世界、驾驭学习、创造未来。六项长期能力位于四维之下,不作总分。共同能动性是群体结果,不是第七项个人能力。 发展目标不会自行变成学习条件。学习系统设计把知识、任务、支持、证据和运行条件组织起来,并按反馈调整。个人学习系统是学习者自己组织的目标、知识地图、资源网络、关系、工具和证据。被设计的学习环境支持这一系统形成,并逐步移交设计权。结构由目标、先后、比例、连接和反馈五种关系构成。人的成长循环与系统的生长循环持续耦合,构成可观察的自生长。自生长不是活动数量的增加。 课程小于系统。个人、家庭、学校和社会使用同一个六步设计循环。社会资源要经过转译才成为学习。课程与项目用真实议题、可迁移技能和认知透镜服务于长期能力。这一机制不是计算公式。评价分开知识、运用和长期能力,也分开项目、学习者与系统。证据不足时不作判断,不记为零。人机档位由学习目标、任务风险和自主程度决定,不按年龄发放。 四个尺度使用同一套概念,责任和数据规则不同。家庭中的不确定性由成年人承担。学校是枢纽,不是唯一中心。社会尺度上,对学习的认识先于资源的数量,这一句仍待检验。学生个人成长数据不是交易对象。 本文没有原始数据。定义是概念规定。关于效果和历史范围的判断是有待检验的命题。结构完整不等于效果已经得到支持。修订在证据改变时修改文本,不靠增加平行术语。低资源条件下仍须有站得住的最小版本。差距是否因此被放大,仍待检验。 The English PDF is the primary file. The Chinese PDF is a parallel version of the same text (zho).