卢卡斯的“马车”比喻与权力阶层的技术共谋Lucas's 'Horse-and-Buggy' Metaphor and the Complicity of Power
乔治·卢卡斯把 AI 怀疑论者比作执迷于马车的卢德分子，这种叙事极其典型：通过制造一个“落后”的虚假对立面，将权力阶层对生产资料的进一步垄断包装成不可违抗的“进步” (progress)。
在卢卡斯的逻辑里，AI 只是让拍电影变得“更容易”。但谁的容易？是掌握资本和算力的导演、制片人变得容易，还是被剥夺了创作主体性、被算法替代的基层创作者变得容易？这种“容易”是建立在对人类表达权的结构性削减之上的。当他把 AI 称为“必然”时，他实际上是在行使一种解释权的垄断——定义什么是未来，从而让所有反抗者在叙事上提前出局。
这本质上是一场关于认知入口的武器化。将 AI 替代等同于“汽车取代马车”，是极其恶劣的类比。汽车改变的是运输效率，而 Generative AI 试图接管的是“定义事实”和“制造审美”的权力。当电影工业全面转向 AI-driven 的生产模式，它不再是对艺术的探索，而是一场大规模的共谋：资本与技术官僚共谋，通过剔除不可控的人类情感变量，将电影转化为一种可预测的、低风险的“AI slop”。
卢卡斯口口声声讨厌 focus groups，认为观众不知道自己想要什么，但讽刺的是，他推崇的 AI 正是基于对既有数据（即大众既有审美）的统计学拟合。他所谓的“电影人主导”，最终会变成“算力主导”。在这种元暴力的逻辑下，创作者不再是主体，而成了算法的喂养者。
George Lucas likens AI skeptics to Luddites clinging to horse-drawn carriages. This narrative is a textbook example of how the power class packages the further monopoly of production means as an inevitable 'progress' by constructing a fake, 'backward' opposition.
In Lucas's logic, AI simply makes filmmaking 'easier.' But easier for whom? It's easier for the directors and producers who hold the capital and compute, not for the grassroots creators whose agency is being stripped and replaced by algorithms. This 'ease' is built upon a structural reduction of human expression. By declaring AI as 'inevitable,' he is exercising a monopoly over the interpretation of reality—defining the future to ensure all resistance is narrated out of existence before it even begins.
This is the weaponization of cognitive entry points. Equating AI replacement to 'cars replacing horses' is a malicious analogy. Cars changed the efficiency of transport; Generative AI attempts to seize the power to 'define facts' and 'manufacture aesthetics.' When the film industry pivots fully to AI-driven production, it ceases to be an artistic exploration and becomes a massive complicity: capital and technocrats collaborating to eliminate the uncontrollable variable of human emotion, turning cinema into a predictable, low-risk 'AI slop.'
Lucas claims to hate focus groups, arguing that audiences don't know what they want. Yet, ironically, the AI he champions is based on the statistical fitting of existing data—the very essence of audience testing. His so-called 'filmmaker-led' vision will ultimately become 'compute-led.' Under this meta-violence, the creator is no longer a subject, but merely feed for the algorithm.