From Recommendations to Generation: Tools Without Interfaces 从推荐系统到生成式人工智能：无接口工具

Summary 摘要

Lev Manovich explores how generative AI tools for media creation mirror the evolution of recommendation systems, both functioning within a “probabilistic culture” shaped by overwhelming digital content. He argues that GenAI relieves users from making every creative decision, instead allowing them to guide outcomes through natural language—offering a seamless, interface-free creative process. This shift democratizes creativity, enabling non-experts to access and remix the web’s vast cultural archive. Manovich positions GenAI as a logical step beyond search and recommendation technologies, redefining creativity not as invention from scratch but as intelligent synthesis—marking a fundamental transformation in how we both consume and create culture.

列夫·马诺维奇探讨了用于媒体创作的生成式人工智能工具如何反映推荐系统的发展轨迹，这两者都运作于一个由海量数字内容塑造的“概率文化”中。他认为，生成式 AI 使用户不必亲自做出每一个创意决策，而是通过自然语言来引导结果，从而实现一种无界面、无门槛的创作方式。这种转变让创意民主化，使非专业用户也能访问并重新混合互联网庞大的文化档案。马诺维奇将生成式 AI 视为继搜索和推荐技术之后的逻辑延伸，它将创造力重新定义为一种智能合成的过程，而不是从零开始的发明，标志着我们在文化消费和创作方式上的根本转变。