Abstract: The rapid development of modern AI technology has shifted philosophical research from traditional metaphysical speculation to more practical  methodological reconstruction, marking a profound transformation of philosophy of AI from conceptual analysis to an engineering philosophy paradigm. Confronted with fundamental differences between AI and human intelligence, the complexity of technological systems, and the dynamic nature of intelligent generation, it is essential to systematically explore the methodological innovations in philosophy of AI across multiple dimensions: types of intelligence, research paradigms, and the philosophy of scientific experimentation. We argue that the innovations include the adoption of a multidimensional perspective on types of intelligence, a paradigm shift to a systematic, generative, and contingency-sensitive engineering philosophy, and a new mindset based on fallibility, complexity, and openness. By generalizing the concept of space through the interpretive framework of signifier topology and folded spatial concepts, research in philosophy of AI can also reshape intelligence. This approach not only addresses the theoretical challenges posed by technological practice to philosophy but also opens new avenues for intelligent cognition in the “post-human” era.

Key Words: Philosophy of AI; Types of intelligence; Philosophy of engineering; Philosophy of scientific

experimentation; Research paradigm