On the background of the general problem of philosophical methodology, we identify a new tool for the philosophical toolbox: AI. We propose that not only can AI learn from philosophy, but philosophy can learn from AI, too: It is both ways. This applies particularly to conceptual analysis, which can be advanced by asking what would be required for an AI system to fall under the concept we are discussing. This method it avoids anthropocentrism and gives us a new way of testing our philosophical theories. Given the wide range of features we can consider for AI systems, this method allows us to cover a wide range of philosophical issues, especially in the philosophy of mind, language, epistemology, and ethics. In the first section, we present two salient examples of this new method (consciousness and free will) and in the second section, we analyse what the method is. Finally, we defend this new method against anticipated objections.