I've done it since the very beginning. ChatGPT is a decent philosopher and can make good arguments for and against positions when prompted. It can also provide

I've done it since the very beginning. ChatGPT is a decent philosopher and can make good arguments for and against positions when prompted. It can also provide insights on how apparently unconnected arguments can be related. The approach I've taken is very similar to what the article describes; it's an area where it's hard to argue against the usefulness of a good LLM.