If philosophy is as an endless debate, then you might be able to extract some "philosophy" from LLMs by studying their attractor states. This suggests a practical method for probing the philosophy of any system, or corpus. Though perhaps the fixed-point attractors, in particular, are more akin to dogma. In any case, there are some serious mathematical and computation tools that can be brought to bear on the problem.
If philosophy is as an endless debate, then you might be able to extract some "philosophy" from LLMs by studying their attractor states. This suggests a practical method for probing the philosophy of any system, or corpus. Though perhaps the fixed-point attractors, in particular, are more akin to dogma. In any case, there are some serious mathematical and computation tools that can be brought to bear on the problem.