The thought experiment

I was reading one of Paul Graham’s essays where he mentionned a thought experiment. Somehow, I forgot what the essay was about, but the concept of thought experiment stuck in my head.

The point of an experiment is to uncover data that will support or kill a theory. Thought experiments have the same goal, but don’t involve any external data. They work exclusively with the experimenter’s model of the phenomenon. But if the model was only based on the current understanding of the phenomenon, how could a thought experiments manage to extrapolate data from that model to understand the phenomenon better!

The models we design seem to naturally explains more than the knowledge it was built on. This suggests our brain has evolved a natural ability to produce models that is superior to our ability to understand. Does our subconscious encode knowledge for us to discover in the models it creates for us?

So next time you are stuck on a problem, go back to your model and try to understand it better. You might know the solution better than you think!


Leave a Reply