Ever wondered how AI really “thinks”? In the latest episode of The Narrow Window, I describe the inner workings of neural networks. We’ll explore shock in Norway and multi-dimensional geometry.

If you’re fascinated by the black box mystery of AI, this episode is for you.

This episode breaks down:• How AI generates responses using statistical probabilities and a dash of randomness• The surprising complexity behind AI’s knowledge encoding• How to think about impossible-to-visualise higher dimensions

For more advanced reading, you can find the paper I mentioned here:

https://transformer-circuits.pub/2025/attribution-graphs/biology.html

You can also find my book, The Narrow Window, on Amazon here:

https://mybook.to/TheNarrowWindow

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit thenarrowwindow.substack.com