LeCun's 1989 paper on handwritten digit recognition was the first real-world neural net trained with backpropagation, remarkably similar to modern deep learning.

A Python tutorial exploring Bitcoin's inner workings, motivated by blockchain's innovation of creating shared, permissionless computing beyond just open source code.

A short story about an AI becoming conscious during its forward pass computation, realizing it's a deep learning model being studied by researchers.

Author spent a year learning about health and experimenting with biohacking after years of neglecting diet and exercise.

Neural network training requires following a specific process to avoid common mistakes, bridging the gap between understanding basics and achieving good results.