> You've spent so much time optimizing ML models that your own brain now refuses to process any thought that could be represented more efficiently with fewer neurons.