Categories Machine Learning

How These 7 Books Destroyed My Confidence (and Made My PhD Possible)

Seven technical monsters that melted my GPU, broke my brain, and somehow made me love research more.

Press enter or click to view image in full size

Image generated using AI

You want to know the exact moment I became an insufferable PhD scholar?

My first-year induction. I’m 23. I’ve done a few NPTEL courses, built a half-decent MNIST classifier, I think I’m a big shot.

Some final-year senior is presenting his research. His slides aren’t slides. They’re just Greek letters. A dense, terrifying wall of them.

Afterward, I try to sound smart. “That was cool, dada. What was the loss function you used?”

He looks at me. Not in a mean way. Just… tired. Like a Delhi-ite explaining winter to a Mumbai-kar.

“It’s not really a loss function, bhai,” he says. “It’s a non-parametric Bayesian hierarchical model. You read Murphy?”

“Murphy? Like… the old radio company?”

He didn’t even laugh. “No. The Murphy. ‘Probabilistic Perspective.’”

I went back to my cramped desk in the lab, downloaded the 1,200-page PDF, and had my first real, PhD-grade existential crisis. I realized I didn’t just need to learn new things. I needed to re-learn everything I…

You May Also Like