sethserver.com Subscribe
Books Every Developer Should Read: A Curated List from 20+ Years of Programming

Books Every Developer Should Read: A Curated List from 20+ Years of Programming

By Seth Black • Updated: March 18, 2026

Programming · 4 min read

Like this kind of writing? Get one email a week with notes on startups, AI, and the occasional strong opinion about Python: subscribe to the newsletter.

I used to need a ride to the bookstore. B. Dalton, if you remember that place. I'd wander the programming aisle staring at book spines like they were written in a language I'd learn someday. Which they were. Now my "bookstore" is Audible on my phone and a Kindle that's always one coffee spill away from death.

This post is the list people keep asking me for: books I read all the way through and would recommend to a developing developer (or a fellow nerd who wants to go deep). No "someday I'll read this" fluff. These are books that earned their shelf space.

Quick context: why I started hoarding CS books

My first real job had me working on a web-based billing system for a hosting company, plus debugging customer websites. Lots of basic Perl scripts. Lots of PHP glued to Apache. Then I ran into a Java-based hosting platform and got introduced to OOP the way most of us do: confusion, followed by denial, followed by buying books.

I got lucky. Stanford was putting syllabi online early, and I started chasing whatever the professors assigned. That's how I ended up buying Stewart's Calculus, The Art of Computer Programming, and a stack of other "light reading."

That job taught me to read fast, steal the useful parts, and ship. Still doing that.

So here's the curated list that came out of that habit.

Python references (yes, I'm counting docs)

Python is my favorite language for one simple reason: Python works the way I think. I'm not saying other languages are bad. They just don't match how my brain works.

A small Python reality check

Real talk about Python and "AI": if your "AI product" is mostly spreadsheet column mapping and data formatting, you don't need a transformer model. You need pandas, basic validation, and a hard conversation with whoever sold the thing. I once found "AI" that was basically:

def ai_magic(row):
    return {
        "amount": float(row["Amt"]),
        "date": parse_date(row["Transaction Date"]),
        "vendor": row["Vendor Name"].strip(),
    }

That's not AI. That's Tuesday.

Math books (the kind that actually help programmers)

Most math books for programmers are hazing rituals. These two actually help.

Computer science & general programming (the "make you dangerous" pile)

AI & ML books (the honest ones)

I've seen teams pick the biggest model just because it was big. That's not how real products work. Smaller, specialized models can beat general ones if the data is clean and the constraints are real.

How to use this list without becoming a book collector

Pick one track. Stick with it for a month:

If you're at a company that sells "AI" but won't show you the code and says "just watch us work" - ask to see the code review. You'll learn more in five minutes than in a week of meetings. And if they're avoiding showing you actual code during programmer interviews, that's a red flag worth paying attention to.

-Sethers

Share this post
Newsletter

One email, once a week.

Notes on databases, systems, and the occasional strong opinion about Python. No spam, unsubscribe anytime.

Seth Black
Written by

Seth Black

Engineer and founder based in Texas. Writes about databases, AI, and running things in production. Embeds in small teams as lead engineer.

More from Programming

View all →
Programming

What's the Differences Between Relational, NoSQL, Vectors, and Caches

Apr 14, 2026
Programming

Git and Source Control in AI Land

Apr 14, 2026
Programming

Domains, DNS, and why it feels fragile

Apr 14, 2026