Artificial Intelligence in Finance: A Book Worth Reading in 2025
Book: Artificial Intelligence in Finance Author: Yves Hilpisch Publisher: O’Reilly, 2020 ISBN: 978-1-492-05543-3
Why This Book, Why Now
Here’s a question that bugs me. Can AI actually beat the stock market? Not in a sci-fi movie way. In a real, consistent, make-money-while-you-sleep way.
That question is exactly what pulled me into Artificial Intelligence in Finance by Yves Hilpisch. The book was published in 2020 by O’Reilly, and it tries to answer something that most finance textbooks avoid: what happens when you throw modern AI at financial markets?
I picked this book up because I kept seeing two worlds that rarely talk to each other. On one side, you have AI people building insane things. Neural networks that beat humans at chess and Go. Self-driving cars. Language models that write code. On the other side, you have finance people with their own set of theories. Efficient markets. Portfolio optimization. Risk models built on assumptions from the 1960s.
Hilpisch tries to connect those two worlds. And that’s what makes this book interesting.
What the Book Actually Covers
The book starts by comparing AI across different fields. Gaming, autonomous driving, and finance. It asks a simple question: if AI can master Atari games and beat world champions at Go, can it also figure out financial markets?
Then it gets into the building blocks. What AI actually is. How algorithms learn from data. How neural networks work and why they need so much data to be useful. This part is the foundation for everything else in the book.
From there, things get wild. Hilpisch talks about superintelligence. Not in a hand-wavy futurist way, but in a structured breakdown. He covers the different levels of AI intelligence, from narrow AI that does one thing well, to artificial general intelligence, to the theoretical idea of superintelligence. He looks at what DeepMind accomplished and what that means for the future.
And then comes the finance part. Traditional financial theory. Expected utility. Mean-variance portfolios. The Capital Asset Pricing Model. Arbitrage pricing theory. These are the ideas that have shaped how Wall Street thinks about risk and return for decades.
But here is the real point of the book. Hilpisch argues that AI could challenge one of finance’s most important ideas: the efficient markets hypothesis. That’s the theory that says markets already reflect all available information, so you can’t consistently beat them. What if AI changes that equation?
What This Blog Series Will Look Like
I’m going to walk through this book chapter by chapter. Each post will cover a major section, break down the key ideas, and add my own thoughts on what holds up and what doesn’t.
Here’s the rough plan:
- The Preface and how AI in gaming, driving, and finance compare
- AI fundamentals covering algorithms, neural networks, and why data matters so much
- Superintelligence and whether it’s real science or just speculation
- Traditional finance theory and the models that have ruled Wall Street
- Final thoughts on the book’s big argument about AI versus efficient markets
Each post will explain the ideas in plain language. No finance degree required. No CS degree either. If you understand what a stock is and you’ve heard of machine learning, you’ll be fine.
Why Bother With a 2020 Book?
Fair question. AI has moved fast since 2020. Large language models changed everything. But here’s the thing. The fundamentals Hilpisch covers haven’t changed. Neural networks still work the same way. Financial theory hasn’t been rewritten. And the core question, whether AI can find patterns in markets that humans miss, is more relevant now than it was when the book came out.
If anything, reading this book in 2025 gives you a useful perspective. You can see what the field expected back then and compare it to where we actually ended up. Some predictions look smart in hindsight. Others, not so much.
Who Should Follow Along
This series is for you if:
- You’re curious about how AI applies to real-world finance
- You want to understand the theory behind algorithmic trading
- You’ve heard terms like “neural networks” and “efficient markets” but want the full picture
- You like reading about big ideas without wading through academic papers
I’ll keep things honest. When the book makes a strong argument, I’ll say so. When something feels like a stretch, I’ll call that out too.
Let’s get into it.