Learned, Not Programmed: A Beginner's Guide to How Artificial Intelligence, Machine Learning, Neural Networks and Generative AI Actually Work
English | September 24, 2026 | ASIN: B0HKWXRYQJ | 271 pages | Epub | 4.22 MB
How does a chatbot actually produce an answer? Why can an image generator create a picture no one has painted, yet struggle to count the objects in it? And why do AI systems sometimes state false things with complete confidence?Learned, Not Programmed explains what is really happening inside modern artificial intelligence, one step at a time. It requires no programming and no math beyond everyday arithmetic. The central idea is simple: today's AI systems are not given their behavior line by line. People design them, choose their data and set their goals, but the detailed behavior is learned from examples. Once you see that, much of what AI does, and much of what it gets wrong, starts to make sense.Inside, you will learn:How machine learning turns examples into behavior, and how a system measures and reduces its own mistakesWhat a neural network is, and how it learns by adjusting its settingsHow tokens and embeddings turn words into numbersHow transformers use attention to weigh contextHow a text predictor becomes a helpful assistantWhat happens, step by step, from the moment you press Enter to the answer on your screenHow generative AI creates images, voices and videoWhy AI makes mistakes: invented facts, bias, brittleness and uneven abilityHow to use AI with sound judgment, and how to protect yourself from deepfakes and scamsThe book is organized in ten short parts: what AI actually is, how machines learn, neural networks, turning language into numbers, and how a predictor becomes an assistant. It then follows a single question from start to finish before moving on to images and sound, why AI makes mistakes, using AI responsibly, and the debates about AI's future.It includes 27 original diagrams, a glossary of about 90 terms, and notes that document the sources behind its key claims. Where experts disagree or the evidence is still uncertain, it says so.This book is for you if:You use AI tools and want to understand what is happening behind the screenYou have heard terms like "neural network," "token" or "transformer" and want them explained in plain languageYou want to judge claims about AI, both the hype and the fear, for yourselfYou are a student, parent, teacher or professional who needs a solid foundation without a technical backgroundWritten for curious readers who want understanding, not just tips.
Links are Interchangeable - Single Extraction
