Artificial Intelligence for the Z80 is a hands-on introduction to AI that teaches its core ideas by building programs in BASIC for the Z80, including the ZX Spectrum and Spectrum Next. Its central premise is that working within an 8-bit computer’s tight limits makes algorithms and their trade-offs easier to see: readers write code, run it, measure or observe what happens, and experiment with changes. The 12-chapter progression starts with AI history and sorting algorithms, using execution times to show why efficient methods matter, then explores rule-based expert systems, an ELIZA-style chatbot, and game-playing search through Minimax and alpha-beta pruning.
The later chapters move from explicit rules to data and learning. Examples include fitting a regression line to real Mauna Loa CO₂ measurements; classifying fantasy characters with k-NN and Naive Bayes; and discovering unlabeled groups with k-means. Perceptrons introduce weights, training, generalization, and the limits of linear decision boundaries, while multilayer networks explain activation functions, error correction, and backpropagation through a digit-recognition task on a 6 × 6 grid. The final chapter turns to the Game of Life, showing how simple local rules can produce complex patterns without training. Its demonstration uses a toroidal 20 × 20 board and separate arrays for successive generations. Complete code and experiments are available for study and adaptation on hardware or in an emulator, and the book is offered in Spanish and English.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.