Librería Samer Atenea
Kálamo Books
Librería Elías (Asturias)
Librería Kolima (Madrid)
Librería Proteo (Málaga)
Joshi’s book shows how AI turns engineering design around: engineers state goals and constraints, and computers propose the designs. Its key idea is the generate-evaluate-select loop, with the engineer refining the specification in an outer loop. A single bracket example links all twelve chapters. The first part covers foundations: design representations, optimization methods and topology optimization, with runnable code. The second covers learning to design: surrogate models, Bayesian optimization, VAEs, GANs, diffusion models, physics-informed networks, and quality-diversity search. The third covers practice: language-model design agents, manufacturing constraints, verification and uncertainty, data operations, oversight and ethics, plus four case studies. Throughout, Joshi warns that optimizers exploit flawed models, so results need independent re-checks. His closing rule: specify carefully, generate broadly, evaluate honestly, select thoughtfully, and keep the engineer accountable.