Harvard Principles and Practices of Engineering Artificially Intelligent Systems (github.com)

A new initiative has been launched to formalize AI engineering as a foundational discipline, parallel to software and computer engineering. This effort aims to teach practitioners how to design, build, and deploy robust AI systems in real-world environments, moving beyond isolated models. An integral part of this initiative is the upcoming textbook titled "Introduction to Machine Learning Systems," which will be available in hardcopy with MIT Press in 2026. Additionally, an open repository has been established to provide a learning path consisting of theoretical concepts, hands-on activities, and performance optimization techniques for AI systems. Significantly, this project focuses on the transition from theory to practical application, incorporating elements like TinyTorch for building machine learning frameworks from scratch, hardware kits for deployment on devices such as Arduino and Raspberry Pi, and co-labs for controlled experiments. By emphasizing the importance of understanding trade-offs in AI system performance—like memory, latency, and energy consumption—the initiative addresses critical skills needed in the AI/ML community. As this resource expands, the goal is to engage a million learners by 2030, ensuring educators and engineers are equipped to tackle future challenges in responsible and sustainable AI development.