GenAI for Systems: Recurring Challenges&Design Principles from SW to Silicon (arxiv.org)

A recent paper titled "GenAI for Systems: Recurring Challenges and Design Principles from SW to Silicon" explores the transformative impact of generative AI on computing system design, emphasizing the need for a unified approach across software, architecture, and hardware communities. The research identifies five persistent challenges in the integration of generative models across various system layers, including issues like the feedback loop crisis and the need for trust and validation. To address these challenges, the authors propose five corresponding design principles—such as embracing hybrid approaches and designing for continuous feedback—that can guide the development and implementation processes. This cross-layer analysis, drawing from over 275 studies across multiple application areas, underscores the necessity for a shared engineering methodology, which includes common vocabularies and systematic design practices. Such a framework aims to facilitate communication and collaboration among diverse stakeholders, ultimately enabling more efficient progress in the field. By mapping challenges to design principles, the paper not only highlights recurring themes but also points out open research questions that could propel further innovation in AI-driven systems engineering.