LLMs, Board Layout and Human Engineering (gregtidanian.substack.com)

OpenAI recently unveiled its latest model, Astra, which has the capability to perform printed circuit board (PCB) layouts using KiCAD, significantly streamlining a traditionally manual task that has long caused delays in electronic design. This announcement has sparked concerns within the embedded engineering community regarding the potential displacement of skilled engineers. While large language models (LLMs) like Astra can now automate PCB design, engineers worry about the impact on their professions and the reliability of AI-generated electronics. The significance of Astra's capabilities extends beyond mere automation; it introduces the possibility of refining engineering processes by integrating LLMs into design workflows. While LLMs could reduce technical debt by identifying human errors in specifications—such as inadequately sized track widths for high-current applications—there remain questions about the quality of context they provide. The role of engineers as systems thinkers, who evaluate user problems and ensure requirements are met, will likely become even more critical as AI tools evolve. The future may see a shift toward a landscape where human engineers focus more on overarching system design and validation, while LLMs handle more routine tasks, pointing to a new paradigm in engineering that harnesses technology to enhance creativity and efficiency.