The classic, linear design process is obsolete because AI tools allow engineers to build and iterate so quickly. Designers must shift from a gatekeeping, mock-heavy process to a more fluid, collaborative role that supports rapid execution.

For physical design, simulation shouldn't just be a final verification step. Instead, it should be a tool used during model training to build the AI's intuition or "taste." This allows the model to generate high-quality designs quickly at inference time, mirroring how expert human engineers develop their skills.

Designers should consider the human operators and machines that will assemble their product. By making choices that simplify manufacturing—providing clear instructions and avoiding known difficulties—the process becomes smoother and more efficient, akin to 'riding a bike downhill.'

The common practice of offshoring manufacturing, exemplified by Apple, creates a critical flaw by severing the feedback loop between designers and producers. This leads to suboptimal product design and simultaneously transfers advanced manufacturing skills and capabilities to other nations, like China.

To ensure a smooth transition from development to production, an operations or manufacturing SME must be part of the design process from the start. Otherwise, products are developed without manufacturability in mind, leading to expensive, reactive fixes and subjective quality control during scale-up.

Unlike text-based AI that relies on descriptive prompts, some advanced design tools for physical components work in reverse. The user defines 'no-go' zones and constraints, and the AI then generates numerous optimized design possibilities within those boundaries.

AI tools like LLMs thrive on large, structured datasets. In manufacturing, critical information is often unstructured 'tribal knowledge' in workers' heads. Dirac’s strategy is to first build a software layer that captures and organizes this human expertise, creating the necessary context for AI to then analyze and add value.

Designing for AI is less about crafting pixel-perfect UIs in Figma and more about creating the underlying system or "harness." This involves enabling the agent to perform long-running tasks, verify its own work, and operate effectively within technical constraints, which is where the real design work lies.

Full automation in electronics manufacturing doesn't require robotics breakthroughs. The existing robots are sufficient. The challenge is designing circuit boards that are 100% compatible with current automation, eliminating the 20% of manual labor caused by non-standard components. AI can create these constrained, manufacturable designs.