The last decade proved something uncomfortable: Innovation is no longer limited by ideas—it is limited by infrastructure. We can imagine autonomous vehicles, intelligent medical devices, sovereign AI, and planetary-scale communication. But we cannot build the silicon fast enough to keep up.

That is the real crisis hiding behind today’s headlines.
For 50 years, Moore’s Law masked inefficiency. Transistors got cheaper, faster, and denser, and brute force carried us forward. That era is over. Modern chips take years to build, cost hundreds of millions of dollars, and depend on shrinking pools of specialised talent. Verification alone consumes more time and money than entire startups can afford. As a result, only a handful of companies can shape the hardware future—and everyone else must adapt to it.
This is not sustainable.
The next century of change will not be defined by who has the most fabs or the deepest balance sheet. It will be defined by who can collapse the distance between intent and silicon.
We have seen this movie before. Software once required elite teams, custom infrastructure, and long release cycles. Then abstraction happened. Platforms emerged. Apps replaced monoliths. Suddenly, a small team—or a single developer—could change the world.
{{/usCountry}}We have seen this movie before. Software once required elite teams, custom infrastructure, and long release cycles. Then abstraction happened. Platforms emerged. Apps replaced monoliths. Suddenly, a small team—or a single developer—could change the world.
{{/usCountry}}Hardware is next.
The inevitable destination is clear: Building a chip must become as easy as creating an app for your smartphone. If you can describe a product, you should be able to produce silicon for it—securely, rapidly, and economically.
This transformation will not come from incremental tool improvements. It requires a fundamental shift in how engineering work is done. For decades, EDA tools automated physics. The real bottleneck has always been human effort—writing specs, building verification environments, debugging regressions, and preserving tribal knowledge. That is where time, cost, and risk live.
Agentic AI changes this equation. Not as a chatbot or copilot, but as an autonomous collaborator—one that understands specifications, reasons across entire codebases, works inside existing tools, and executes workflows at machine speed. Humans set intent. AI executes, checks, and iterates.
The consequences are profound. Custom silicon becomes accessible. Innovation decentralises. Entire industries—from healthcare and robotics to energy, defense, and space—gain the ability to build purpose-designed hardware instead of waiting for it.
The decade of disruption showed us what happens when software scales faster than institutions. The century of change will be defined by something even more radical: When silicon finally learns to move at the speed of imagination.
The question is no longer whether this future arrives.
It’s who has the courage to build it.
This article is authored by Shelly Henry, CEO, Moores Lab AI,- Global Business Summit 2026.