Z Advanced Computing: AI4America
by S. Rob Sobhani · The Washington TimesOPINION:
For all the headlines about artificial intelligence, most of today’s attention focuses on giant tech companies building ever‑larger models that demand staggering amounts of data, hardware, electricity, and money. The race in Silicon Valley resembles a competition to build the largest jet engine the world has ever seen. But a small Maryland company, Z Advanced Computing (ZAC), is quietly pursuing a very different vision—more Wright brothers than jumbo jet.
ZAC’s founders, brothers Bijan and Saied Tadayon, are not trying to win by outspending everyone else. Instead, they are rethinking how AI should learn in the first place. Their approach, called Cognitive Explainable AI (CXAI), teaches computers to understand concepts the way humans do—by learning from just a few examples and clearly explaining why they reach a conclusion. It is a groundbreaking shift, and the U.S. Air Force seems to agree: Last month, it awarded ZAC a $25 million sole‑source contract, a rare and powerful endorsement.
Most artificial intelligence systems today operate like massive pattern‑matching machines. To recognize a chair or identify a vehicle, they need to study tens of thousands—sometimes millions—of images during training. That demands enormous data center capacity and specialized chips that consume huge amounts of power.
People were interested in these podcasts
ZAC’s technology is built on a different philosophy. Instead of memorizing patterns, it understands concepts—the essential qualities that define an object or idea. This allows it to learn from dramatically fewer examples. Rather than millions of samples, ZAC’s system typically needs just five to 50.
The difference is more than academic. It means ZAC’s models can run quickly and efficiently on everyday computers rather than on giant server farms. It also means they can explain their decisions, something traditional AI systems struggle—or fail—to do.
In an era when AI is influencing decisions in medicine, national security, transportation, and finance, the ability to understand how a system reached its answer is essential. Trust depends on transparency.
ZAC’s partnership with the U.S. Air Force didn’t begin with the recent contract. It stretches back several years, including a 2019 project in which the Air Force funded ZAC to develop advanced 3D image‑recognition technology for unmanned aerial vehicles. The goal was ambitious: identify complex objects from any angle using only a small number of training images.
Traditional AI methods struggle with tasks like this unless they’re fed enormous datasets and run on expensive, power‑hungry hardware. ZAC showed it could deliver the same or better results using a fraction of the data and running on low‑cost computers. Over the years, the company has continued publishing updates demonstrating consistent progress, reinforcing its reputation as a team focused on real engineering rather than hype.
Advertisement Advertisement
That steady execution—and the unique capabilities of ZAC’s approach—helped secure the Air Force’s latest investment.
The United States is in a global competition for AI leadership, and winning will take more than massive data centers in a handful of states. It will require AI systems that are affordable, explainable, energy‑efficient, and deployable everywhere—from satellites to submarines, from smart homes to manufacturing floors.
Right now, much of the AI industry is moving in the opposite direction. Some large companies are planning to spend hundreds of billions of dollars on new data centers, electricity, cooling systems, and specialized computing hardware. This “bigger is better” model has severe economic and environmental costs—ones that will only grow.
ZAC represents another path. Because its technology runs on ordinary hardware, it can be deployed widely at low cost. Because it learns quickly from small datasets, it avoids the sky‑high expenses of collecting and storing massive amounts of information. And because its decisions are explainable, it can be trusted in high‑stakes environments.
If America wants AI that can be used in “a thousand places” instead of a few high‑tech hubs, the country needs systems that scale down as well as they scale up. ZAC offers that advantage.
Advertisement Advertisement
There is also a patriotic dimension to this story. In 2018, ZAC was offered $30 million by a group of Chinese investors, on the condition that the company relocate to Nanjing. Many startups would have accepted. But the Tadayon brothers declined. As CEO, Dr. Bijan Tadayon put it, “America gave us a second chance after the Iranian Revolution. We owe it to our adopted country.”
At a moment when China is investing heavily in AI dominance, ZAC’s commitment to remaining an American company is both principled and strategically important.
The Wright brothers changed history not by building the largest machine, but by reimagining how flight could work. ZAC is attempting something similar: Proving that AI doesn’t need to be massive to be powerful, and that understanding and efficiency matter as much as raw scale.
With a growing record of real‑world results, a long‑standing partnership with the U.S. Air Force, and a technology built for trust and practicality, ZAC has earned serious attention.
Advertisement Advertisement
As America looks toward the next decade of AI, the breakthrough it needs may not come from Silicon Valley but from a small, determined team in Maryland charting a new paradigm.
• S. Rob Sobhani, Ph.D. is an Adjunct Professor at Georgetown University and Board member at ZAC