The world of design and manufacturing is being quietly but powerfully transformed by a new wave of artificial intelligence.
Helping lead this revolution is Paul Powers, the founder and CEO of Physna, a company whose name is derived from “physical DNA.” In a recent interview, Powers offered a glimpse into how Gen AI is being used to decode the complexities of 3-D modeling, creating tools that not only understand the physical world but help shape its digital twin with precision.
Unlocking the Language of 3-D Geometry
Physna interprets 3-D models and translates them into code. This translation forms the basis of a novel approach where geometry is no longer locked in static visualizations but becomes a searchable, comparable, and analyzable data form.
Powers describes this as building a “bridge between what’s physical and digital,” allowing machines to understand shapes and structures as intuitively as they parse text.
At the core of Physna’s value proposition is its ability to extract more meaningful data from 3-D models than traditional systems. By transforming these models into a format readable by AI, the company has developed capabilities far beyond surface-level geometry. Their tools can decipher internal structures, assess proper scale and measurements, and ultimately enable both analytical and generative AI applications.
From Insight to Innovation: The Lifecycle of an AI-Driven Project
Physna’s projects follow a sequence that begins with ingestion and analysis of a company’s 3-D model data.
“Instead of starting with a giant training dataset to make something generalized,” Powers explains, “we focus on data specific to a company or industry.”
This specificity is crucial. A part designed for aerospace functions vastly differently than one built for automotive applications, even if they might appear similar on the surface.
The early stages of a project are analytical. By understanding the relationships between a part’s geometry and its metadata — such as how it’s manufactured or how it performs in the field — Physna provides immediate benefits.
Companies can deduplicate parts, identify alternative suppliers, and streamline design processes by leveraging existing knowledge. This alone can significantly reduce costs and engineering cycle times.
Only after these foundational insights are established does generative AI come into play. Here, Gen AI doesn’t attempt to create entire planes or vehicles from scratch. Instead, it is used to optimize individual components — say, a more wear-resistant fuel valve under a specific weight threshold.
Powers emphasizes that reliable generative outputs stem from deep analytical understanding. “You need to understand what works and what doesn’t based on past data before you can responsibly generate new designs.”
Enhancing, Not Replacing, Human Ingenuity
The specter of job displacement often shadows AI discussions. But in Powers’ experience, engineers and designers inside companies aren’t threatened by these tools.
“They’re thinking, ‘You’re making my job way easier,’” he notes.
Physna helps these professionals move away from redundant tasks like redesigning already existing parts and toward higher-value activities such as innovation and optimization.
The fear of obsolescence, he argues, is largely external. Inside organizations, the reaction tends to be pragmatic. If a tool prevents engineers from inadvertently reinventing the wheel, it’s not eliminating jobs; it’s enhancing human capability.
“You’re making them able to make faster, more accurate, and better decisions more rapidly,” he says.
Breaking the Learning Curve with Intuitive Design
One of the clearest insights from Powers is his view on adoption.
“If we have to spend too long training somebody, we’ve messed up,” he says.
His goal is to make Physna’s platform so intuitive that even highly technical tasks become accessible without extensive onboarding. This is achieved by designing interfaces that mirror natural human interactions — point, click, and increasingly, converse.
He draws parallels with ChatGPT, noting how the lack of a traditional menu system allows users to engage in a more natural, conversational flow. While complex design tasks still require precision, the goal is to move toward an interface where users can communicate intentions directly and get results without elaborate instructions.
Trust, however, remains the biggest barrier. Not blind trust, but a willingness to “dare to try something new.” Powers sees this psychological leap as more significant than any technical hurdle.
Once users see the productivity gains, resistance fades.
The Future of Gen AI in Design
Looking ahead, Powers anticipates a future where AI systems become both more specialized and more generalized. It sounds paradoxical, but what he means is that domain-specific tools like Physna will offer generalized, intuitive interfaces.
Users won’t need to understand the inner workings of 3-D modeling or AI to get meaningful results, as they’ll simply describe what they need, and the system will do the heavy lifting in the background.
He also sees deeper personalization on the horizon. Gen AI tools, like language models today, will become increasingly attuned to individual users. Preferences, habits, and workflows will be learned and adapted to, allowing the AI to anticipate needs instead of just responding to commands.
And the physical-digital divide will continue to shrink. Physna is already experimenting with integrations involving augmented reality, providing real-time interactions with 3-D models. In the near future, your glasses might not only display a model but also predict what you’re about to ask about it, and take action accordingly.
Designing the Future, One Model at a Time
Paul Powers and his team at Physna aren’t just digitizing design — they’re reimagining it. By blending the strengths of analytical and generative AI, they’re creating a platform that empowers engineers to innovate faster, smarter, and with more context than ever before. In this world, design becomes less about drawing from scratch and more about evolving with insight.
As Gen AI continues to evolve, so too will its role in shaping the physical world. With leaders like Powers at the helm, the promise of AI in design feels less like a distant horizon and more like a rapidly unfolding reality.
