There’s a quiet identity crisis running through the software profession, and Arthur Miller thinks it’s built on a misunderstanding.
For an entire generation, a technologist’s worth was measured in production. Code shipped, features delivered, velocity sustained. Then AI made production nearly free. Early enterprise deployments of AI-native delivery are compressing three years of work into six weeks, and the people whose careers were organized around producing are asking what’s left for them.
Miller’s answer, delivered in his new book AI-SDLC and the New Enterprise Architecture Approach, is that the most human part of the job was always the job.
Miller speaks from the far end of the career the industry aspires to. He was the enterprise architect behind Zelle and Real-Time Payments infrastructure, led R&D engineering at the Defense Information Systems Agency, served as CTO of the credit union lending platform equipifi, advised oncology clinical trial software, and now runs his own financial technology company, FiQIS.
Across all of it, he says, the scarce skill was never typing.
“The most impactful solutions emerge not from brilliant code alone,” Miller says, “but from cross-functional teams that trust each other, challenge assumptions with respect, and stay anchored to the why behind the work.”
His book formalizes that belief into a division of labor for the AI era: agents produce, humans judge. Machines generate the system. People decide what should exist, verify it was built right, and answer for it when it wasn’t. In his framework, every stage of the software lifecycle has a gate, and every gate has a human name on it, because accountability is the one thing that cannot be generated.
Read one way, that’s governance. Read the way Miller intends it, it’s a promotion for the entire profession. The engineer’s job stops being the production of artifacts and becomes the exercise of judgment, which was always the harder and more valuable work. What to build. Whether it’s right. Who it serves.
It also explains why a book about software architecture keeps sounding like a book about leadership.
“Leadership is not about being the smartest person in the room,” Miller says. “It’s about creating space where others feel safe to bring their brilliance forward. When we lead with empathy and purpose, innovation naturally follows.”
The trap he wants his industry to avoid has a familiar shape. “In fast-moving organizations, it’s easy to confuse speed with success,” he says. “True progress comes when our decisions are grounded in shared values, not just technical velocity.”
The stakes of that confusion are the book’s sharpest line: a 97 percent faster path to the wrong system is a 97 percent faster mistake.
Miller calls his larger project architecting trust at scale, and he’s explicit that the architecture includes the people. Systems designed with care and stewardship, teams where learning matters more than being right, careers that survive the machines by doing what the machines can’t.
“Technology is a powerful tool, but it’s people who drive change,” he says. “Our legacy isn’t just in what we build, but in how we treat the people who help us build it.”
The machines took over production. Miller’s argument is that they handed the profession back its judgment, and judgment was always the point.
