
Sign up to save your podcasts
Or


Dustin Schafer is Head of Research at Henderson Engineers, where he focuses on translating emerging technology into organizational strategy. In this episode, Dustin shares the career path that took him from architectural engineering and mechanical design into project management, commissioning, quality, standards, and ultimately technology leadership — including a recent shift from CTO into a role dedicated to the intersection of operations and technology.
The conversation centers on a core argument: as AI advances, it will dramatically reduce the cost of production work (routing, sizing, checking, documentation), but it won’t automatically create “good” design. Codes set a floor, not an answer — and without encoding a firm’s priorities, history, and decision-making, generic AI tends toward what Dustin calls “compliant slop”: technically valid outputs that fail to differentiate or serve real client intent.
Dustin explains why service firms need to articulate their work in terms that map to software methods (e.g., framing routing as a pathfinding optimization problem), and why AI should often be used as a translation layer between well-defined standards — not as a substitute for intention and care. He also dives into why “just start doing something with AI” can lead to shallow, edge-of-workflow wins, and why real leverage comes from working outward from business model and client outcomes.
Finally, Dustin reframes disruption through the lens of risk and value: clients don’t just pay for drawings — they pay firms to take on the risk of being wrong. In an AI-driven world, the opportunity (and threat) is in using these tools to take the right risks faster and more consistently — and repricing around outcomes, decisions, and differentiated judgment.
Links:
By Evan Troxel & Candace KitchenDustin Schafer is Head of Research at Henderson Engineers, where he focuses on translating emerging technology into organizational strategy. In this episode, Dustin shares the career path that took him from architectural engineering and mechanical design into project management, commissioning, quality, standards, and ultimately technology leadership — including a recent shift from CTO into a role dedicated to the intersection of operations and technology.
The conversation centers on a core argument: as AI advances, it will dramatically reduce the cost of production work (routing, sizing, checking, documentation), but it won’t automatically create “good” design. Codes set a floor, not an answer — and without encoding a firm’s priorities, history, and decision-making, generic AI tends toward what Dustin calls “compliant slop”: technically valid outputs that fail to differentiate or serve real client intent.
Dustin explains why service firms need to articulate their work in terms that map to software methods (e.g., framing routing as a pathfinding optimization problem), and why AI should often be used as a translation layer between well-defined standards — not as a substitute for intention and care. He also dives into why “just start doing something with AI” can lead to shallow, edge-of-workflow wins, and why real leverage comes from working outward from business model and client outcomes.
Finally, Dustin reframes disruption through the lens of risk and value: clients don’t just pay for drawings — they pay firms to take on the risk of being wrong. In an AI-driven world, the opportunity (and threat) is in using these tools to take the right risks faster and more consistently — and repricing around outcomes, decisions, and differentiated judgment.
Links: