Artificial intelligence is rapidly becoming a standard feature of BIM software, but AEC Magazine argues that the bigger opportunity for architecture firms lies outside vendor platforms. Rather than relying only on AI assistants from Autodesk, Graphisoft, Bentley, and others, practices could turn their accumulated professional judgment into reusable intelligence they own.
The article explores this idea through ALPA’s open-source Skills for Architects project. Its repository contains 39 modular AI Skills covering tasks across the project lifecycle, including zoning analysis, occupancy calculations, Environmental Product Declaration parsing, specification generation, materials research, and furniture schedules. Unlike a simple prompt, a Skill captures the specific way a practice wants a task performed, combining instructions, reference material, context, and defined objectives.
Together, these capabilities form what the article calls the judgment stack. Vendor AI represents the first layer, providing generic automation available to all customers. The second layer consists of a firm’s own Skills, workflows, rules, and governance. The third is institutional memory, where AI can draw on previous projects, technical standards, specifications, BIM models, and lessons learned.
This approach could make decades of tacit knowledge more accessible. Expertise that traditionally resides in senior architects, office standards, project templates, and mentoring could become structured, version-controlled, and reusable. Agents could coordinate multiple Skills, while shared rules and validation hooks maintain consistency and check outputs.
The shift could also change BIM’s role. Authoring tools would remain important for geometry, coordination, and documentation, but professional intelligence could increasingly operate independently, accessing data across different applications and writing results back. Interoperability therefore becomes essential.
The article argues that competitive advantage in AI may ultimately depend less on access to foundation models or vendor features and more on a practice’s ability to codify its distinctive judgment. Firms that capture, refine, and retain their institutional expertise could transform professional knowledge into a lasting digital asset.