From automation to infrastructure
Before starting this course, I primarily understood Artificial Intelligence in the AEC industry as a productivity tool—something used to automate repetitive tasks, accelerate modeling, or optimize isolated steps within the design workflow. AI was, in my mind, an add-on: powerful, useful, but largely standalone.
Over the duration of the course and through exposure to real-world applications, this perception shifted significantly. Today, I see AI less as a discrete tool and more as an infrastructure for decision-making across the entire AEC ecosystem. Rather than simply making processes faster, AI has the potential to help us manage complexity, coordinate systems, and make informed decisions based on data rather than assumptions.
AI as a system not a replacement
one of the most impactful insights came from the seminar by Libny Pacheco, which emphasized that AI creates the most value when it’s embedded within real production environments. instead of replacing design intent or creative judgment, AI becomes a layer that supports scalability, coordination, and consistency across complex workflows.
This reframing is critical for AEC, where projects involve multiple disciplines, long timelines, and high levels of uncertainty. Successful AI adoption does not depend on isolated experimentation, but on deep integration with existing tools, standards, and processes—from BIM environments and construction logistics to facility management and performance monitoring.
Agentic AI and the Question We Haven’t Asked Yet
Another pivotal concept introduced during the course was agentic AI—systems capable not only of generating outputs, but of taking actions, coordinating tasks, and adapting to changing conditions. This idea became clearer through examples shared by the co-founder of GenSpark, where AI agents were deployed as operational tools across multiple fields. This raised an important question for me:
How do we meaningfully deploy agentic AI in the AEC industry?
While many discussions focus on AI for form generation or visualization, the real opportunity may lie in engineering systems, operational optimization, and lifecycle performance—areas where decisions are complex, interdependent, and data-rich, yet often managed through fragmented workflows.
From Design to Lifecycle Performance
This shift in perspective was further reinforced by examples shared by Andrea Paindelli from Veolia, where AI was used to improve efficiency while delivering measurable environmental impact. These applications moved the conversation beyond design phases and into the operational life of buildings and infrastructure.
For me, this was a turning point. AI is not only about making better drawings or faster simulations—it is about connecting design decisions to long-term environmental, operational, and economic outcomes. This expands AI’s role from a design assistant to a strategic system embedded throughout the building lifecycle.
Looking Forward: Predictive and Scenario-Based Decision Making
Looking ahead, I believe AI in AEC will increasingly support predictive and scenario-based decision making, enabling professionals to evaluate long-term impacts early in the design process. Instead of reacting to problems during construction or operation, teams can anticipate performance, test alternatives, and understand trade-offs before committing resources.
In this sense, AI becomes less about automation and more about responsibility—helping architects, engineers, and planners make better-informed choices in a world where environmental, social, and economic pressures are only increasing.
