The future of building design isn’t AI versus humans — it’s AI working with AEC specialists to scale and preserve expertise, writes Amar Hanspal, CEO Motif
The question everyone in AEC is asking — how will AI change building design? — is generating two confident but opposite answers. One camp says general-purpose AI agents will simply “solve” buildings from simple inputs, that anyone will be able to vibe-code their way to a building design. The other insists that the complexity of real buildings will keep us anchored to the geometry- and document-centric workflows we’ve relied on for decades. Both are wrong.
The future isn’t AI replacing human expertise, and it isn’t human expertise ignoring AI. It’s a genuine collaboration between the two — where the accumulated wisdom of experienced practitioners is captured, codified, and amplified by intelligent systems purpose-built for building design. That’s a more demanding vision than either side wants to admit, and getting there requires rethinking not just our tools but what we expect a platform to do.
Start with what firms actually have. Every architecture practice, every structural consultancy, every MEP engineer sits on decades of hard-won intellectual property: standard details, design rules, preferred assemblies, lessons learned from a thousand projects. That accumulated knowledge is their competitive advantage — it’s what makes them worth hiring. But almost none of it is captured in a form that software can use. It lives in people’s heads, in email threads, in markups on drawings that get filed and forgotten. When a senior designer retires, their judgement walks out the door.
Discover what’s new in technology for architecture, engineering and construction — read the latest edition of AEC Magazine
👉 Subscribe FREE here
The real unlock of AI in AEC isn’t generating generic geometry from a chatbot prompt. It’s giving firms a way to encode their best thinking into a platform and then collaborate with it — a dialogue between designer and system that draws on your firm’s specific intelligence to arrive at the right result. Methods developed once can be reused. Good ideas don’t just spread — they scale. This is the path from individual genius to institutional memory, from drawings to systems, from services to something closer to software.
That vision demands a platform that deeply understands building design. AI is moving up the stack — from geometry authoring to design intelligence — and the platform has to move with it. A general-purpose foundation model can do remarkable things, but it can’t on its own navigate the interdependencies of a real building, where a structural change triggers weeks of coordination across disciplines, consultants, and deliverable sets. Firms don’t need AI to draw faster. They need systems that understand the ripple effects of a design decision across an entire project and can orchestrate the response. Domain-specific intelligence has to be embedded in the platform itself, not just draped over it.
Not every task needs that full platform, and that’s fine. The number of AEC workflows that have never been touched by good software vastly exceeds the number that have. Software is getting radically easier to build, and AI-assisted tools — even vibe-coded ones — can deliver real value for these workflows right now. We are nowhere near meeting the demand for what software can do in this industry, and the barrier to entry is finally collapsing. That’s a genuine liberation.
But there are also systems of record — BIM, ERP, contract management — where professionals stake their licences, their liability, and their livelihoods on the data being correct. These demand a level of trust and reliability, and that AI-generated software simply cannot provide today. What the industry needs is a modern platform that serves both: a professional’s auditable source of truth and the foundation for an ecosystem of AI-native applications. That platform needs machine-readable geometry, version-controlled building state, and robust APIs. It needs to be agent-native from the ground up.
Firms don’t need AI to draw faster. They need systems that understand the ripple effects of a design decision across an entire project and can orchestrate the response
That’s where the current generation of tools breaks down. Legacy authoring environments were architected decades ago for a single human operator drawing on a digital sheet. They cannot be retrofitted into agent-native platforms. The data models are wrong, the APIs are afterthoughts, and furiously renaming the product or wrapping it in a cloud shell doesn’t change the DNA. Worse, the business models are hostile to the shift the industry needs: licensing terms that claim broad rights over customer data, proprietary formats that trap institutional knowledge, and pricing structures designed to extract rather than enable. The platform that hosts your data will determine whether your accumulated expertise compounds for your benefit — or someone else’s.
This is why we believe it’s time for the “I” in BIM to evolve — from Information to Intelligence. We call this Building Intelligence Modelling: a platform where modelling isn’t the job, it’s the outcome of a conversation between humans and intelligent systems. Design becomes a dialogue. Knowledge is shared, not just geometry. And the firms who built that knowledge remain in full ownership of their work and their wisdom.
That’s not a faster horse. It’s a different animal entirely
Amar Hanspal will be presenting at AEC Magazine’s NXT BLD 2026 conference in London 13-14 May.
