About

Artificial intelligence is accelerating construction in the same way it is transforming software engineering — but construction is a high-liability, safety-critical domain. Errors are not bugs; they are structural failures, regulatory breaches and long-term risk.

This session explores the distinction between essential complexity (life safety, regulation, multi-disciplinary coordination) and accidental complexity (manual QA loops, file fragmentation, version confusion). It argues that AI should eliminate friction, not obscure responsibility.

Introducing concepts such as context compression, deterministic orchestration and structured document ingestion, the talk demonstrates how AI systems can be architected to reduce hallucination and drift in complex built environment projects.

Using the graph-native newBIM platform as a case study, the session shows how continuous compliance and structured governance can coexist with AI acceleration preserving human judgement while improving productivity.

Takeaways

• Understand the difference between essential and accidental complexity in construction.
• Learn how AI can reduce friction without increasing systemic risk.
• Discover architectural patterns (context compression, deterministic workflows) for safer AI deployment.
• See how graph-native systems enable continuous compliance in high-liability domains.