AI VS ESTIMATING: are we panicking? 🤔

Everyone’s asking: “Will AI change the way we estimate projects?” and there’s a lot of talk about AI “taking over the world.”

Even in our team, we’ve tested some of the latest tools to see whether they actually improve productivity and deliver real value to clients. We conclude: AI has clear strengths, but also clear limits.

AI is well-suited to “cookie cutter” projects, where cost rates and design elements don’t vary much. For this type of work, AI can already assist with take-offs and estimating with decent accuracy.

But when it comes to complex projects, its capabilities are still very limited. There are too many variables, assumptions, and context-specific nuances that AI simply doesn’t understand without heavy prompting and oversight.

Equally, AI (specifically large language models, or LLMs) doesn’t “think” the way people do. It generates answers based on probability patterns in language, not lived experience or genuine reasoning. That means it may give you something that sounds convincing but is actually wrong.

You always need an experienced professional to validate the outputs.

This raises the question of accountability. When people make mistakes, we can trace the issue, refine processes, and assign responsibility. When AI gets it wrong, the risks land squarely on the business without the same transparency or recourse.

Where we see AI flourishing is in pattern recognition and support tasks. For example, automating simple take-offs and estimates. As well as assisting in bid writing and strategy by highlighting risks or opportunities that might not be obvious at first glance.

So while full automation of complex estimating is unlikely anytime soon, AI is already becoming a valuable support tool provided it’s guided, validated, and overseen by experienced estimators.

#AI #Construction #Estimating #Preconstruction #DECProjects