Using AI as a tool, not a definitive

By Emma Peterson. 

Trent Cotney breaks down where AI is useful; and where it can lead you into trouble.  

It seems like artificial intelligence (AI), specifically generative AI, is the only thing anyone is talking about these days. To learn more about its role in roofing and construction operations, Megan Ellsworth sat down with Trent Cotney of Adams and Reese for a Coffee Conversation discussion of what contractors need to know about the technology as it continues to evolve.  

One of the most common places you see companies, whether in the trades or beyond, using AI is in customer service and marketing. But those aren't the only places it can be used. Trent explained, “AI is great in marketing and service, it saves time and money. But where it really is transforming the industry is in estimating, material management, scheduling, forecasting and job costing. That’s where it’s really giving people a competitive advantage.”  

How exactly does AI play into these areas? A great example is scheduling software. Trent shared:  

I had the opportunity to see some scheduling software that used AI and what it did was look at the efficiency of each crew member on each crew with each system, and determine, based on their upcoming projects, what crew should be matched where for most efficiency. And then if, for some reason, a project is suspended or ends, it automatically fixes that. So, the amount of efficiency that you are gaining on that equals money. And if you're not using it, someone else is and you can’t adequately compete.  

It is important to note that even with incredible benefits like this, AI is not a perfect solution for everything. Trent highlighted three main reasons to be cautious about AI – accuracy, risk and compliance. He elaborated:  

AI is a lot like a golden retriever, it wants to please. But it will come up with stuff in order to do that. I can think of at least three cases where lawyers used AI, it came up with false citations and those lawyers were sanctioned. So you have to independently verify it for accuracy... As for legal compliance, you need to remember that your searches on these open AIs, like ChatGPT, are available, even ones that get deleted. It’s all publicly discoverable so you have to make sure no one is uploading confidential information that puts you at risk.  

Trent’s best advice for using AI effectively while managing these concerns? Guardrails. He explained, “The best thing you can do from a compliance standpoint is put up some guardrails about its usage and make sure everyone knows its limits. AI is a tool and the issues occur when people use it as a definitive. You have to put eyes on its outputs because it won’t spot issues that aren’t there. If you put an issue in front of it, it can solve that, but it doesn’t know the holistic picture. It’s a tool and that’s how you have to use it.”  

Listen to the rest of the conversation or Watch the recording to learn more about safe and effective AI use in the industry.