Artificial Intelligence: Evaluating Risks and Rewards

Throw “AI” into Google search and you get a mere 249,000,000 results. That’s a lot of information. Some IEC members have already implemented AI technology into their daily work, some are deep into investigation, and others are beginning to contemplate these tools for their business. In March 2024, the Forbes article, Illuminating The Future: The Electrifying Impact Of AI On Electricians, stated: 

AI is redefining electrical work, a field traditionally reliant on manual expertise and human judgment. The integration of AI isn’t about replacing the skilled electrician; it’s about augmenting their capabilities with the precision and efficiency of AI. Today, AI emerges as a tool that enhances the expertise of electricians, allowing them to deliver services that are not just effective but also futuristic. 

One thing is clear: AI is here to stay. Among the findings reported in an October 2023 report from the Small Business and Entrepreneurship Council (SBEC): 

  • 75% of the surveyed small businesses (507 out of 676) utilize AI tools in their operations. 
  • AI adoption increases with revenue: 66% of businesses with less than $100K in revenue use AI, while 86% of businesses with over $1 million in revenue employ AI. 
  • Key drivers or reasons small business owners adopted AI include: their own research, peer influence and comfort with technology (54%), time and cost savings (51%), competitive pressures (29%) and inflationary pressures (26%). Small business owners also noted that high labor costs (25%), augmenting skilled labor or lack of access to skill labor (19%) drove AI adoption. 
  • The median annual business investment in AI-powered tools is $1,800. 
  • 93% agree that AI tools offer cost-effective solutions, driving cost savings and improved profitability. 
  • Nearly half (48%) of small business started using AI tools in the past year, and 29% have been using them for one to two years. 


Research and Evaluate 

Ros Bayley, vice president practice leader at Hiscox US, has more than 25 years of experience in the insurance industry and recently has focused her efforts on industrial, contracting, and construction. She helps her customers mitigate and protect themselves against risk. This includes making recommendations regarding AI use.

“AI is a technology the same as any other technology and it’s subject to the same risks,” Ros says. “Don’t be afraid of it just because it’s something new.” 

Ros says Hiscox construction customers are using AI quite effectively for increased productivity, backoffice support, and enhanced design applications. Her first piece of advice when considering AI applications for your business: do NOT rush into a purchase. 

“You need to do your due diligence and not rush into an AI technology purchase,” she says. “Make sure the application you are buying actually speaks to what you do.” 

She advises deep dives into the choices available and taking full advantage of the expertise of the people at the companies making the tools. 

“Sit in on the webinars, and use those sales opportunities as learning opportunities,” Ros says. “Find out what other people have to say about that company’s software. Then join a webinar from another company and learn about their products. You can’t rush learning; you’re going to have to go through a couple of these to get your head wrapped around the technology, to get your head wrapped around the opportunity. The best place to learn is from the people who are actually making the tools.” 

Next steps? Ros says to look at the reviews and see who their customers actually are and who is buying at the moment. See if there is anyone who seems like you in terms of business model and size. 

“Get referrals,” she suggests. “Ask the company if they will connect you with one of their customers so you can have a discussion with them.”  


Protect Yourself 

As an insurance provider, Ros has numerous tips to offer — some which must be considered before any purchase is made. Others kick in once you implement the new tool into your business.  

Contracts: Pay attention to contracts. Ros says the biggest thing to look at is limitation of liability as this defines the amount of liability for which you can be held responsible. This is where an insurance claims department looks when attempting to collect for its customers. 

“Remember, it works conversely as well,” Ros states. “Does the AI provider also have that limitation in their contract and is our customer comfortable with that? We recommend to our customers that they use a contract lawyer to review any contract before signing.” 

Insurance: Number one is errors and omissions (E&O) and, specifically, E&O that covers technology use. Make sure not only that you have the right insurance but that the people you are buying from have the right insurance. 

“It’s great having a contract that makes the seller liable for something but two million dollars is not going to be forthcoming if they don’t have the right insurance,” she says. 

Training: Train your people and train the AI. Treat the AI as your intern; believe that it knows nothing and then train it the way you want it to react. AI software is not plug and play; it is definitely something that needs to be trained by learning from your data and the way you work as well. 

“We always want to be sure our customers are asking the AI the right questions,” Ros explains. “When you give it instructions or ask it questions, make sure that you phrase it correctly because junk in equals junk out. AI learns iteratively and it’s not going to know whether those questions are right or wrong; it is just going to do exactly what you told it to do.”  

Review output: Just as you took time to research and evaluate before making your AI purchase, take all the time necessary to be sure the software is performing. 

“Don’t become complacent and don’t become over-reliant,” Ros instructs. “You still need to double check to make sure the output is exactly as you expect it. Institute policies of random regular checking and stick to it.” 

Ros offers a couple examples of why the above is so important, based on problems they anticipate contractors may run into that causes them to come back to Hiscox with a claim. The field is new, and there is very little precedent in the courts and very few case studies. 

“For example, AI likely allows contractors to push four bids out a day as opposed to one,” she says. “The AI provider has told them they need to run software updates every Monday, and one Monday the contractor forgets to do it. The company issues 15 bids that week based on the old data. They win one of those jobs and that bid is based on old information. It’s based on a lower price. The contractor goes to do the job and realizes ‘oh, I actually haven’t charged enough’ and wants to upcharge the customer. We all know where that ends up. Either the contractor has to take a loss on the job or try to get the money and there’s a potential for lawsuit. All because they didn’t run an update.” 

Ros provides another example. Her client buys AI that isn’t necessarily conversant with the construction industry and the metrics are not quite right.  

“They’re doing a distance warning system to make sure that no systems overheat,” Ros says. “There is a mix-up in language, for example using Celsius vs. Fahrenheit. The wrong temperatures hit, the machine overheats, no flag is given, no alarm is sent, and the warehouse burns down.” 


Get in the Game 

“There are so many opportunities for contractors to use AI and to use it safely,” Ros concludes. “There are massive cost-saving and time-saving benefits allowing more time for you to do the work you love to do.” 



The Status Quo Might Be Right for You 

While Ros believes AI is a route to growth, she acknowledges that some companies may not be looking to increase their business.  

“There will always be a need for small contractors — the ones people come to when they want jobs done by someone who is trustworthy, someone who is reliable, and someone they know,” she says. “But every electrical contractor needs to understand AI, especially as it relates to smart home and smart business customers. Those customers are likely to look for contractors who are in that technology and it’s a good connection. Even a smaller electrician that’s not using AI in the back office or with BIM still has to be educated. It’s going to become an expectation.”