In early November, I attended UA Week, an event created by Utility Analytics Institute, a sister brand to T&D World within Endeavor Business Media's energy group. One of the event's focal points was Generative Artificial Intelligence (GenAI). Prior to this event, most of what I'd heard and read about GenAI centered on the risks and dangers associated with it. I had heard little about its potential and positive impacts, especially in the electric utility industry, until I attended UA Week.
The most in-depth GenAI discussion I encountered during the week featured a panel of four data science experts, three representing large investor-owned utilities, and one representing a software consulting company. Given that utilities are often considered conservative and slow to change, I was surprised to learn that these three utilities are embracing GenAI. Therefore, I'm using this month's column to simply share their thoughts and insights with you.
Andy Quick of Entergy, Andy Kapp of Evergy, Valli Mithukaruppan of Exelon, along with David Bess from technology company Neudesic, were eager to share their views on GenAI. Evergy's Andy Kapp stated that GenAI will have an impact on society comparable to the introduction and adoption of computers. He also believes it will fundamentally change the way work is done and cautioned that companies, including utilities, risk losing their talent if they don't embrace it.
David Bess agreed with Kapp’s computer analogy and added that GenAI will improve and advance by “magnitudes” every year for years to come. He believes the technology will be adopted at home before it enters the workplace, like apps were adopted when they first became available. Bess stressed that utility employees must be trained to use GenAI and that it must be effectively managed.
Valli Mithukaruppan of Exelon echoed many of Kapp and Bess's sentiments, emphasizing that utilities must learn to scale the technology and stay ahead of the curve as the adoption is progressing rapidly.
Andy Quick mentioned that Entergy is actively working on a comprehensive GenAI strategy. He anticipates that much of the innovation supported by GenAI will come from the employee level. Quick said that Entergy plans to educate and train employees on the technology and its potential, followed by a rollout of the technology. He predicts that "enterprise GenAI" will transform the way work occurs within the utility, shifting some tasks from human workers to "digital workers." Entergy's data science group is currently developing a new AI governance plan and organization, which will soon be proposed to Entergy's board.
Kapp shared a similar message, describing GenAI as employees' new co-worker. Evergy hasn't released a formal training program for employees yet, but Kapp and his fellow data analytics and science experts have begun offering "lunch and learns" on the topic. In addition, a GenAI tool called Co-Pilot has already been adopted by Evergy's coders
“In March and April, we were really bullish to get GenAI rolled out, but I pumped the brakes on that a little,” he said. That doesn’t mean, however, that Evergy is not using any GenAI technology. Kapp’s group has layered Chat GPT on top of some current tools, and employees are testing it for accuracy and other features. Kapp pointed out that Chat GPT has been used for tasks traditionally performed by company employees, freeing up their time for more complex assignments.
In general, the four experts expressed more positive views than negative ones about GenAI. However, they all agreed that the technology raises valid concerns. Some UA Week audience members voiced apprehensions about data privacy and security. Bess, who works with many utilities, acknowledged the importance of this concern but suggested that utilities leverage their existing data privacy and security practices rather than reinventing the wheel. He also highlighted another concern, which is that GenAI is likely to render entry-level and junior coders obsolete. Bess said some universities are already seeing students leave programs focused on coding and programming, potentially leading to a shortage of senior-level coders in the future. Kapp spoke about his concerns with some of Evergy’s solution providers adding GenAI or similar AI components to the products and services they provide without consulting with Evergy’s staff. He emphasized that it's important for vendors to engage with the utility before making changes to their systems.
The four individuals participating in this panel discussion believe that GenAI should be embraced. However, results of an audience survey conducted during the panel session reveal they might be outliers. The survey showed that 45% of the companies represented in the audience don’t have a GenAI strategy, and 72% have no unique governance structures for AI. For the 55% of companies that are working on some type of GenAI strategy, nearly half (44%) said their companies’ internal data analytics team is driving that initiative. One additional polling question asked audience members whose companies are using GenAI, what common use cases are being piloted. The top answer was coding, followed by supply chain and rate cases.
While the four panelists were enthusiastic about the positive impacts of GenAI, they acknowledged that well-trained and knowledgeable individuals are essential to realize its benefits. They also cautioned that GenAI must be thoroughly understood and effectively managed to ensure it doesn't produce unintended consequences. Bess encouraged the audience to empower themselves to learn about GenAI and reminded them that it is not the first technology disrupter. Mithukaruppan recommended creating a long-range plan to guide the adoption of GenAI. Likewise, Quick also emphasized the importance of creating a GenAI strategy, stressing it should be created intentionally rather than letting it evolve organically. Kapp stressed that GenAI is a co-pilot, not a pilot, and will not replace humans in the workplace. He also told the audience to adopt, adopt, adopt…don’t resist.
