AI is the technology du jour. Sales teams are begging for it, C-levels are insisting on it, and customers are demanding it. That’s not to say it’s a fad, because it can deliver tremendous value, but too many companies are rushing to incorporate it out of fear of being left behind in a perceived technological arms race. With the ability to sift through large data sets and identify relevant trends, AI can significantly benefit companies looking to build out their capabilities by providing actionable insights that drive informed decision-making and strategic initiatives. However, companies should remember that AI can sometimes complicate matters if not used properly or if implemented hastily.
Sometimes, we do too much when enough would have sufficed or even been great. Our team learned this lesson while exploring large language models. We thought we were starting with our customers in mind, but we realized we were overcomplicating things.
We knew that fleet managers were sitting on mountains of data that were becoming increasingly complex and that AI could help organize it all. We implemented an insights engine as a proof of concept to ascertain the operational questions fleet managers were trying to answer. We then used our data and leveraged LLM to help users ask a wide array of questions and quickly answer them. We also wanted to explore use cases with generative AI, as there was a lot of demand for that.
But before long, we realized many fleet managers more often just wanted to know answers to basic questions. They didn’t need answers to all of these other complex questions, even if they said it might be interesting. So, we pivoted our approach, tailoring the technology to the immediate needs of the customer, even if the tech could do more.
Read more: How AI is becoming the ultimate shop assistant
Our learning curve involved optimizing when and how to use AI best for our customers. However, AI can be misused, leading to mishaps like the following. A North American airline introduced an AI-powered chatbot that incorrectly informed a customer he could receive a partial refund on his next-day bereavement ticket and the return ticket, provided he applied for the discount within 90 days. Unfortunately, that information was incorrect. The airline’s policy, as stated on its website, was to provide bereavement discounts only if the customer applied in advance. This snafu led to a civil lawsuit and negative publicity. The point is not to fault AI itself, but to illustrate the pitfalls of applying the technology prematurely or impulsively without fully understanding the consequences, both intended and unintended.
On a more positive note, The World Bee Project is using AI to save the bees. The organization is gathering data about bees through IoT (Internet of Things) sensors, microphones, and cameras on hives. AI then analyzes this data to identify patterns and trends that could help identify opportunities for intervention to help bees survive. AI expedites the sharing of real-time information on a global scale with the goal of taking action. As critical as AI is to the success of this project, expertise about bees is also needed to evaluate opportunities, highlighting another truth about AI: it’s not a set-it-and-forget-it tool. We must continuously evaluate its findings and adjust its parameters. Human oversight is key.
AI is an incredible tool, but it’s not always the right tool. It can enhance solutions and supercharge efforts, but some teams are over-indexing on AI. It’s important to remember that AI is computationally intensive and expensive. AI server cooling consumes a significant amount of water, which has environmental implications. This may change with the advent of quantum computing and its accompanying surge of processing power, but for the foreseeable future, AI’s cost and impact on internal labor resources must be a consideration when designing solutions. In addition, some applications of AI are simply not suitable for providing the best customer solution. Start with the challenge, not the solution, and the right tools will present themselves.
April Durrett is the global head of Experience at Verizon Connect, where she leverages her extensive expertise in product design, service design, and research to drive innovative user experiences such as Verizon Connect’s award-winning AI Dashcam. Throughout her career, April has honed her skills in transforming complex challenges into seamless solutions. Prior to joining Verizon Connect, April held senior design leadership positions at high tech companies such as Fjord @ Accenture, DataRobot, and Enterprise @ Deem. Her diverse background encompasses work in SaaS, AI, and sustainability, enabling her to bring a unique perspective to emerging technology-enabled products and services. April holds a bachelor’s degree from Pratt Institute and an MBA from Trinity College Dublin.
