In my previous column for Fast Company, I outlined critical initial steps that companies should take as they strategize how to put AI/ML and location intelligence to work to address their most complex business challenges. At many organizations, there is significant internal and external pressure to be an early adopter of AI, but this is uncharted territory for most. To help companies take those first steps in an informed way, I am sharing lessons from work we are doing on implementations for the earliest of early adopters using AI/ML and location intelligence.
My prior column asked and answered two critical questions your team will face: Does ChatGPT need a helping hand? And where should your data live? In this column, I will continue the discussion by answering three more questions that every organization should ask itself. My answers to each are based on real-world experience supporting companies that are several steps ahead of the game when it comes to harnessing the power of AI with the insights that can be extracted from location-based data.
THERE ARE SO MANY USE CASES. WHERE SHOULD I BEGIN?
This is a question I am hearing from senior leadership at many organizations. It’s an important question with any new technology, but I think it is particularly critical with AI because the potential scale of investment is so large. The magnitude of investment can be eye-popping because of the costs associated with building and training large AI models, building out the applications, and rolling out those major implementations across an organization.
To mitigate the risks and costs of doing AI on that scale, I have been advising companies to take a use case-driven approach to AI. Rather than letting the wide-ranging possibilities of AI drive decisions right now, the smarter strategy may be to start by making a short list of the thorniest operational challenges that other technologies have not yet been able to solve. Then start talking about how AI and location intelligence could be put into action as a potential solution.
In my experience, this approach quickly narrows down an endless list of potential paths forward into a set of highly defined projects that have a significant impact while building your organization’s experience with AI/ML—all while mitigating the risks and costs that organizations are taking if they go too big too soon.
IS MY DATA READY?
Data quality is critical to the success of an AI/ML implementation, but it is an issue that is too often overlooked when organizations are mapping out their strategy for putting AI to work. Unlike other kinds of data that may largely exist within a single system (financial metrics, for example), location-based data lives in multiple systems across an organization. Because these systems are so often disconnected from one another, data likely exists in silos that would undermine an AI initiative.
In addition to being siloed, location-based data also often varies dramatically in accuracy and quality. For example, one data point just captured by a field worker using a mobile tablet with 3D scanning might check every box for accuracy and precision. But another data point may be from a legacy system based on handwritten notes on paper maps and reports that are years or decades old. Is that information accurate enough to drive the insights you need? Does it have the level of detail you need? Inaccurate, poor-quality data cannot deliver the results that companies want from their AI investments.
That is why I recommend that every organization conduct a comprehensive assessment of its location-based data that determines where it lives, what format it is in, and how accurate it is. With the latest technologies and techniques, you can rapidly and dramatically improve the quality and richness of this data so that it is ready when you need it to be.
HOW DO YOU SHRINK THE GAP BETWEEN EXECS AND THESE AI INSIGHTS?
Historically, there has been a chasm of distance between location-based data and executives. In the past, small geospatial departments were necessary as intermediaries because of the complexity of geographic information systems (GIS). Those professionals will continue to be invaluable for organizations, but the vast gap between execs and this data can be dramatically reduced using AI.
Natural-language AI interfaces can make it possible for people without advanced geospatial skills to interact directly with this data and extract insights. Companies get this intuitively when they strategize about ChatGPT as an interface with this data, but the temptation is to think too small. It’s not just about putting this data at the fingertips of field crews and operations teams and department managers. All of that is great, but the impact will be even greater if you plan from day one to also build this with senior leadership as core users.
To illustrate how powerful these insights can be, imagine executives asking the following questions via a chat interface:
- “How many vehicles do we currently have deployed in Blake County?” to find out how much a new tax on capital equipment in the county could impact their bottom line.
- “Where are my company’s assets most vulnerable to wildfire risk?” to assess risk seeing news reports on fires raging across the West.
- “What was our 2023 spend on construction projects in the 8th Congressional District?” to prepare for a meeting with an elected official.
The answers to all of these questions could take days or even weeks for an analyst to research and summarize. With the power of generative AI and conversational (chat) interfaces, the answers are at the company’s leadership’s beck and call. This not only drives faster, better decision making in the C-Suite, but also expands accessibility of critical insights for those with limited technical skills, subject matter literacy, and varying abilities (e.g., sight, dexterity).
Organizations that ask themselves these critical questions will reduce the risks of their initial steps into harnessing the power of AI/ML and location intelligence. It will also establish the right foundation for successfully scaling these programs over time.