The Geospatial Edge: Issue 13, Spring 2025

The Geospatial Edge is Esri Canada’s periodic newsletter for managers and professionals tasked with growing their organizations’ geospatial capabilities. In this issue, Matt Lewin asks what skills GIS practitioners need to cultivate to stay ahead in the AI era—and what geographic information system (GIS) programs can build that AI can’t easily replace.

I highly recommend reading "Strategy in an Era of Abundant Expertise" from the March 2025 issue of Harvard Business Review. If you've been thinking about generative AI and how it might disrupt your day-to-day work or your broader GIS strategy, it's a must-read.

The authors make a simple but powerful point: as GenAI improves, the expertise that used to set organizations apart is becoming widely available—to competitors, partners and even customers. It's not that expertise no longer matters. It's that the way we create value with it has to change. In fact, some organizations are already adopting an AI-first philosophy for many job functions—for example, Shopify's CEO recently telling staffers they need to prove jobs can't be done by AI before adding headcount.

That got me thinking about GIS.

If there's any field where expertise was once a significant barrier to entry, it's GIS. Collecting, analyzing and visualizing spatial data used to be complex, technical work that only trained specialists could do. Today, AI tools are at the brink of knocking those barriers down. Fast.

This raises some tough but necessary questions: What happens to GIS programs when users can do the basics themselves? What skills do GIS practitioners need to cultivate to stay ahead? And what can GIS programs build that AI can't easily replace?

My take on the three big questions every GIS program should be asking right now:

Q1: Which GIS tasks will users handle themselves with AI?

Traditionally, GIS programs have been responsible for a broad range of services: managing spatial data, running analyses, producing maps and supporting decision making processes. In the past, these services depended heavily on specialized software and deep technical knowledge, which, to a degree, protected a GIS team's role in an organization.

Today, those barriers are rapidly falling.

Across the geospatial spectrum, AI is already supplanting human expertise. Generative models can automatically discover and organize datasets from public and private sources, a task that once took analysts hours of searching and cleaning. Data preparation—historically one of the most time-consuming aspects of any GIS project—is now being accelerated by AI that can identify errors, harmonize datasets and even predict missing values with minimal human oversight.

For instance, when an analyst needs to update a city's land cover map, they might spend days sifting through data portals, downloading files, fixing projects and preparing data. Now, however, they can use an AI agent to search cloud repositories like ArcGIS Living Atlas, AWS Earth on Demand or Microsoft Planetary Computer. The agent can select the best imagery based on the project's area and specific criteria, such as cloud cover and resolution. Once it identifies the appropriate data, it manages all the tedious preprocessing steps, including cloud masking, band normalization, clipping and projection alignment.

Another example could involve conducting a site suitability analysis. Traditionally, an analyst would gather data layers, such as zoning maps, traffic flows, demographic data and environmental constraints. They would then manually weigh these factors according to project priorities. After building weighted overlays and running spatial models, they would produce a final suitability map. Instead of doing this manually, an AI agent developed with a tool like LangChain could handle much of the data planning and organization. It could then trigger a tool like PyLUSAT to run the suitability model. A user would simply prompt the agent with a natural language request, such as, "Find the best location for housing based on proximity to transit and low flood risk." And just like that, a suitability map would be generated!

In other words, the services users once relied on GIS teams to perform are increasingly available on demand, often at the click of a button. This means that, in the future, the real value won't come from performing standard tasks faster. It will come from tackling the complex, strategic and ambiguous problems that AI tools aren't equipped to solve—challenges that require human judgment, contextual understanding and nuanced decision-making.

Which brings us to question two:

Q2: Which skills does a GIS team need to evolve to stay ahead?

The shift in cognitive workload driven by GenAI doesn't mean GIS expertise is becoming obsolete. It means the nature of that expertise is evolving, and quickly!

A useful way to understand this evolution is through the DIKW model: Data ➔ Information ➔ Knowledge ➔ Wisdom. I've also posted about these AI advancements previously, but here's a simplified breakdown:

At the base level, data is raw and unprocessed — coordinates, satellite imagery and survey points. A step up, information is organized data that starts to have meaning, like a shapefile showing flood zones. Knowledge builds on information by adding context, experience and interpretation. For example, understanding not just where the flood zones are but why they matter to vulnerable communities. At the top is wisdom, i.e. the ability to make sound judgments, anticipate consequences and act ethically in complex, real-world situations.

For much of GIS history, professionals operated primarily at the data and information levels. Finding, preparing and analyzing spatial data was where much of the value was created.

Now, AI is rapidly taking over those lower layers.

This means the future of GIS expertise lies firmly in the upper layers with knowledge and wisdom. GIS professionals will need to curate data, not just collect it. They'll need to validate and question AI-generated outputs, not just accept them at face value. They'll need to frame better problems, guide organizations through uncertainty and connect spatial insights to real-world decisions that balance economic, environmental and social priorities.

In practical terms, this means investing in higher-order skills, such as:

• Critical thinking: Asking smart, strategic questions that AI can't anticipate

• Solution conceptualization: Developing use cases that address complex business uses centred around a geographic approach (try this)

• Interpretation: Seeing beyond the patterns in the data to understand what they mean in context

• Ethical judgment: Leading conversations about privacy, fairness, bias and the responsible use of spatial data

• Strategic communication: Telling compelling, clear stories that move decision-makers to act

The benefit to GIS programs that focus on higher-order skill sets will primarily be a shift in perception—from being viewed as a service provider to becoming a trusted partner. Instead of merely being assigned tasks like "create this map" or "run this model," you'll have the opportunity to engage in discussions earlier in the process. This engagement will help leaders understand which spatial questions are truly relevant.

Rather than just producing outputs for others to interpret, you will take on the role of guiding interpretation, challenging assumptions and framing decisions through a spatial lens. As a consultant myself, I am often asked how to elevate GIS on senior leadership's radar. Perhaps AI will be the catalyst that makes this happen!

Q3: What assets can we build to enhance our ability to remain relevant?

As AI gets more sophisticated, GIS managers face a tough reality: skill and expertise alone won't protect your team's relevance. To stay essential, GIS programs need to build assets that AI can't easily replicate. Think of these assets as intellectual moats—durable advantages that protect your team's value over time.

I can think of three:

Authoritative data. Most AI systems are trained on open, generic datasets, the kinds of data that are readily available but often lack depth, specificity or context. These models are excellent at recognizing broad patterns, but they struggle when nuance matters, when small local differences, historical subtleties or regulatory complexities change the meaning of spatial information.

That's why owning or curating high-quality, authoritative spatial data is one of the strongest moats a GIS program can build. It's not just about collecting more data. It's about creating datasets that are richer, more accurate, more timely and more deeply connected to the realities of the organization or community you serve. For a city, for example, data reflects the local "ground truth"—not just current zoning maps but the political forces that shaped urban boundaries. It weaves in domain-specific insights like how soil composition affects land valuation or how informal transit networks really operate in a city. When GIS teams build and maintain these kinds of specialized datasets, they create a strategic asset that AI can't easily copy or replace.

Organizational fluency. Every company or organization operates within a web of political dynamics, cultural norms, legacy systems and informal networks. These forces aren't written down in any dataset. They're invisible, but they're powerful, and they often matter more than the technical solution itself. AI might be able to suggest the "optimal" solution on paper, but it won't understand the undocumented realities that will determine if it succeeds or fails inside an organization. GIS managers who know how to navigate those realities and who understand where GIS can create real impact and how to deliver it in a way that resonates with decision-makers bring a kind of wisdom that AI can't replicate.

It's not just technical expertise; it's knowing how to make GIS matter in your specific environment.

Creating and brokering GIS-AI agents and ecosystems. GIS managers who develop the skills to design, train and orchestrate specialized AI agents (lightweight AI assistants built for specific geospatial tasks) will put their organizations on a very different footing. Instead of relying on off-the-shelf models that treat every organization the same, they'll build bespoke AI ecosystems that reflect the unique goals, constraints and realities of their environment. Imagine having agents trained on your spatial data and you becoming the orchestrator of those agents to the extent that your team's primary job becomes brokering an agent ecosystem—essentially a fleet of AI agents built to serve your business. And because these agents are trained on your organization's own data, they get smarter and more aligned over time, evolving in a way that generic AI tools never can.

The risk of standing still

There's a real risk here, and it's not just about falling a little behind. If GIS programs don't examine their strategy, they risk becoming irrelevant.

When users can generate their own maps, run their own analyses and produce their own spatial insights without our help, the old "service provider" model simply doesn't hold up. If the value you offer is limited to technical outputs, it's only a matter of time before someone asks, "Do we even need a GIS team for this?" Staying focused on technical execution could result in a missed opportunity to lead and shape how AI is used.

Standing still right now is a bigger risk than moving forward. Adaptation isn't optional.

Final thought

The rise of AI marks a fundamental shift for GIS. It automates tasks across the entire geospatial data life cycle. It democratizes access to powerful analysis and visualization tools. It also changes users' expectations about what GIS can and should deliver.

However, it doesn't diminish the importance of GIS. It makes human GIS expertise more important than ever as long as that expertise evolves.

The future belongs to the GIS programs that move beyond technical execution to strategic leadership. It belongs to the teams that invest in human skills, durable assets and deep relationships. It belongs to those who are willing to ask themselves the hard questions now and act boldly on the answers.

Let’s talk

Are you a GIS manager or leading a geospatial or AI function? Are you concerned about the state of your strategy in the face of AI? Do you need help developing a workforce strategy that reflects the changing GIS landscape? Send me an email or connect with me on LinkedIn. I’d like to hear from you!

All the best,

Matt

The Geospatial Edge is a periodic newsletter about geospatial strategy and location intelligence by Esri Canada’s director of management consulting, Matt Lewin. This blog post is a copy of the issue that was sent to subscribers in May 2025. If you want to receive The Geospatial Edge right to your inbox along with related messages from Esri Canada, visit our Communication Preference Centre and select “GIS Strategy” as an area of interest.