Shadow comparisons
Continuing my thoughts on deep horizontals, looking at AI and Geospatial economics. Another point of comparison is the use of consumer tools in enterprise workflows.
A very brief continuation of yesterday’s post.
In the MIT study we discussed yesterday, there is a note about an emerging “shadow AI economy.” This is a simple idea that individuals were using AI to accelerate personal workflows. This is a counterpoint to how those individuals might have been using the pilot GenAI workflows designed corporately. A number of reasons were suggested for this. One was the personalization that individuals had put into their own AI experiences. In essence, their prompts were vibing with the individuals more, and as a result, giving better responses. Another might be that the corporate experiences lacked the foundational depth of experience that the consumer tools have. Or for whatever reason, the corporate experiences were a little clunky and forced the user into a new and uncomfortable workflow.
Any IT manager would be horrified by this, knowing they do not control any sensitive corporate IP or private data that might be discussed with the consumer AI in question. They might be so concerned that they ask ChatGPT for advice on handling this situation…
This is another parallel with the geospatial world. While often provided with corporate routing or mapping tools, how often do individuals drop to Google Maps because their corporate toolkit lacks data recency or a decent user interface? Trust me when I say this is a common situation. I’ve seen it in logistics, municipal/civic environments, defence and intelligence applications, and insurance workplaces. Individuals in some of the most sensitive organizations will drop to Google or other mapping providers when necessary because they need to execute.
So, there is a shadow economy in Geospatial, too. However, the transaction here is not overtly financial; instead, it’s one of data access and acquisition. Google knows where individuals are looking, and that feeds its revenue streams. Given the lack of financial barrier, perhaps this becomes even more tempting? There is also a very grey area around licensing and derivative products created from a Google base map. No, I’m not going to cite any sources on that.
This is all to say that the comparisons between the deeply horizontal AI and geospatial economies are startling.