Professor Renee Sieber: Where did all the humans go in GeoAI?
GeoAI is fast becoming the algorithm du jour in geospatial research, promising to automate many tedious activities and extract knowledge from big datasets. How do we ensure humans are not rendered irrelevant (or worse yet, entities to be controlled) as GIScientists begin applying the shiny new toy inside the black box? To contextualize the talk I describe past involvement in public participation in computational technologies. I then briefly describe the shiny new toy—the promise of GeoAI—via several applications. Since AI appears to be all about automation of human intelligence it wouldn’t seem there is a role for humans. However, humans are extensively involved in cleaning and validating the training data and model output, which is surprising for what is supposedly an automated process. We should not be sanguine about explicating a role for humans, since AI could be seen as the epitome of technocratic control, sublimating political power and bias deep within the algorithm. I conclude with thoughts about the implications of GeoAI for state governance, human comprehension, individual privacy, public engagement and public and private sector accountability.