Paul Allen's new machine learning center for impact is figuring out what poachers will do next

"They were trying to run their operation from that physical board," says Ted Schmitt, principal business development manager for conservation technology at Vulcan, the Seattle-based philanthropic tech company founded by Microsoft cofounder Paul Allen (who died on October 15), which partnered with the park to help it move to the company's digital system, called EarthRanger, in April. "They all know that poaching goes up during a full moon, for obvious reasons," says Schmitt. "But what they don't know, and they all expect, is that there are patterns like that latent in the data that they just can't pull out. That's the promise of machine learning…it's going to let them be proactive." The machine learning is still in early stages of development, but some analytic tools are already in use. A new heat map feature, for example, first tested at Grumeti Game Reserve in Tanzania and Liwonde National Park in Malawi, showed that most incidents were happening near the borders of each park, so rangers could focus on those areas with the highest risk.