In a paper from California Institute of Technology’s Jet Propulsion Laboratory, the author laments the loss of connection between machine learning and solving real-world problems. The main question that seems to come up is: what good can machine learning or artificial intelligence (AI) do? A lot, as it turns out. AI could play a big role in aiding and even solving the issues of world hunger and poverty in ways that are surprisingly simple, but increasingly necessary.
AI’s Natural Abilities Used in New Ways
When you think of AI, it’s possible that fixing world hunger isn’t the first application that comes to mind. However, when broken down to their most basic components, world issues need more of one thing: data. The right kind of data can help identify issues that impoverished regions face, which, in turn, can lead to solutions. Information processing and solution finding are the backbone of AI, and enlisting its help can generate the data required by global initiatives to identify regions in the greatest need of aid.
Global programs like the United Nations and World Bank have launched initiatives to reduce world hunger and poverty that need the massive amounts of data and analytics that AI technology can mine and sift through. The UN is developing strategies at conferences like AI for Good Global Summit to figure out how AI can help them fulfill their Sustainable Development Goals, while the World Bank is taking a more direct approach. They’ve partnered with Orbital Insight, a geospatial analysis company, to determine whether satellite data can improve ways of tracking poverty and socioeconomic trends globally.
Capturing images of distressed areas via satellite, for example, provides markers that measure how a village is doing. A team of scientists from Stanford University’s Department of Earth System Science created a poverty mapping technique that uses satellite images to capture and measure luminosity levels to determine what potentially impoverished areas look like during the day and night. The daytime images highlight the homes, roads, cars and other aspects of the village to provide a baseline of how many people live there and what resources are nearby. If the population is sparse or the area is rural, less light in nighttime satellite images would be expected. However, if a village is well populated but has less electricity than expected, that can be an indicator of poverty.
Gathering this type of data could make it possible to pinpoint the areas most affected by poverty. Providing that information to humanitarian groups, NGOs and even governments could better equip them to deal with a region’s unique issues.
In Times of Crisis …
As smartphones gain popularity in the developing world, the possibilities for AI to contribute to struggling communities grow. Researchers from the Data Science Institute and Department of Computing at Imperial College in London developed an algorithm that can look at mobile data from users in a region to predict the gender of the user. When a disaster hits, those traceable demographics could potentially help identify and track the most vulnerable users (such as women with young children) and get aid to them faster, even in remote or dangerous areas.
Another method of tracking crisis victims combines smartphones with our innate human need to socialize or let our friends and family know when something is going on. Crisis-mapping technology uses social media, email and texts to identify and map areas where a natural disaster or other humanitarian crisis is happening. An open source software program created by researchers from the Qatar Computing Research Institute called Artificial Intelligence for Disaster Response (AIDR) collects and classifies tweets posted during an event. Using a keyword and tag tracker, this software marries machine learning and the natural human impulse to share and connect during times of crisis to create virtual maps that can show events as they are happening.
Despite all of the good that planning, gathering data, and developing programs may do, helping those in crisis does require some on the ground (or in the sky) legwork. Humanitarian aid is an important part of the mission that the Northrop Grumman autonomous aircraft system Global Hawk has taken on. In 2013, it was able to find a safe zone for humanitarian aid during Typhoon Haiyan in the Philippines using data collected through detailed imagery. It was also able to determine safe roads and fields to guide rescue teams to displaced survivors to get them help. In addition to being an “eye in the sky” for rescue missions, Global Hawk also supported requests from international partners to surveil areas of Haiti after the earthquake in 2010 and the Fukushima disaster in 2011.
AI and Humanity Entering the Future, Hand-in-Hand
Next steps for injecting AI and machine learning into humanitarian aid vary from collecting and developing crops that can adapt to a changing climate to drones and smartphone accessories that can deliver medical attention to remote areas. These methods aren’t tactics being saved for a far-off future; they can happen right now. Jack Hidary, a serial tech entrepreneur and Senior Advisor to XLabs, told the Telegraph that “In low-income areas, agriculture and healthcare are two critical ecosystems that we can apply AI to immediately.”
While the debate rages on about artificial intelligence and how it will affect the workforce of those in developed nations, those in developing nations could benefit from the intervention of technology into their lives. Partnerships between world health organizations and AI-driven initiatives like Orbital Insight, AIDR and Global Hawk show that the efforts to combat world hunger are becoming more focused. As AI technology evolves, its role in providing solutions and data to help combat world problems will grow, providing us with an insight into our world and society that we may have never seen without it.
