Guest post by Dr. Anshu Roy
Today, artificial intelligence (AI) is often misdescribed as a video game, dystopian technology, or relegated to discussions about drones. Recently there’s been a cascade of caricatures drawn depicting AI exclusively as a weapon of repressive regimes bent on weakening open societies. Some of the mischaracterization is a generational anxiety about something new and different. Previous generations didn’t have access to the synergy of 24-7 satellite imagery and AI-enhanced ability to connect thousands of seemingly unconnected anomalies to search for predictive patterns. They couldn’t augment human intelligence and planning with a technological edge to anticipate flashpoints and give policymakers more time to make good decisions before a quickly ticking clock leaves only bad options on the table.
But it’s in our interests—all of our interests—to insist on more accurate portrayals and comprehensive dialogues. AI is an innovative tool borne of knowledge. Knowledge about the power stored inside the atom led to the nuclear bomb. That same knowledge cures cancers. AI is no different. In the right hands, it is a solution for good. Instead of decrying only one dismal potentiality, we should want and should be taking concerted, determined steps to see artificial intelligence deployed not to oppress, but to empower decisions promoting the endurance of the liberal order, climate action, and humanitarian response to mass migration—issues about which everyone in national security should be passionate.
Perfect storms gather an ocean away; it’s in our interests to weaken them before they crash onto our shores. But you can’t weaken what you don’t anticipate. Today, we can anticipate—and act, in time.
How might we use AI to look ahead? Using a combination of thousands of commercial and publicly available data, explainable AI models operational today predict that rising prices will likely drive political instability in Africa with 80% confidence over the next 6-12 months. Why? Global shocks due to concomitant and compounding crises have driven up the price of oil and wheat by as much as 300% in parts of East Africa. AI models show that a decade of worsening drought is weakening fragile public infrastructure and leaving populations vulnerable to a vicious cycle of shortages, ethnic warfare, crime, corruption and predatory influence from malign foreign powers. Because AI models account for variables beyond the capacity of any one expert, they can help us discover novel strategies to overcome these threats.
AI can also help us direct dollars, diplomacy, and development to the right places to plug leaks before they become gushers. It can help us get ahead of “threat multiplier” issues like climate change. AI can be a practical tool in daily analysis and policy planning—whether implemented to anticipate climate refugee flows or make informed investment decisions about energy.
We can use AI to expose disinformation, or to save lives amid humanitarian crises. AI can aid the delivery of refugee relief, all optimized using AI-informed global logistics platforms. Risks to supply chains can be minimized using anticipatory indications and warnings capabilities. Cutting-edge logistics save money and lives; it’s “Moneyball” for diplomacy and development.
Rather than drawing caricatures about technology, the great enterprise should be to champion harnessing it to reinvent alliances for information dominance and writing rules of the road.
AI is neither a panacea nor Pandora’s box. We can make it a tool to thwart the march of authoritarians and help tackle society’s biggest challenges. That’s better than merely lamenting a dystopia that’s far from inevitable.
Dr. Anshu Roy is a Life Member of the Council on Foreign Relations and CEO of Rhombus, an AI company serving the national security community.