AI just recently began making its way into the public consciousness, but over the next several years it will step in to transform entire industries. Key to this transformation is AI’s unlimited capacity to process and analyze mass amounts of data, and then act autonomously on the insights. Compare this to the status quo: We have tons of data but only a limited capacity to put it to work.
Having worked in AI for almost a decade, I was able to ask scientists and researchers I've collaborated with—and who are working on AI solutions that exceed what any of us has seen so far—which industries are likely to be transformed by AI first and what that will look like. Based on their answers, here are five industries I think will lead the charge, and how AI will fuel their disruption.
Medicine
Doctors can’t properly treat patients until they know what's wrong, but determining this quickly and precisely leaves a lot of room for error. Recognizing that there's a wealth of medical knowledge available—but that humans can’t possibly access it all in short timeframes—newer tech solutions are using AI to quickly search large databases, such as Modernizing Medicine (which includes case data from thousands of providers and patient visits), to identify similar cases and their treatments.
These solutions, however, are limited to the data in the network and don't take specific patient information into account. According to Dr. Calin Muntean at the University of Medicine and Pharmacy of Timisoara, "AI will team up with IoT devices to begin closing this gap by performing semi-automatic diagnostics on patients before they see their physicians, identifying their problems and recommending treatments based on a connected network of patient data." Physicians will then act as the final gatekeepers, reviewing and signing off on the recommendations or suggesting alternative treatments.
Emergency Along With Search and Rescue
When major disasters strike (think earthquakes, tsunamis, fires or collapsed buildings), search-and-rescue teams don’t always know where people are or how to get to them. Civilian groups, like those involved in volunteer group SWARM, have taken it upon themselves to address this problem by volunteering their drones during crises. Others, including researchers at Dalle Molle Institute for Artificial Intelligence Research, are working to train drones to locate missing persons by identifying which paths they should follow and how to navigate them.
These efforts focus on groups of drones working independently of one another. The next step is to create AI-driven drone networks that communicate with one another during large-scale disasters. "Imagine swarms of drones being released into disaster sites to detect the presence of human bodies," says Dr. Razvan CIOARGĂ, Politechnica University of Timisoara. "Through communication with one another and the use of intelligent sensors, they’ll map previously unknown environments and generate real-time 3D models that show rescue teams where people are and the safest way to get to them."
Public Transportation
Have you ever sat at a red light on an empty street, waiting for it to turn green? Or been held hostage in a subway car waiting for train traffic ahead to clear? These service disruptions are symptoms of a greater problem: non-centralized public transportation.
Different modes of transportation and travel paths (streets, rails and otherwise) are currently handled independently from one other, but larger cities need a single system that monitors the interaction between them. In the U.S., the Federal Transit Administration has formalized its commitment to "Intelligent Transportation Systems for Transit," a program focused on enhancing existing systems with new technologies and processes.
Rather than fix the current system, researchers such as Dr. Ciprian David, a member of the European Space Agency, are working to introduce a new, interconnected one that relies on AI to process quantitative and qualitative sensor data. "AI will push out information to keep things running smoothly and autonomously control the number of vehicles on the road," says David. "This will improve available transportation routes in response to demand, traffic congestion and external conditions."
Education
Schools have come a long way in recognizing the different learning styles and needs of students, but no matter how progressive a school is, meeting the needs of every single student is virtually impossible without help from technology.
Developments over the last decade — such as AI’s ability to track a learner's problem-solving process and prescribe personalized learning activities — will act as building blocks to individualized AI in education. The next step is to fully realize "the potential for AI to create unique learning pathways for individual learners in blended and online learning," as described by 2013 research by Chaudhry, et al.
Building on existing intelligent tutoring systems, which provide immediate, customized instruction or feedback to learners, AI will act as the aggregator and deliverer of data between system, student and educator.
"As the AI learns more about students, it will interact with them directly," says Dr. Vasile GUI, Ph.D., researcher and member of the European Space Agency. "If it identifies a student as particularly interested or proficient in a certain subject, it may inform them (and/or an educator) of opportunities for diving in deeper, perhaps through virtual reality or even further reading material the student wouldn’t otherwise have access to."
Retail
Shrinkage, or loss in inventory at checkout, is a $45.2B/year problem in the retail industry. Retailers attempt to combat shrinkage in a number of ways, but according to Bogdan Ciubotaru, our company's CTO, “current solutions focus on putting a bandaid on flawed checkout hardware.” The next phase for will be to completely replace physical checkouts with AI-driven systems.
Amazon, for instance, recently announced its checkout-free technology concept Amazon Go. Microsoft has introduced Skip, an app that lets customers scan groceries as they go. And we’re working on retailer-agnostic AI-technology that identifies customers and automatically charges them as they leave the store.
These industries are just the beginning. AI-fueled disruption will ultimately affect all industries that suffer from a lot of information but a limited ability to process and act on it. So, basically all of them.
