When the English inventor Thomas Newcomen created the world’s first practical steam engine in 1712, he didn’t ignite the Industrial Revolution—that would come much later. What Newcomen made was an unwieldy device for pumping water out of mines. It would take another 70 years for James Watt to invent a more efficient design—one that could be mounted onto carriages and boats to create railroads and steamships. Steam made the pistons go, but it wasn’t until Watt and his contemporaries found the proper applications that the world as we knew it changed.
What steam was to the first Industrial Revolution, artificial intelligence will be to the fourth, according to the World Economic Forum. (Electrification and digitization were the second and third, respectively.) In both cases, “it’s not the end, it’s just the power,” argues anthropologist and Intel Fellow Genevieve Bell. “AI is the thing that will make everything possible, but like the steam engine, it only gets interesting when you know what it’s going to do.”
Steam trains not only remade the world by virtue of laying thousands of miles of track, but shrank it, too—and then synchronized it with the invention of time zones. It also remade work-as-we-know-it. The first Industrial Revolution created a class of engineers, says Bell, followed later by accountants and computer scientists. What will a world filled with AI produce, and what are the applications that will change—well, everything? Here are five areas in which AI is poised to make a profound impact.
One way to think about AI is to imagine a plunge in the cost of making predictions. Once predictions are cheap, they will be applied everywhere, says Avi Goldfarb, chair in artificial intelligence and health care at the University of Toronto’s Rotman School of Management. “My favorite thought experiment is the airport lounge,” he says. “It exists because we’re bad at predicting how long we’ll be waiting. But what if we knew better?” Imagine AI wringing the slack out of nearly every industry through continuously updated predictions.
And they won’t be doing it alone. Twenty years ago, chess grandmaster Garry Kasparov became the first human to lose to an AI in an epic match against IBM’s Deep Blue. These days, he advocates for human-AI pairings, with the human handling strategy while the machine tackles tactics—making each stronger. AI appears destined to change how we make decisions.
Wall Street’s high-frequency trading bots were just the beginning. As AI evolves and gains agency, we’ll see the rise of what Rethinkery Labs founder Devin Fidler calls “self-driving organizations.” These software constructs will clamber across the Internet searching for opportunities, hiring human staff when necessary. One might assemble a real estate empire of underpriced assets; another might learn how to spot high achievers on the cusp of a promotion, offering executive coaching to help seal the deal. “There are so many opportunities to find situations in which a little tweak delivers an outsized return,” says Fidler. Instead of hierarchical corporations, think ecosystems, as human-AI hybrid organizations appear to exploit new niches in the food chain.
Few industries have a better safety record than aviation. One reason is the famous flight recorders essential to re-creating the moments before a crash in order to learn exactly what happened—and avoid similar mishaps in the future. In a world suffused with well-designed AI, such instruments will be everywhere. A case in point is autonomous vehicles. Ford, General Motors, and Waymo have published details about their approaches to designing AVs that will fail safely in the event of an error or faulty component. Mobileye, an Intel company, has gone a step further, proposing a clear set of predetermined rules to formalize what it means to be a safe driver, transparently defining what safety means beforehand, and applying them to every vehicle on the road.
Great teachers are invaluable—and scarce. So why not use AI to create more of them? That’s the idea behind Nine Billion Schools, a non-profit promoting lifelong, individualized learning. “Not all children will be educated by AI, but you can add it to the mix to obtain different insights,” says Brian David Johnson, a futurist-in-residence at Arizona State University. AI instructors will track students’ progress, tutoring or calling for human assistance when needed. They’ll also connect adult learners to local employers and institutions willing and able to help retool their skills. “This is where the AI can really understand you as a human being,” Johnson says. “It can understand when you’re ready to make your career leap.”
One of AI’s greatest strengths is its ability to infer patterns and relationships hiding in massive amounts of data. And nowhere is that ability better put to use than in health care, which prizes careful record-taking and structured information. For example, Montefiore Health System in the Bronx has created what it calls a “data lake”—a secure pool of anonymized patient data—which it plumbs using models to identify patients in imminent risk and design optimal treatment plans. “A lot of medicine is based on prediction,” Goldfarb says. “And a lot of the critical jobs are in diagnosis. Pairing an AI tasked with taking in data about new symptoms with human specialists will only lead to better outcomes.”
This story was created for and commissioned by Intel.