Select any filter and click on Apply to see results
Table of Contents
Every student embarks on a higher education journey with a question: “What do I want to be?” The answer, in an economy complicated by machine intelligence and rapid automation, is necessarily complex. Technology is a catalyst that is reconfiguring every profession, from finance to medicine to media. Old verities about “useful” skills are disappearing into the cloud. Many students give the existential question “What do I want to be?” a simple response: “Employed.” But it’s not obvious what skills that will require in a world in which algorithms rule.
The most helpful way forward is to ask, What skills and abilities are machines unlikely to learn in the foreseeable future? These will be the skills and abilities that help us outsmart artificial intelligence (AI) and maintain our place in the professional world, and that students should therefore acquire if they hope to adapt to a technologically complex economy. In other words, are there skills and abilities that are likely to be beyond the scope of machine intelligence—at least for many years to come—and will lie only within the grasp of human minds?
Our foremost technologists seem to think there are. In a 2013 internal study, Google determined that the seven most important traits for successful employees were what are commonly dismissed as “soft” skills—which in fact are just the sorts of skills and abilities that will be difficult to automate. These skills and abilities include: communication, empathy, effective coaching, insightfulness into other people, critical thinking, problem-solving, and making complex connections.1 Far from emphasizing purely technical expertise, the study suggested that the top employees possessed the most “human” skill sets. Of course, these are the textbook qualities of successful managers, regardless of the industry, but it’s telling that the world’s most prominent technology company singled them out.
We would further observe that human minds display two innate characteristics that no machine, regardless of its power, is likely to imitate soon. First, humans are exquisitely good at modulating what we think and do based on specific contexts. Context influences how we analyze situations, how we define and approach problems, how we make decisions, and how we interact with other people. Second, we humans are especially good at doing this when we need to take into account motivation, emotion, and shades of human experience.
One reason that context is so hard for machines is that it is open-ended. It’s the nature of reality that contexts are infinite and sometimes inscrutable. New dimensions of context emerge as the world unfolds (another unexpected policy is announced in Washington, DC, for example), and thus one cannot anticipate every possible context in advance—let alone how it will affect decision-making. Deep learning in machines works by extracting patterns in existing data sets, but it cannot address the full universe of all possible stimuli and hence cannot reduce them to data patterns. Machines cannot be programmed to cope with every possible context.
Moreover, think of what goes into our intuitive understanding of other people; a lifetime of experience allows us to detect shades of expression, nuances of meaning. Emotion plays a central role in such human affairs, sometimes in novel ways. Elements of our current or anticipated situations continually interact with our knowledge of such factors to guide our analyses and decisions.
Interactions among variables (such as those that occur when a change in tax laws affects the economics of higher education), by their very nature, are complicated—in part because there are so many of them. We need to know which interactions are important, and here’s where we humans have a real edge: our brains have evolved mechanisms for quickly sorting among a plethora of constant interactions between context and decision-making—and they do so in part by relying on our emotional reactions. Neuroscientists have documented that emotion plays a crucial role not only in setting our goals but also in guiding our decisions and behavior—in part by helping us prioritize relevant factors.2 And researchers simply don’t understand how emotion affects cognition well enough to emulate such processes in a machine.
To preserve our human place in the professional world, education in the twenty-first century should focus on developing skills and abilities that involve appreciating the effects of context, particularly as it bears on human emotion, motivation, and experience. Jobs that do not require these skills and abilities are most likely to fall prey to machines sooner rather than later.
According to a comprehensive study sponsored by Pearson Education that focused on multiple industries in the United States and United Kingdom, skills that will be increasing in demand include the ability to educate and train people, the ability to coordinate people, and originality.3 These are all distinctly “human” skills and abilities and require context awareness and the ability to read and respond to emotion. The skills and abilities that are predicted to be in decline—control precision, wrist-finger speed, manual dexterity, reaction time, equipment maintenance—are all routine, not context-dependent or connected to distinctly human capacities. The importance of soft skills is not limited to Google.
Moreover, it’s not just a question of which types of jobs are likely to be automated and which types are not. We can parse the responsibilities of many jobs into two types of activities: those that can be easily automated and those that will be difficult to automate. For example, diagnosing illnesses may soon be done by smart machines (and probably better than by most human physicians), but it is unlikely that machines will ever replace humans when the results of such tests need to be discussed with patients and their relatives to consider treatment options. Similarly, machines may soon conduct much of the research in legal proceedings, but they are unlikely to be put in charge of swaying a jury with an emotionally charged closing argument.
The human difference doesn’t end there. Although we can program machines to be motivated to do certain things, our best attempts to program their motivations to take into account context sensitivity and human qualities pale beside what humans can do. Despite the giant leaps made by machine learning, artificial intelligences don’t feel the drive to pioneer new ventures or the curiosity to create new knowledge. Smart machines don’t wake up in the morning wanting to solve social ills. And even if they did, they wouldn’t know how to solve such problems in ways that would work for humans.
Clearly, the most valuable human skills and abilities will be those that combine context sensitivity with an emotional or social element, such as cultural agility, empathy, and leadership, as well as the ability to ignite a generative spark to be creative and entrepreneurial.
We can help people cultivate their human advantages, but this requires a new approach to education. Instead of just educating people in robotics, we should educate people in what we call “humanics,” an integrated curriculum that combines a foundation in technical domains with human literacies.4 As a baseline, a humanics curriculum teaches mathematics, coding, and basic engineering principles. It also grounds students in data literacy, giving them the analytical tools to become fully functional actors in the digital world. Furthermore, to thrive in the human milieu, students study humanities and design, acquiring the power to communicate well, to think ethically, and to engage with a diversity of perspectives. A humanics education will equip students to thrive alongside AI, not compete with it.
Crucially, a humanics education should be cemented through experiential learning—internships, independent research, global experiences, co-ops, and the like. Immersion in the world is the most direct and effective means of gaining and practicing context sensitivity. Experience catalyzes the suprarational aspects of learning, pushing students to respond to life’s infinite variety of contexts by leveraging their invention, imagination, and mental agility.
Of course, advanced machines are going to continue to evolve. But growing AI dictates that human intelligence, too, must continue to grow. To that end, it is vital for human learners to upgrade themselves throughout their working lives, engaging in sustained, lifelong learning as they adapt to the shifting technological landscape. Consequently, higher education will have to reorient a large part of its focus toward serving working adults, developing customized, lifelong learning experiences for people who are engaging firsthand with AI in the workplace. This will necessitate a significant shift in the traditional model of higher education, moving lifelong learning from the sidelines to the center of the university’s mission. By breaking free of disciplinary silos and old-fashioned degree formats, we can offer learners content personalized for their immediate needs, delivered on demand.
Despite the miraculous power of our technologies, we can outsmart artificial minds, but only if we think like human beings. Given an education that develops what are likely to remain uniquely human skills and abilities for many years to come, tomorrow’s students will be able to answer questions about what they want to be with confidence: “A fully contributing human being!”
1. Valerie Strauss, “The Surprising Thing Google Learned about Its Employees—And What It Means for Today’s Students,” Washington Post, December 20, 2017, https://www.washingtonpost.com/news/answer-sheet/wp/2017/12/20/the-surpr....
2. Antonio Damasio, Descartes’ Error: Emotion, Reason and the Human Brain (New York: Penguin, 2005).
3. Hasan Bakhshi, Jonathan M. Downing, Michael A. Osborne, and Philippe Schneider, The Future of Skills: Employment in 2030 (London: Pearson and Nesta, 2017), https://futureskills.pearson.com/research/#/homescreen.
4. Joseph E. Aoun, Robot-Proof: Higher Education in the Age of Artificial Intelligence (Boston: MIT Press, 2017).
To respond to this article, email firstname.lastname@example.org with the authors’ names on the subject line.
JOSEPH E. AOUN is president of Northeastern University. STEPHEN M. KOSSLYN is president and CEO of Foundry College.