At Stanford University, students who major in the arts, humanities and social sciences are known as “fuzzies,” a lighthearted moniker contrasted with the “techies,” or those who study computer science or engineering. While over the past decades, the university has become synonymous with Silicon Valley, it remains a liberal arts institution focused on broad ideas, and some of its most successful future graduates might well be the fuzzies, not the techies.

While some in Silicon Valley point to technological advances and herald the irrelevance of studying anything but computer engineering, in fact, the opposite is becoming true. As our technology tools become more and more accessible to fuzzies, the great irony in our technology-led world is that our soft skills become increasingly relevant. They become the very way we differentiate from, and have an advantage over, machines.

Routine tasks can be foreseen, scripted and therefore programmed. It’s only a matter of time before these processes are organized into code so that they can be repeated in endlessly more efficient ways. Look no further than the factory floor to recognize that highly routine environments that can be scripted will, over time, give way to machines that can perform these tasks with little error or downtime. Manual automation has been happening in the form of robotics for decades, but it’s also important to note that many people work with these robots. In Korea, there are 478 robots per 10,000 manufacturing workers; in Japan, 315; in Germany, 292; and in China, 36. There are still many humans in manufacturing.

Today, machine learning and artificial intelligence are often seen as the harbingers of immense job elimination. But more than eliminating jobs, these tools will redefine and augment tasks. Within each job, there are many tasks that are predictable and therefore highly routine. They can be scripted and programmed, and eventually, these tasks will be assigned to those that can most efficient when performing them: machines.

But where the fearful news hours stop is in explaining that many tasks are not predictable, they’re not routine, and they cannot be scripted or programmed. These tasks often contain a highly human component such as critical thinking, communication, creativity and collaboration. These four Cs of soft skills highlight the many tasks that are immune to redundancy in our epoch of increasingly competent machines. They should also be expanded to include compassion and complex problem-solving.

As machines hollow out the routine aspects within our jobs, what remain are the nonroutine tasks that are best performed by humans. In our offices, humans and machines will each operate according to their comparative advantage. For example, we’re not going to ask a human colleague to perform a highly mathematical analysis of data, but we will ask a human to creatively frame a problem, to ask questions of data and to communicate results to a client.

In this world where machines carry out routine tasks and humans manage complexity, compassion, communication and creativity, what becomes most important is not only the ability to interact with technology, but also the refined soft skills to differentiate from it.

We are already seeing this reality borne out. Harvard’s David Deming has shown that the highest-growth jobs are not techie-only in scope. Rather, the highest growth category are what he calls “high math, high social” jobs, which involve basic technical fluency and relevant soft skills to be the human layer atop machine analysis.

As our jobs come to consist of higher proportions of complex problems, we each will have skills that we’re good at. As we come to specialize in types of problem-solving and trade tasks to efficiently get things done, communication is the skill that reduces friction. In other words, we require communication and compassion in business, because they reduce the transaction costs of trading expertise, interacting and therefore solving hard problems.

As we invest in the techie, we ought to also not forget the value of the fuzzy.