The Future of Work – Rethinking Skills for a New Era of Automation

With the progressive expansion of AI, machine learning and automation, I find myself spending a lot of time contemplating the future of work and how societally we are preparing – or not.

2018 was another year of intensifying debate about the good, bad and ugly aspects of accelerating technological developments. It’s abundantly clear that policy makers are lagging behind the speed of technology development, and often don’t understand the uneven consequences their policies can inflict. Too busy tackling populist fears and concerns related to growing inequality, they’re mostly treating symptoms, not root causes. 

A lot of energy is being spent arguing the overall stock of jobs; will those newly created make up for those lost to automation? While that has happened at every phase of the industrial revolution over the past 250 years, fears abound that this time will be different.

Of course it’s relevant to ask how much and how soon automation will kill (at least some) jobs, but that question isn’t very actionable. Other immediate questions are not being asked enough―such as the paradox of how employers struggle to find employees with the right skills when most employees complain that their employers underutilize their skills. Or how it is that younger generations are more anxious and rigid at a time when most jobs need more adaptable and open-minded talent. Clearly some changes are required.

While we won’t, and shouldn’t, slow down technological advancement, we can prepare better for the future rushing toward us. To that end, I’d like to share some thoughts that may help reframe the debate. To make it more digestible, I am offering these thoughts in three parts, with this first one focusing on the most important skills for the future.

The conundrum of the growing STEM shortage

Teaching our kids science, technology, engineering and mathematics (STEM) is increasingly important. The ones who excel in these subjects will build and program the machines that will automate all other jobs. There is a wide-spread perception that STEM students are almost guaranteed higher pay and more professional choices. Let’s explore some facts.

McKinsey Global Institute estimates that working hours for advanced IT and programming workers will nearly double between 2016 and 2030, and demand for all technological skills will grow about 50 percent. But even in 2030, these skills will only account for one-sixth of working hours in the US and Europe.

I increasingly believe this long-predicted skill shortage isn’t so much the result of an education or parenting failure; the deeper you investigate, the more riddles appear. Curiously, most STEM graduates don’t go on to work in STEM jobs, and most STEM jobs are not held by STEM graduates. Consequently, STEM grads today don’t enjoy better long-term occupational outcomes.

The biggest STEM shortage is in computer sciences and engineering, which also happen to be to the least diverse workforce. Women account for about 50 percent of all STEM jobs, but only for 14 percent of engineering jobs and 25 percent of jobs in computer science.

A Pew Research Center survey in the US shows that the imbalance extends beyond gender. Whites and Asians are overrepresented, holding 85 percent of engineering and computer science jobs, while Blacks and Hispanics are underrepresented, holding only 13 to 14 percent. As issues of racism and sexism intertwine, those at the intersections of both (e.g., black women) are even more under-represented.  

These imbalances have early and deep roots. Social and cultural influences start at a young age and affect long-term motivation. An interesting example is my own eight-year-old son, who joined a coding class in North London and was surprised to find not a single girl in the group. His slightly older sister just isn’t interested in coding. Societally, real and perceived stereotypes interplay, making it very hard to reverse these imbalances. We now face the added risk that AI will immortalize these biases, as happened recently for Amazon’s recruiting algorithm.   

It’s not that educators and employers can’t do a lot better, and I passionately feel that we must work harder to overcome unconscious bias because it’s the right and the smart thing to do. But we also need look at underdeveloped skills beyond STEM that impact the accumulated diversity debt (and frankly the fact that even among White (or Asian) men, STEM careers may be undesirable if not unobtainable.

Skills we stop teaching after kindergarten

Robert Fulghum’s 1986 essay “All I Really Need to Know I Learned in Kindergarten” today rings truer than ever. Before we learned the alphabet, we were encouraged to treat others with fairness and generosity, to use our imagination and act on our curiosity, to take care of ourselves and our surroundings, to express regret and show gratitude. Almost everywhere, children are taught the same lessons. This advice denotes the most inimitable human qualities, and highlights those that will be most valuable when machines learn how to do all the imitable things.

Young children learn by doing multitude micro-experiments with no pressure to meet anyone’s standard of success― until standards are imposed on them. The most creative people in history retained their curiosity and individuality, and continued to experiment all their lives. In the approaching knowledge economy, success will require such ability to connect different ideas and think out of the box, taking advantage of both the current internet and new superpowers that machines will give us, which are exponentially more than any previous generation could enjoy. 

However, we’re very far from the day that machines can genuinely empathize. To borrow a timeless quote from Teddy Roosevelt, “No one cares how much you know until they know how much you care.” Consider the situation where machines may provide a great medical prognosis, but human empathy will be needed to help explain it, support patients and administer many aspects of physical treatment.

In fact, in many areas where machines are expected to share recommendations, a large class of jobs (requiring no STEM background) is being created to train machines so they appear somewhat more sympathetic (or “human”), and to review, translate and explain machine insights and recommendations to a broader human audience.

The bottom line? There are many opportunities outside of STEM, and everyone should still look for areas that fit their interests and strengths. Potential matched with talent and passion can achieve great things. Perhaps there is less necessity of pure STEM education than current thinking suggests; employers will continue investments in boot camps and other learning programs to home-grow talent. Even later in one’s career, there can still be opportunity to get in. What’s most important is each individual’s ongoing investment in learning, and even the very ability to learn and change. 

Fulfilling a growing societal demand for authentic servant leadership may prove to be the most rewarding career path when AI commoditizes the monopoly of expertise in many professions. The first three items in Fulghum’s Credo expose the rarity of true servant leadership: share everything, play fair and clean up your own mess.

*****************

This is the first of a three-part series of posts exploring the future of work. In the next part, I will explore two critical but endangered traits needed for resilience in the face of drastic change in the job market.