Last week, our publisher ServiceNow debuted at #1 on the Forbes Most Innovative Companies list for 2018. This ranking measures what Forbes calls the “innovation premium,” essentially the difference between a public company’s market capitalization and the net present value of cash flows from its existing businesses. Companies with larger innovation premiums rank higher on the list because investors expect them to generate outsized cash flows from new products in the future.
The innovation premium seems like an effective way to gauge financial market expectations. It’s certainly a more rigorous measure of corporate innovation than the typical magazine ranking, where editors stick their collective fingers in the wind to decide which company is coolest.
However, it doesn’t capture the ambivalence with which people often view technological innovation. From mechanical looms to factory robots and chatbots, many of the most innovative technologies of the past two centuries have been tools that increased labor productivity by performing tasks that humans used to do. The flip side of productivity growth is that you need fewer people to achieve the same output. Naturally, this tends to spark fears that the new tech will eliminate jobs.
Human anxiety about being displaced by our own creations goes back at least as far as Mary Shelley’s 1818 novel Frankenstein, written at a time when the Industrial Revolution was provoking widespread concern that mechanized production would put artisans out of work. Carriage makers and stable hands felt similar unease when they encountered horseless carriages in the late 19th century. So did industrial workers when robots entered manufacturing during the 1960s and 1970s.
This anxiety has persisted in the digital age. In 1996 and 1997, Russian grandmaster Garry Kasparov played a series of chess games against a custom‑built IBM supercomputer named Deep Blue. Newsweek described the matchup as “The brain’s last stand,” while Britain’s Guardian newspaper said Kasparov’s job was to “defend humankind from the inexorable advance of artificial intelligence.” Kasparov won the first best‑of‑six contest, but lost to Deep Blue in their rematch.
In his 2017 book, Deep Thinking, Kasparov writes that he found it unsettling to compete with a tireless, infinitely resourceful machine. But he goes on to argue that it’s foolish to fear the rise of AI, which he sees as the latest in a long line of innovations that have eliminated jobs in the short term, but created vast opportunities for humankind in the longer term.
“Romanticizing the loss of jobs to technology is little better than complaining that antibiotics put too many grave diggers out of work,” he writes. “The transfer of labor from humans to our inventions is nothing less than the history of civilization.”
All major advances in automation have disrupted labor markets and caused measurable pain to workers. Real wages stagnated for nearly 50 years during the first Industrial Revolution of the 19th century, according to a recent McKinsey Global Institute report. Another recent study found that each new industrial robot eliminates six jobs in the surrounding metropolitan area.
In the long run, however, automation tends to increase employment. New industries and occupations absorb the labor lost to automation. Productivity increases spur demand for goods and services by reducing their cost. As a result, companies create new jobs to meet this demand. In 1850, for example, 58% of the U.S. labor force worked on farms. Thanks to the tractor, the combine harvester and chemical fertilizers, just 2.5% of Americans work in agriculture today. Yet our farms are vastly more productive than they were in 1850, and our economy is at or near full employment.
Over time, automation also drives sustained rises in living standards. By many measures, working class citizens in advanced economies live better than millionaires did in the 19th century.
Given the ever‑increasing sophistication of AI and the exponential growth in computer processing power, it’s natural to worry about the impact of machine intelligence on work and employment. In fact, 72% of Americans are worried about a future in which intelligent machines do much of the work that’s currently done by humans, according to a recent Pew Research Center survey.
The McKinsey study estimates that worldwide, about half the tasks that people are paid to do can technically be automated. That’s true for nearly every role in the modern economy, from warehouse pickers to doctors, lawyers and even CEOs.
The authors of the study predict that advanced economies will create enough new jobs to offset those lost to automation by 2030. Globally, however, as many as 375 million workers will need to change their occupation because of automation, and will need to learn new skills or increase their level of education in order to find work.
In future, the job skills in greatest demand will be those that augment machine intelligence rather than compete with it. That’s already true in chess, where top‑level players prepare for matches using sophisticated software that can analyze billions of possible moves to help develop a winning strategy. It’s also true in medicine, where AI tools help doctors mine vast troves of medical literature to help them diagnose and cure ailments.
Recently, Pittsburgh’s child protective services department deployed a machine‑learning algorithm that helps human case workers decide which “child in danger” calls to prioritize for action. The New York Times reported that the algorithm did a better job than unassisted humans of making these potentially life‑or‑death decisions.
The startup ROSS Intelligence has pioneered a legal AI that combs through millions of pages of case law to find relevant precedents that it then synthesizes into coherent legal memos, complete with citations, analysis and conclusion. That AI won’t be joining the Supreme Court bar anytime soon, however. According to Wired magazine, it still needs human lawyers to hone its arguments and enliven its prose.
ServiceNow sells cloud‑based software that automates routine processes across organizations, freeing up employees to do more meaningful and strategic work. Like the examples above, this technology speaks to a centaur‑like paradigm for the relationship between human and machine intelligence, one where people and machines are more effective as partners than competitors. Regardless of how you define or measure innovation, that seems like a sound way to think about the fruits of human creativity.