Artificial intelligence is already displacing workers — managing its rise should start now

Secretary of Treasury Steven Mnuchin recently publicly dismissed the potential job loss impacts of emerging technologies associated with artificial intelligence (AI), saying that such possibilities remain 50 to 100 years away and that its “not even on (his) radar”. Mnuchin may have been talking about artificial general intelligence (AGI) — i.e. robots with intelligence that meets or exceeds that of humans. If so, then his statements may bear out over time — much remains disputed the timing and possibility of developing AGI.

Artificial narrow intelligence (ANI), however, which uses AI-related technologies to carry out specific tasks through robots or software, is already here and its potential opportunities as well as its risks are well documented by numerous business reports, evidence in global markets, and even the actions and findings of Secretary Mnuchin’s former firm, Goldman Sachs. If Mnuchin’s opinion indicates a broader lack of U.S. policy understanding and anticipation of these issues, it will create significant risks to U.S. economic competitiveness and potentially social stability. AI and its related technologies present immense opportunities for our society — of which we are already reaping the benefits of on an increasing, everyday basis. But we must be clear-eyed about the overall picture as well, and pro-active about managing the potential downsides.

In January, on the same day as Mnuchin’s confirmation hearing, Marty Chavez, Goldman’s Deputy CFO and former CIO (also Mnuchin’s former position), told a Harvard symposium that in 2000 Goldman Sachs employed 600 equity traders. They now employ 2, supported by 200 computer engineers. Indeed, one third of the firm’s 9,000 employees are now computer engineers. Over the same time frame, the number of floor traders on the New York Stock Exchange went from 5,500 to 400, with the positions taken over by server-run algorithms. Other major firms are also increasingly automating much of their operations, including back office tasks, financial transaction research, compliance auditing, and even AI-based financial advisors.

In manufacturing, a 2015 study by the Ball State University Center for Business and Economic Research found that an estimated 88 percent of jobs lost from 2000–2010 were due to increased automation and information technology, with only a small fraction lost to trade. And while most of these jobs were not lost to ANI technologies, they serve as a preview to an even more accelerated ANI-driven future.

In addition, the Boston Consulting Group (BCG) predicts that, due to falling costs, we are on the cusp of a rapid increase in commercial investment in robots. Using welding as an example, BCG cites that a welder today makes a total of about $25/hour, while the total cost of a robot is about $8/hour, when amortized over 5 years — and this is expected to drop to $2 in the next 15 years. Overall, BCG found the tipping point for when it makes sense for relevant companies to move to investment in robots over people is when the cost of a person becomes over 15% higher than a robot. With technology costs plummeting in many areas, a lot more companies may soon make this choice.

Indeed, on March 28, the New York Times highlighted a recent M.I.T. report that “appears to be the first study to quantify large, direct, negative effects of robots” and that “the paper is all the more significant because the researchers, whose work is highly regarded in their field, had been more sanguine about the effect of technology on jobs” just a year ago in concept, before doing the due diligence.

This is not to say we will definitely have fewer total jobs in the future, or that people will no longer have work to occupy themselves. The U.S. and its leaders, however, should expect major workplace and potentially societal disruptions as a rapidly increasing number of mid-wage jobs become automated and the more advanced jobs created require new skill sets. We are already seeing this among American manufacturers. A Deloitte study found that in 2011, in the midst of an economic downturn, there were 600,000 unfilled manufacturing jobs in the U.S. because employers could not find enough people with the necessary higher-level STEM (science, technology, engineering, math) skills necessary for the modernizing work environment. This could increase to almost 2 million by 2021 if the level of workforce education does not improve.

If companies cannot find the workers they need in the U.S., they are more likely to move operations to countries where they can. One example is Germany, which despite using three times as many robots per hour worked than the U.S., lost only 11% of its manufacturing jobs from 2001 to 2010, compared to 33% in the U.S. What accounts for this? Among other things, after under-performing in the 1990s, Germany implemented a broad set of economic reforms starting in the early 2000s to improve their economic competitiveness and increase the technical capacities of its workforce. This included tax reform, a defined technology strategy, and several programs designed to cushion the impacts of job cutbacks on workers, help them to learn new skills, and be prepared for when the economy rebounded. Many other international competitors — including China, Holland, Japan, South Korea, and Taiwan — have taken comparable measures.

Returning to Goldman Sachs, in a 2016 report on jobs, the firm stated “Today, the pace of this (automation) evolution is accelerating … putting more jobs at risk.” The report went on to say:

“This dynamic has helped create a ‘jobs gap’ — the gap that often exists between the types of jobs that people want and the types of jobs that are available. Closing the jobs gap requires a new approach to risk-sharing, one that spreads the burden of investing in human capital more broadly. This risk-sharing approach should include a greater educational focus on social skills, creativity and judgment, not only STEM subjects; expanded incentives for corporate job training; standardized labor contracts; innovative financing structures to support investments in human capital and career transitions; lower barriers to entry into certain professions; increased support for small-business creation; and regulation that supports the growth of the ‘freelance economy.’”

As the pace of this evolution picks up, so will the need for novel and creative solutions. Our policymakers, academics, and think tanks are predisposed to tools at the ready or that have already exhibited success. Unfortunately, in the markets of the future, successful tools of the past are unlikely to provide solutions that move fast enough to keep up — in short, the jobs that a truck driver may retrain for could very well then also be supplanted as well. How do we as a society, and a species, then catch up?

The first stated part of Mnuchin’s responsibility as Secretary of Treasury is to “maintain a strong economy and create economic and job opportunities by promoting the conditions that enable economic growth and stability at home and abroad.” To do this, it is important that he not reject outright the vast and growing evidence that while the total number of U.S. jobs may not drop precipitously, the world of work is changing quickly and many of today’s mid-skilled jobs will no longer exist in a decade, much less 50–100 years. Artificial intelligence and its associated technologies are inarguably fueling this reality. If the U.S. is to keep up and create opportunities for the working middle class, a Trump administration stronghold, government has a crucial role to play, working together with the private and education sectors, in setting the conditions and policies necessary to support business and workers in a more automated America.

Technology for Global Security is working with partners to help manage the implications of the AI-powered revolution — advocating for innovative policies to enable innovative technologies. Get in touch if you want to learn more and get involved: email us at info at