In the first three parts of the AI series we covered the potential of AI-driven job displacement, current societal and economic conditions, and several characteristics distinguishing this fourth industrial revolution from its predecessors. Now let’s talk about what’s next and where do we go from here.
Most AI practitioners, policymakers and industry participants today have a similar vision for the future of jobs. Technology is much better than we are in computation, data analysis, pattern recognition, repetitive tasks and movements. Human beings, on the other hand, are better in creative and strategic tasks, tasks that require emotional intelligence, empathy, and also tasks that require fine motor skills. Kai-Fu Lee, author of “AI Superpowers”, similarly analyzed AI job replacement potential based on how social, creative or strategy based jobs are for cognitive labor and, for physical labor, how structured the environment is and whether high dexterity is required.
If jobs of the future simply require us to be human then most of us should be able to adapt, right?

Unfortunately, the majority of jobs that are not at risk are some of the lowest-paying jobs in the US. For Personal Care and Service Occupations, for example, the median annual wage is just $27,270 as of May 2017 based on data from the Bureau of Labor Statistics [1]. This category includes occupations like barbers and hairdressers, concierges, childcare workers and personal care aides. What about teachers? There are plenty of news headlines suggesting that widespread adoption of AI will lead to teachers focusing more on individual learning outcomes. However, the US is currently undergoing a historic shortage of teachers. The non-profit Learning Policy Institute reported a 35% drop in people studying to be teachers between 2009 and 2014 [2]. Teacher salary data is inconsistent and differs by state but the Bureau of Labor Statistics reports mean annual salary of $58,780 for over 4 million people in the Preschool, Primary, Secondary, and Special Education School Teachers category.

Another problem is the perception of social and personal care work in our society. During the third industrial revolution, the digitization of society, the US continued to lose mostly blue-collar jobs. Historically, these were predominantly male jobs. The occupation predicted by the Bureau of Labor Statistics to decline the most from 2016 to 2026 is locomotive firers, shrinking 79 percent. It is 96 percent male [3]. At the same time, employment in the health care & social assistance sector is projected to grow faster than any other sector. This sector also has 8 of the 20 fastest growing industries across the economy [4]. Cutting to the core of Donald Trump’s core support base, men do not want these jobs, because of the social stigma and the economics of the so-called pink-collar jobs. For example, the median annual wage for a locomotive firer in 2017 was $60,360 compared to $27,270 for Personal Care and Service Occupations.

“It’s not a skill mismatch, but an identity mismatch.” – Lawrence Katz, an economist at Harvard
And fixing this identity mismatch is a significant challenge. One that requires a new narrative of societal and corporate values, coupled with a revised economic structure. This new economic structure needs to empower an individual, at the expense of Corporate America. It can be achieved through strict regulations on lobbying and dark money campaign contributions, as well as new legislation and enforcement of antitrust laws.
In a September 2018 WSJ/NBC poll, 77% of surveyed registered voters said “reducing the influence of special interests and corruption in Washington” is either the most important or a very important issue facing the country. The House of Representatives passed the For The People Act in March of 2019, designed to address federal ethics laws, expand voting rights and regulate dark money campaign financing. The bill has not been put to a vote in the Senate as it has no Republican support.
When it comes to antitrust, the Justice Department is still relying on legislation from the late 1800s and early 1900s, specifically the Sherman Antitrust Act 1890 and the Clayton Antitrust Act of 1914, to police and challenge industry consolidation. Both acts used very broad language leaving a substantial amount to interpretation. The current interpretation of those laws is based on the views of Robert Bork, former Solicitor General, also known for the firing of Watergate Special Prosecutor Archibald Cox. His views are based on the concept of consumer welfare. In other words, unless it can be proven that a certain merger will harm the consumer, it does not violate antitrust laws. In the case of Google, the argument, the company cannot be violating antitrust laws in search because the product is free. This is an antiquated view. There is a strong case that the concentration of economic power in the hands of a few companies leads to negative economic and social outcomes. New antitrust legislation for the 21st century is necessary. The Justice Department needs a new road map and a broader set of tools for antitrust enforcement.

Limiting corporate power in American society is crucial to addressing large scale issues, be it AI-driven changes to the labor market or global warming. Whether it’s the Green New Deal or the Universal Basic Income, or some other proposal, addressing these generational challenges will require substantial funding. Increasing taxes on the wealthy is getting increasingly popular among the 2020 Democratic candidates and the general public. Some polls show that 74 percent of Americans, including 65 percent of Republicans, support a tax on the wealth of the country’s richest individuals. But individual income taxes already account for 48% of the US federal revenue [5]. In 2015, American taxpayers reported $10.14 trillion in adjusted gross income [6]. At the same time, the U.S. Bureau of Economic Analysis shows total seasonally adjusted corporate profits before tax at $8.6 trillion for the last four quarters ending in Q3 2018. While making 83% of the income of the American taxpayers, corporate income tax accounted for just 9% of the US federal revenue. This imbalance is counterintuitive. Corporate America has the most financial capacity to facilitate the necessary changes to our economy and society. Limiting its power will allow the government to develop a new economic model, refocusing it on supporting the individual members of the society.

1 %
In 2017, individual income taxes account for 48% of the US federal revenue
1 %
In the same year, corporate income tax accounted for just 9% of the US federal revenue
While making 83% of the income of the American taxpayers, corporate income tax accounted for just 9% of the US federal revenue.
When it comes to AI-driven labor displacement, Corporate America is, once again, best positioned to drive change. But the government has to be aggressive in setting the policy direction. As we change the perception of what it means to be a shareholder and alter the incentives to encourage socially beneficial corporate behavior, we can put the capital markets to work. This is already happening with the emergence of socially responsible, or ESG, investing but the pace of change is too slow and the breadth too limited. Let’s not forget that AI could potentially replace between 9% and 47% of all jobs in the US, pushing millions into an inadequate social safety net. We need a robust mechanism to ensure that people can maintain a certain standard of living while contributing to the new society. My proposal centers on requiring all publicly listed companies to create a not-for-profit, “social benefit” division and fund it with a percentage of its pre-tax corporate profits. The level of funding can be stepped-up over time as the pace and scale of labor displacement becomes more apparent. A “social benefit” division would be structured just like any other business division, with its CEO and executive team. It would work with local governments, ideally based on guidelines from the federal policymakers, on social impact projects in the area of the parent company’s expertise. When employees of a parent company are displaced due to AI, some can be transitioned to the “social benefit” division. The wages will be lower, but they will be able to maintain employment. Combined with reducing work hours, stretch and re-train provisions by the federal government, this could have a substantial impact on smoothing out the effects of AI-driven labor market displacement.

We benefited tremendously from the first two industrial revolutions. We muddled through the third one, going back just as much as we surged forward. Our economies have grown, elevating global standards of living. We nearly extinguished global wars, reduced famine and improved life expectancy throughout the world. But within countries, societies are fractured and polarized, with clear winners and those who have been left behind. Every year, we leave ever more people behind as we shatter norms and structures we used to rely on. All in the name of technological advancements and the “blitzscaling” mantra of Silicon Valley, built on the premise of moving fast and breaking things. Industry 4.0, this age of “cyber-physical systems”, is only going to accelerate the deterioration of the social fabric. We can no longer keep up with the pace of change, fueled by cheap money and non-existent hardware barriers. While the status quo is cancerous, potentially fatal, we should always have hope. After all, this is a systems design problem we can solve. But we have to move fast. In the next few years, AI will not disrupt our human relationships and the nature of being human itself. But it is coming for our jobs and our livelihoods. We need a new narrative of societal and corporate values, coupled with a revised economic structure. And any economic structure, just as any societal structure, should be built around jobs.

Even Karl Marx, according to Fromm, saw labor as a “genuine activity where man develops himself, becomes himself; work is not only a means to an end — the product — but an end in itself, the meaningful expression of human energy” [7]

[1] https://www.bls.gov/oes/current/oes_nat.htm#39-0000

[2] https://learningpolicyinstitute.org/product/coming-crisis-teaching

[3] https://www.bls.gov/emp/tables/fastest-declining-occupations.htm

[4] https://www.bls.gov/opub/mlr/2017/article/projections-overview-and-highlights-2016-26.htm

[5] https://www.taxpolicycenter.org/briefing-book/what-are-sources-revenue-federal-government

[6] https://taxfoundation.org/summary-federal-income-tax-data-2017/

[7] https://www.marxists.org/archive/fromm/works/1961/man/ch04.htm