CEO of SingularityNET. Also a cross-disciplinary scientist, entrepreneur & author, recently relocated from Hong Kong to rural Seattle area.
The jobs aren’t coming back — well, at least, a lot of them aren’t.
Eventually, Covid-19 will be beaten — vaccines and therapies will be found and widely deployed. However, that doesn’t mean the jobs that the pandemic has taken are coming back.
Of course, some will return. For instance, restaurants will return to in-house dining and hire more waitstaff. But the rethinking and reorganization that Covid-19 has induced will have longer-term impacts. Many businesses have already figured out how to get their requirements filled differently — without the need for as many humans.
Stone X Group director of global macro strategy Vincent Deluard has encapsulated the situation quite crisply: “I would summarize 2020 as the bear market for humans. Like many things, Covid-19 is just accelerating social transformation, concentration of wealth in a few hands, massive inequalities, competition issues and all that.”
Successful companies are relying increasingly on intellectual capital and social capital such as brand value. Deluard created a crude measure of the number of intangible assets a company has per employee — the value of the company’s IP and brand recognition divided by their total amount of staff. Dividing the S&P 500 into deciles based on this measure, he found the top decile has returned 18% this year versus a 19% loss for the bottom decile.
A report from the University of Chicago projects that “Covid-19 shock caused three new hires for every 10 layoffs [and] that 32-42% of Covid-induced layoffs will be permanent.”
Not all this job loss will be due to AI; there are other factors like telecommuting. However, a recent study from Oxford University found that roughly 47% of jobs are at serious risk because of AI specifically. Truck and taxi drivers, retail store staff, physical manufacturing workers, fast-food service workers and miners will likely be among the first jobs to go.
Some jobs should last a bit longer. Delivery and courier jobs should last until suitable drones and automated delivery vehicles emerge from testing to deployment. Logistics management jobs like event planning should last until online task exchange and coordination systems get more sophisticated. Writers should stay employed until NLP systems get a few steps beyond GPT3. Advanced software system engineers should be needed a while, but routine programming tasks will increasingly be automated by machine learning systems. While human lawyers will still be around for a while, the vast bulk of routine business law tasks (e.g. those involving customization of online templates to particular situations) are in the midst of automation by a host of startups.
Many people are already effectively employed by automated digital systems. If you’re a gig economy worker doing tasks for clients on Upwork, the Upwork software framework itself is more like your boss than any particular client. As Upwork and similar software becomes increasingly intelligent and adaptive, this can make your job better, not worse.
The same Oxford study references a concept called the “technological bottleneck” — an attempt to determine how “at risk” a job is of displacement. The bottleneck takes three factors into account:
1. How much creative intelligence the role needs.
2. Whether manual manipulation and perception are required.
3. The role of social intelligence in the role.
The bottom line is, AI is rapidly encroaching on all these areas, and businesses are aggressively refactoring their processes to accommodate AI strengths and weaknesses as they evolve. GPT3 generates prose faster but more erratically than human writers, but this may allow a business to replace 10 writers with one editor and one system administrator — for a few years until these two are also obsolete. An automated fast-food restaurant doesn’t need to have a humanlike burger-flipping robot arm but merely any sort of mechanical burger processor. Humanoid robots like Sophia and similar on-screen avatars can mirror facial expressions and recognize emotions, often providing better empathetic bonding than stressed-out hurried human service staff.
One clear pattern supervening over all the details is that there is a bias toward a more rapid elimination of lower- and middle-class jobs and a somewhat slower reduction of higher-wage technical, knowledge-oriented, managerial and creative positions. This is one factor driving the ongoing increase in wealth inequality; a recent report from Rand Corp. (via Business Insider) found that $2.5 trillion is redistributed from the bottom 90% of Americans to the wealthiest 1% every year. Without this redistribution, the median college-educated American worker’s pay would have doubled.
For an individual needing to generate income to support their lifestyle, the message is clear: Unless you have great traction or expertise in some niche with a special AI-proof characteristic, your focus should be on broad-scope tech-savvy, creativity, social intelligence and, above all, the ability to keep learning and adapting as the unknown unknowns unfold.
For those leading businesses, the clearest path to leveraging the accelerating obsolescence of human labor is to focus on unique technology innovations, elite teams with unique know-how, and building brand assets and social capital. Understanding how to do something few others do and having the ability to rapidly reach a broader community are the core competitive advantages in this next phase.
Geopolitically, we may ask: Once most jobs are gone, then what? Does the West follow the communist pattern and manufacture meaningless jobs just to keep people busy and justify paying them? Or as former presidential candidate Andrew Yang famously advocated, do we introduce a guaranteed universal basic income for everyone?
The potential upside of AI for humanity is tremendous — once benevolent AIs do all the jobs, people can spend their days on social, intellectual, spiritual, athletic and artistic pursuits. However, the odds that the first massively advanced AIs are benevolent to humans will be higher if the transitional impact of AI on the global economy is less inhumane.
One may ask: Who will give UBI to the average citizen of a third-tier Congolese city? Or what will be the orientation of the computer-hacker daughter of an unemployed Ethiopian mom who dies of a curable disease while developed-world youth are living comfortably off UBI and optimizing their psychological well-being?
The decline of human labor, which Covid-19 is accelerating, brings a rich variety of risks and opportunities, which it behooves us to greet with open eyes.