FUTURE OF WORK: Disrupting Human Capital with AI. (Part 1/2)
By Antoine Bruel, Head of Growth, Braincities Lab
For better or for worse, we are living a deep watershed in the economy with first-degree social impacts on society as a whole: human disruption by robots and autonomous systems. You’ve heard it by now, AI is stealing our jobs; it’s faster, smarter, cheaper than humans. So what are we to become? What are we to do and what will our purpose be in a society where our jobs, career or professional activities define who we are?
There is no denying it, as PwC’s latest July 2018 UK Economic Outlook (PwC, 2018a) repeated past predictions that substantial numbers of jobs would be lost through AI and related technologies, though the proportion of jobs predicted to be cut by 2037 has fallen to 20 percent, rather than the previous 30 percent. But while we spend time scaring active populations, debating about taxes on robots or AI ethics (which is paramount of course), we are not preparing ourselves to evolve and adapt through this inevitable shift. It is blinding us as we do not put enough effort and emphasis on demonstrating how it will actually bring value to people’s life by augmenting human capabilities and making us thrive through this new realization of Schumpeter’s creative destruction cycle. We only see what is being destroyed, replaced or swept away. We are in doubt because we have yet to understand the creative side of the process, the storm has risen and we won’t see until the dust settles. My goal with this piece is to put causes and consequences in perspective and help rethink how AI transforms our society by showing aspects of it you might not have thought of.
This first part will focus on analyzing why AI (understood as the development and deployment of systems with a minimum level of autonomy or problem-solving abilities) raises so many doubts and fears, and how it is currently disrupting us as humans in the economy. In the second part next week, I will balance this overall gloomy analysis to shed some light on this inexorable future looming large, showing how AI will also entail a wave of invaluable innovation to make us more relevant in the workplace.
Here’s my conjecture.
Let’s start with a simple statement: our society is traumatized. This trauma stems from a vicious cycle we are stuck in, nurturing a deeply entrenched fear among middle and lower classes that helps understand why AI adds to the overall frenzy. Very schematically: companies strive for more profits and consumers fight for more purchasing power. The first ones nurture a cost reduction obsession while the seconds fuel an obsession for lower prices. The paradox is that lower costs should lead to lower prices (and everybody would be happy), but that along the way companies discover lowering costs means higher margins (all price being equal). And down the spiral goes: down pressure on prices + obsession for higher margins entail down pressure on costs. First costs in the line of sight and easiest to get rid of are usually human labor costs. This means cutting human workforce down, contributing to unemployment, putting pressure on wages and leading purchasing power down. What are people replaced with? Well, until not so long ago, they were replaced by other humans, just further away on the planet and cheaper as most manufacturing companies are offshoring. But even that is not efficient or cost-effective enough anymore. So we replace people, especially on lower added value tasks, with automation and AI-powered robots.
AI, therefore, becomes the new enemy as it becomes more potent. Innovations make things better by making them simpler and cheaper, with faster and cost-efficient production techniques. So the cycle goes on as plummeting purchasing power leads to higher demand for cheaper products. And this might all stop before everything reaches 0. Or not. This momentum pushes research for yet deeper innovation because short-term gratification and benefits are obvious: technology adoption entails immediate positive externalities. Among these benefits are increased productivity, which allows for greater output with fewer inputs. This greater output, in turn, can generate higher levels of aggregate economic growth. Nevertheless, there are great emerging challenges for the economy to face in the wake of the adoption of technology.
One major consequence of this vicious cycle I described is structural unemployment. Investopedia defines it as a longer-lasting form of unemployment caused by fundamental shifts in an economy and exacerbated by extraneous factors such as technology, competition, and government policy. Reasons, why structural unemployment occurs, include workers’ lack of requisite job skills or that workers live too far from regions where jobs are available and cannot move closer. Jobs are available, but there is a serious mismatch between what companies need and what workers can offer.
There is no denying government policies play a major role in structural unemployment as France could show. France has historically particularly been prone to a higher structural unemployment than its neighbors. It reached 9% earlier in 2018 when it was 6.8% in 2008 before the last crisis, which could mean a bigger chunk of the active population has become more difficult to employ. But probably the main reason is the job market and labor regulations being more rigid or inflexible than surrounding markets. This theory is out for debate as we fuel animated arguments regarding labor regulation, but actual figures do not really matter and we ought to be careful as these metrics are usually used to serve different political visions and arguments. 9% is the average unemployment rate in France for the past 30 years. But I digress.
Whatever the specific reasons for a high structural unemployment in France, we can foresee a potential acceleration of the phenomenon worldwide as human capital becomes more complex and expensive to leverage and value compared to exponentially effective and cheaper autonomous systems. We can’t compete with robots for very long, and the simpler the task at hand, the better machines are at solving them. What’s more, this situation understandably creates confusion in how people perceive Artificial intelligence as a whole, as it is associated with deeply harmful social consequences. As Well Fargo puts it in its March 2018 report on structural unemployment report technological changes that outpace the ability of the labor market to adapt can create frictions in the labor market, whereas workers displaced by technology are unable to adapt their skills quickly enough to a changing labor market demands. In addition, the adoption of technology often requires a higher level of training or other improvements to skillsets for both new workers and existing workers. Skills mismatch issues related to technology adoption implies a need to both re-train existing workers and also to ensure that new workers are prepared for the jobs of the future”.
2) Human disruption
The innovator’s dilemma (by Clayton Christensen) introduced the concept of technological disruption where an emerging incumbent with rapid and flexible development would bring simpler/cheaper/more efficient offers to a market, therefore shaking whole industries and their active established incumbents by either rendering their offers obsolete or creating better alternatives. This time around, AI-led innovation, robotization, and automation are disrupting beyond production techniques or product & service offerings. This time around, we as human beings are massively being disrupted on deeper levels than before. This Human Disruption will be an exponentially pervasive phenomenon across all industries and services as we see technological improvements facilitating the automation of, at first, low added value, repetitive, little-complex tasks, but rapidly escalating to anything more complex. Artificial intelligence aims far beyond mere automation, it aims at endowing machines and systems with faculties to achieve complex tasks where both analogical and logical processes of reflexion are required in the way we approach them. Problem-solving, decision-making. Currently existing multi-purpose AI barely start showing their efficiency that we already try to conquer the next frontier, Art, creativity, and imagination. Just to give you some insight, we are now selling AI-painted canvases, watching AI-written ads and listening to AI-composed music. Before exploring the barriers that AI is overpassing and diving into the Creative aspect of this Schumpeter cycle (especially explaining how AI will help us with AI itself), I want to emphasize on what Human Disruption looks like in our society today with a few demonstrative examples.
Appalling as it sounds, at the beginning of the year 2016, Apple and Samsung supplier Foxconn reportedly cut down its workforce from 110,000 employees down to 50,000, therefore, replacing 60,000 with robots. The company representative stated, “We are applying robotics engineering and other innovative manufacturing technologies to replace repetitive tasks previously done by employees, and through training, also enable our employees to focus on higher value-added elements in the manufacturing process, such as research and development, process control and quality control.” This article from BBC also reports that between September 2014 and May 2016, 505 factories across Dongguan, in the Guangdong province, have invested 4.2bn yuan (£430m) in robots, aiming to replace thousands of workers. Such radical managerial decisions are earthquakes in the job market, putting intense pressure on the active population, job demand, wages, and purchasing power leaving hundreds of thousands of people with low qualifications out of a job overnight.
In the US, Walmart, world’s largest retailer by revenue, has started resorting to automation and robotization as they just announced they will be deploying 360-floor scrubbing robots equipped with AI capabilities in early 2019. These machines can navigate autonomously, adapting to their environment, obstacles and crowd movements in their surroundings in real time. Walmart is leveraging the power of automation to take care of a range of tasks as it recently doubled its use of self-checkouts in store. Critics therefore understandably arise as the US first private employer (1.5 million paid workers) implements more and more automation schemes entailing fears of lower wages and fewer jobs.
In a completely different field, armed forces and war operations are also great examples of how human are being disrupted. Human beings disappear from war zones or special forces intervention as they are being replaced by drones and robots with more sensors, more flexibility and precision and of course, no issue of disposability and moral engagement. This represents a major evolution in armies structure as expertise and jobs settle remotely. Wars evolve and take new forms with digitalization as well with new battlefields entailed by cyber crimes and attacks.
When men disappear from conflict the solemn threat becomes dehumanization of war and its consequences.
These instances can give you chills; how entire layers of the economy and in some industries more rapidly than ever, can we replace people by autonomous systems and machines, so easily and with immediate results. It raises existential questions about the future of work and legitimate worries about our place in the economy, especially when it comes to low skilled population and what they will become, how they can adapt. We have to keep a relevant point in mind though, as David Autor of MIT and Anna Salomons of Utrecht University (they recently published a study on automation) mention in a Brookings article, “The concern should not be about the number of jobs, but whether those jobs are jobs that can support a reasonable standard of living.”
In the following part of the article next week, my goal will be to look at how society can be reshaped by these structural shifts and give some insight into how AI could actually solve issues it is generating itself. See you next week!