Existential Questions on the Future of Work (Part 3: Technology & Automation)
As technology becomes more apt in automating tasks performed by people, how might labor abundance impact the social order or how we define identity and purpose in life? Who are really the Makers and Takers and are rewards, incentives and liabilities aligned ethically?
The thoughts that follow in these writings intend to stimulate discussion around questions we already face or will likely confront within our lifetimes:
Part 1: People & Civilization
Part 2: Institutions & Economics
Part 3: Technology & Automation
Part 4: Makers & Takers
Part 5: Values & Questions
Acknowledgements include (links provided throughout):
Andrew McAfee and Erik Brynjolfsson (sequence of books on topic),
Ryan Avent: The Wealth of Humans,
Heather McGowan (starting with Jobs Are Over) and
Kurt Vonnegut Jr., Player Piano.
Created to Serve Us
It is intuitive to assume things exist wholly to serve the betterment of individuals, communities and civilization. Previously we noted how people developed things such as tools, machines, institutions and systems and employed them to transform the world and establish modern civilization.
We live in an age unthinkable even a half-century ago. The sum of human knowledge is now widely available to all — at all times, nearly everywhere. The human population is connected together as never before through global social media. At this moment of unprecedented potential to move civilization forward together, paradoxically some of the very things we created threaten to subordinate people and the planet.
Digital Creations: Automation & Learning Systems
One focus of my job is to transform how complex products are designed and built by harnessing newly abundant and affordable computer technology. From this vantage, a path forward that leaves behind a significant segment of the workforce is not difficult to imagine. With very few exceptions, the content of this (exceptional / must-watch) 15 minute video by CGP Grey is well-within my professional understanding of the potential of existing and emerging technology.
The vast increase in affordability and capability of computers driven by Moore’s Law made possible the creation of sophisticated and adaptable systems (sensors, controls, data storage and logic processing) capable of performing a number of tasks that humans had previously performed to earn wages. The term automation refers to using such machinery for physical (robotics) or non-physical tasks.
“Over the past decades, computers have broadly automated tasks that programmers could describe with clear rules and algorithms.
Modern machine learning techniques now allow us to do the same for tasks where describing the precise rules is much harder.”
Jeff Bezos, Letter to Amazon Shareholders, August 2017
Unlike the physical automation of the Industrial Revolution that began with the power loom in the 1800's, the ability of these systems to perform thinking activities can be more broadly applied. As eloquently noted by Bezos, recent advances in machine learning have introduced the capability for machines to perform a wider array of tasks with far more robustness and flexibility than ever before. Systems can even surpass humans in areas such as consistency, as machines do not get distracted, suffer fatigue, etc. Example: Medical Errors.
The books by Andrew McAfee and Erik Brynjolfsson comprehensively address a multitude of factors along this path, yet remain optimistic on the outcomes. I agree there is strong precedent to safely bet on human ingenuity and entrepreneurship to adapt to the emerging world where technology performs many (most?) of the tasks performed by people today. However, the shape of our economy and society beyond the likely inevitable displacement of today’s workforce continues to elude any credible and intuitive description by the futurists and visionaries.
Opportunity for Symbiotic Technology
Humans retain strong advantages in a number of traits and behaviors valuable across occupations, including: creativity, ingenuity, curiosity, adaptation, abstraction, intuition, reflection and common sense. These can be critical to an institution being opportunistic in growth, resilient during change and addressing unknown unknowns. Institutional systems and technologies persistently evolve to better people in managing variability & inconsistency, defects & waste, slack & surplus, rework & repair and now the comprehension & perception of data and its patterns beyond human ability. Naturally, these have value to institutions in reducing costs, risks and confronting known unknowns — for example, through execution and monitoring of plans and contingencies.
The Future of Workers will be shaped by their ability to collaborate with machines in an ever-changing division of labor between the relative strengths of humans and technology. In Automation should be like Iron Man, not Ultron (Communications of the ACM, Vol. 59 №3), Tom Limoncelli writes:
The compensatory principle says people and machines should each do what they are good at and not attempt what they do not do well. That is, each group should compensate for the other’s deficiencies.
Limoncelli notes simply automating everything possible and leaving the leftover gaps to the humans often produces negative results. For example, employers compelled to short-term financial performance (perhaps by activist investors) may by motivated to fully automate their as-is process. But this comes at the expense of inclusion of observant and innovative workers lending to robustness and long-term strength of the firm. Human-centric automation argues for recognizing the importance of targeting tasks holistically and recognizing the assets humans contribute to processes.
The Race to Stay Ahead of Technology
The unrelenting pace of job transformation through automation technology will change the demand for skills in the labor market. This means workers will need to be continually learning to remain relevant to employers, as systems take on additional tasks within their occupation — workers will be expected to perform new tasks the machines cannot (yet) perform.
Even as workers invest to keep pace with state of the art technology and thereby remain employed among the fewer and more productive, entire classes of jobs may be eliminated. What we consider a “career change” today may become the norm tomorrow — an expected and ongoing shifting between “employment engagements” (as termed by heathermcgowan in Jobs Are Over).
Future socio-economic systems must also consider support needed by workers transitioning from one employment engagement to the next. Many of these workers will not be young, single and mobile — but have planted roots in their communities with families and homes. Additionally, in places where healthcare is coupled with employment, lack of work continuity (as workers cycle between jobs and retraining) could constitute significant loss in productivity to industry and potentially devastating stress and financial risk for families.
Employers of Automation
The Future of Work will be defined by the combination of two trends:
- economic self-interest absolving institutions (corporate firms) of social responsibility (to benefit humanity and the progress of civilization) &
- the emerging potential of technology to broadly automate tasks performed for wage-earning in the occupations of humanity.
Technological innovations are of course as much a human creation (a thing) toward the progression of civilization as the institutions discussed in Part 2. How we elect to use these things we have created now presents a number of questions of relative priority between people and things.
In The Wealth of Humans, Ryan Avent observes systemic forces countering the growth of high-wage jobs when financially-driven employers are able to substitute technology for wages to lower or limit costs. What he terms the “Employment Trilemma” will shape some critical dynamics as continued advancement in the automation of physical and mental tasks disrupt the balance between institutional self-interest (efficiency, economy & quality) and labor market forces (bargaining power impact on wages).
When technology reduces the sum total of needed labor by traditional sources of income, we will face an opportunity to redefine the role of work in human culture.
See also: Existential Questions on the Future of Work (Part 1: People & Civlization) and Existential Questions on the Future of Work (Part-2: Institutions & Economics) as well as Human un-Learning (a result of Machine Learning)?