A few separate conversations with people, who had concerns on our future, especially related with Artificial Intelligence and automation, motivated me to write this article. While some had philosophical questions such as ‘Will AI dominate humanity? Then will we have a dismal future?,’ others had rather practical concerns such as ‘What should we do in the immediate future when AI and automation displace us from labor markets? If we are being educated from outdated Industrial Age frames, then what to do? What do we have to learn and how do we prepare ourselves for our future?’
Since I have been reading and have introduced to you several books related with the Digital Economy, I thought I may be able to find some insights from the books. Here, I will develop my discussion with below subjects and order.
I will refer three books for my discussion.
- The Fourth Industrial Revolution by Klaus Schwab, 2016(hereafter Schwab)
- AI, Automation and Work by Daron Acemoglu and Pascual Restrepo, a paper issued on NBER, Jan 2018 (Acemoglu)
- The Zero Marginal Cost Society by Jeremy Rifkin, 2014 (Rifkin)
- Stagnant Growth, The Productivity Paradox and Employment (Schwab)
- An Alternative View on Low Productivity (Acemoglu)
- Countervailing Forces That Counter Balance The Displacement Effect (Acemoglu)
- Creation of New Tasks, Required Skills and Employment Opportunities in an Age of Automation, AI and Robotics (Acemoglu)
- The Distinction Between Industrial Age System and Digital Age System (Rifkin, Schwab, Acemoglu)
- Implications on Future Talents (Schwab, Acemoglu)
1. Stagnant Growth, The Productivity Paradox and Employment (Schwab)
As I’ve introduced in my previous post on the Fourth Industrial Revolution, Schwab states ‘the productivity paradox’ – declining productivity with the expectations of higher productivity that tend to be associated with the exponential growth of technology – is one of today’s great economic enigmas, and for which there is no satisfactory explanation.
However, Schwab takes an optimistic stance on the future growth despite ‘the productivity paradox’ we’ve been experiencing and the potential ‘deflationary impact’ of technology that involves automation AI and robotics, in terms of capital favored over labor and the squeeze of wages.
To support his practical optimism, he introduces an insightful argument. It suggests that innovative goods and services created in the fourth industrial revolution are of significantly higher functionality and quality, yet are delivered in markets that are fundamentally different from those which we are traditionally used to measuring. Many new goods and services are “non-rival,” have zero marginal costs and/or harness highly competitive markets via digital platforms, all of which result in lower prices.
“Under these conditions, our traditional statistics may well fail to capture real increases in value as consumer surplus is not yet reflected in overall sales or higher profits. Therefore, ‘measuring’ inputs and outputs and hence discerning productivity need to be different and so changed.” I considered this very apt and insightful in understanding the whole economic dynamics in the digital economy.
Regarding employment and labor issues, Schwab clearly sees that new technologies will dramatically change the nature of work across all industries and occupations and create few jobs in new industries than previous revolutions. Also, there will be greater polarization: Employment will grow in high-income cognitive and creative jobs and low-income manual occupations, but it will greatly diminish for middle-income routine and repetitive jobs.
2. An Alternative View on Low Productivity (Acemoglu)
Now, in his latest paper issued on NBER, Acemoglu brings a fresh insight. He argues that the current debate on the implications of automation on the future of work is centered on ‘a false dichotomy,’ in which the alarmist claims that the oncoming advances in AI and robotics will spell the end of work by humans, while many economists on the other side claim that there is no reason to be concerned that this time will be any different because technological breakthroughs in the past have eventually increased the demand for labor and wages.
Instead, Acemoglu argues that the puzzle of ‘the missing productivity’ stems from ‘excessive automation’, meaning faster automation than socially desirable. Excessive automation not only creates direct inefficiencies, but may also hold productivity growth down by wastefully using resources and displacing labor.
He argues two factors contribute to excessive automation. First, US tax code biases encouraging capital over labor result in misallocation of capital and labor. Also, labor market imperfections and frictions play a role. Second, there is a ‘deceleration’ in creation of new tasks and other productivity-enhancing technologies other than automation due to firms’ singular focus on AI and automation.
As already mentioned, he approves neither naive optimists nor alarmists. He states that the implications of automation are much richer than the direct displacement effects might at first suggest. What’s his perspectives then? To conclude first, the answer is toward the positive side but it needs time and collective efforts by our side. Let’s move on.
3. Countervailing Forces That Counter Balance The Displacement Effect (Acemoglu)
To start, his stance is that automation and thus AI and robotics replace workers in tasks that they previously performed and create a powerful ‘displacement effect.’ Automation certainly reduces share of labor in national income and employment.
However, there are ‘countervailing forces’ that counter balance this displacement effect: ‘productivity effect’ and ‘capital accumulation,’ both of which increase the demand of labor and ‘deepening of automation’ which increases productivity itself.
Furthermore, there is a ‘more’ powerful countering force that directly counter balances this displacement effect by increasing both the demand of labor and the share of labor: ‘Creation of new tasks’ where labor has a comparative advantage to machines. Illustrating various cases from history, he affirms periods of intensive automation have often coincided with the emergence of new jobs, activities, industries and tasks. I consider this the most important to take note.
However, the adjustment process from rapid automation is a slow and painful one for mainly two reasons: ‘reallocation’ costs of searching and matching displaced labor and ‘mismatch’ between technology and skills.
Especially so, as Acemoglu asserts, if we don’t know what the required skills are and if educational system doesn’t keep up with the fast development, being unable to nurture talents with certain complementary skills leaving a void and thus lowering productivity from new tasks.
Acemoglu confirms from historical cases that growth came hand in hand only after educational system and investments in human capital expanded the skills of the workforce.
4. Creation of New Tasks, Required Skills and Employment Opportunities in an Age of Automation, AI and Robotics (Acemoglu)
So here comes a critical question at least for us. What are the new tasks and types of jobs required in the age of automation, AI, and robotics?
What are new ‘labor-intensive’ tasks where labor has a comparative advantage relative to capital that can balance the growth process in the face of rapid automation?
Even if he acknowledges the past is also replete with automation technologies displacing workers, Acemoglu argues this need not have disastrous effects for labor. Nor is it technologically likely that AI will replace labor in all or almost all of the tasks it is currently specializing in.
He states that there are still many human skills that we still cannot automate, including complex reasoning, judgment, analogy-based learning, abstract problem-solving, and a mixture of physical activity, empathy and communication.
He asserts that there are good reasons why market incentives will endogenously lead to the creation of new tasks that gain strength when automation itself becomes more intensive. Also, some of the most defining automation technologies of our age, such as AI, ‘may create a platform’ for the creation of new sets of tasks and jobs.
He quotes a report by Accenture that identified entirely new categories of jobs that are emerging in firms using AI as part of their production. These jobs include “trainers” (to train the AI systems), “explainers” (to communicate and explain the output of AI systems to customers), and “sustainers”(to monitor the performance of AI systems, including their adherence to prevailing ethical standards). The applications of AI to ‘education, health care, and design’ may also result in employment opportunities for new workers.
Take education for example. According to him, existing evidence suggests that many students, not least those with certain learning disabilities, will benefit from individualized education programs and personalized instruction. With current technology, it is prohibitively costly to provide such services to more than a small fraction of students. Applications of AI may enable the educational system to become more ‘customized,’ and in the process create more jobs for education professionals to monitor, design and implement individualized education programs. Similar prospects exist in health care and elderly care services.
5. The Distinction Between Industrial Age System and Digital Age System (Rifkin, Schwab, Acemoglu)
So they say the way we deal with many issues in our time should be different from the past. As Schwab observed, leaders from various boundaries not only don’t clearly understand the current development in our digital age, but also are confused about rapidly evolving environments in all aspects, making it difficult to cope with them.
To overcome this difficulty, I believe we need to understand the difference between the industrial system and the current digital system. In my opinion, Rifkin clearly defines it.
He compares the three Industrial Revolutions using a coherent framework of Energy/Communication/Transportation matrices with certain business models organized around them.
“The coal-powered steam infrastructure enabled the convergence of coal-powered steam rail transport and coal-powered steam printing for the First Industrial Revolution. Likewise, the discovery of oil, the invention of the internal combustion engine and the introduction of telephone gave rise to a new and more powerful energy/transportation/communication complex for the Second Industrial Revolution.”
“The important aspect of the First and Second Industrial Revolution is that the technology platforms were designed to be ‘centralized with top-down command and control,’ because fossil fuels are only found in certain places and require centralized management to move them from underground to the final end users. The centralized energies, in turn, require ‘centralized, vertically integrated’ forms of communication in order to manage the momentous speed-up in commercial transactions made possible by the new sources of power. The enormous capital cost in establishing centralized communication/energy/transportation matrices meant that the new industrial and commercial enterprises embedded in and dependent on these technology platforms had to create their own ‘giant, vertically integrated operations across the value chain.”
“On the other hand, the coming together of the Communications Internet with a digitalized renewable Energy Internet and automated Transportation and Logistics Internet in a seamless intelligent infrastructure-the IoT technology platform-in the Third Industrial Revolution, due to its ‘open architecture and distributed features,’ allows social enterprises on the Collaborative Commons to break the monopoly hold of giant, vertically integrated companies operating in capitalist markets by enabling ‘peer’ production in laterally scaled continental and global networks at near zero marginal cost.”
The current educational systems were initially set up and developed during the first and second industrial revolution to nurture talents to manage the ‘centralized and hierarchical’ industrial giants. As Acemoglu states, without the demand for workers from new factory jobs, engineering, supervisory tasks, accounting and managerial occupations in the second half of the 19th and much of the 20th centuries, it would have been impossible to employ millions of workers exiting the agricultural sector and traditional labor-intensive tasks. (p9)
These traditional functions in factory, engineering, supervisory tasks, accounting and management roles are still continued as of today, and I assume many of the roles are or may become obsolete in current digital economy. For example, massive production of identical products is giving way to more customized production of sophisticated products enabled by lower cost digital technology. As such, we are seeing more digital-based businesses striding into the top 10 lists of Fortune 500 than decades ago.
Different business models require different thought process and leadership. Firstly, the digital based business models imply imminent changes in the existing supply chain as well as value chain that were designed and more fit for those of industrial system. Secondly, the organizational and management models that were developed around centrally managed, bureaucratic and hierarchical firms may need to evolve into flatter and more empowering ones.
6. Implications on Future Talents (Schwab, Acemoglu)
So I believe understanding of all of these differences can become a starting point for future talents to prepare and equip themselves. That may also explain why, as Schwab mentioned, they claim our traditional approach to the measurement of growth, profits, output and input may fail to catch clearly the value created through different value chains in the digital economy.
What qualities are required in an age of AI, automation and robotics? Acemoglu states that AI and other new automation technologies necessitate a mix of numeracy, communication, and problem-solving skills different than those emphasized in current curricula, which means curricula apt for industrial system. (p33)
Also, in the same context, Schwab states in the foreseeable future, low-risk jobs in terms of automation will be those that require social and creative skills; in particular, decision making under uncertainty and the development of novel ideas. (p40)
Schwab also states we have to reconsider what we mean by “high skill”. Traditional definitions of skilled labor rely on the presence of ‘advanced or specialized education and a set of defined capabilities within a profession or domain of expertise.’ Given the increasing rate of change of technologies, the fourth industrial revolution will demand and place more emphasis on’ the ability of workers to adapt continuously and learn new skills and approaches within a variety of contexts. (p45)
In a recent speech at the World Economic Forum at Davos, Jack Ma, founder of Alibaba in China said that ‘Everything we teach should be different from machines.‘ (You can find the video clip either at @leadingbyreadin or at @wef on Twitter)
Also, the latest article by Bill McDermott, SAP CEO, on World Economic Forum, left a positive note for humans, saying “Machines can’t dream.” And I believe this explains everything! Imagination, creativity, strategy, intuition, inspiration or more soulful aspect of humanity are something only humans can add value to the new tasks and that cannot be replaced by AI, robotics or automation. As Schwab stated as well, humans will perform tasks where we need to work ‘with’ machines in the future, not something else.
Lastly, what’s Jeff Bezos’ view on AI? According to his interview with Fortune in 2012, instead of looking at AI as the harbinger of humanity’s downfall, he sees it as an “enabler” that “will empower and improve every business, every government organization, every philanthropy.”
I have tried to find some common insights from prominent authors and scholars through their books. I hope I added some value to readers who are keen on this subject. One thing I want to ask you all is, especially if you are in a leadership role in an organization or you yourself is leading your own giant, that ‘we need to be centered.’ I believe it’s a leader’s role to stay centered and pragmatically optimistic.
Fear and worry always stem from ‘uncertainty’ which our time of transition represents. And I believe this uncertainty has been faced by our predecessors as well throughout our long human history. So let’s not buy pessimists’ or alarmists’ call. Or at least be discerning or build a balanced view.
Thank you for taking your time for reading my article! I hope this article helped you gain insight albeit small!
- Machines Can’t Dream (https://www.weforum.org/agenda/2018/01/machines-can%27t-dream/)
- Amazon’s Jeff Bezos: The Ultimate Disrupter (http://fortune.com/2012/11/16/amazons-jeff-bezos-the-ultimate-disrupter/)