Work in a New World

Abstract

Work has always played a central role in human lives, from ancient times to the modern age to contemporary history. How is the way we work changing in the digital economy, and how can we make sure that it is changing for the better and not worse? This chapter provides an overview of what we currently know about the future of work, such as potential job losses through automation, new emerging jobs, new skills, future social security systems, and wage inequality, and of chances and risks more generally of working in the digital economy. The focus is on presenting an overview of currently existing socioeconomic research on the “future of work” and to identify policy areas in which policymakers may have to become active, with either established labor market and social policies or with entirely new policy ideas.

1 Introduction

Artificial intelligence (AI) keeps making the headlines about its progressing capabilities in carrying out activities that have been considered exclusively human behavior (see, e.g., very recently, the BBC reporting on ChatGPT and its successor GPT-4 on March 15, 2023)¹ Yet, the discussion about how digitalization is going to change the world of work dates back at least to the 1960s (see Simon, 1965); this is more than half a century ago. In 1964, the International Institute for Labour Studies (IILS) in Geneva, created by the International Labour Organization (ILO) in 1960, gathered international experts for a 5-day conference to discuss “Employment Problems of Automation and Advanced Technology.” The papers and presentations of this conference have been preserved [see Stieber (1966)], and they offer valuable insights about how international experts at the time evaluated the challenges of new technologies for employment, education and training, management, and industrial relations. There was broad consensus at the conference that all these areas would be strongly affected by the perceived switch away from technological advances of so-called energy converters, i.e., machines that eliminate or reinforce human and animal muscle power, toward so-called “information converters,” defined as machines that eliminate or reinforce human brain power, and which could “facilitate” man’s acquired skills in writing, reading, and reaching conclusions along definite lines of reasoning or computing. It is easy to consider the development of ChatGPT as a current (and temporary) peak in the development of information converters, a journey that already started decades ago. Interestingly, the concerns about the world of work today appear to be very similar or even identical to the ones in the 1960s.

There are several reasons to first look into the past before reflecting about the future. Not everything that appears to be new to the current generation or to contemporary societies is or has been entirely new to mankind. Obviously, the introduction of new technologies into work processes, even of information processors, may be such a challenge for the current work generation, but previous generations have already dealt with this situation. A crucial question therefore is: To what extent are these changes truly revolutionary or disruptive in historical context? Will digitalization of our workplaces and our societies lead us into a very different (digital) society with new types of labor relations and new institutions, or will we merely see gradual adjustments to the existing world of work? The changes of the industrial revolution² were so profound that new labor market institutions, labor laws, and labor rights emerged: for example, trade unions, ministries of labor, regulation of working hours, the statuses of “employees” and “employers,” unemployment insurances, maternity protection, and many other labor-related developments are all outcomes stemming from the industrial revolution. It is impossible to already give a definite answer to the question whether we will transform into a new type of digital society, with new types of labor relations and labor market institutions. Historians will be able to answer this question in the future. This chapter provides a non- exhaustive overview of the ongoing discussion about the future of work: First by looking at some debates from a more “traditional perspective,” about how digital technologies might alter the existing world of work through job losses, new jobs, job transformations, skills training, productivity increases, and other changes. Second, the chapter highlights some considerations about how the world of work might fundamentally change our work culture in a future digital society.

Another interesting observation from the 1964 conference is that the participants attempted to evaluate the value of new technologies. Many concluded that, ultimately, the value of any technology must be determined on the basis of its benefits and harm to humanity, Digital Humanism.

2 What Kind of “Digitalization”?

The terms “digitalization” and “artificial intelligence” (AI) are rarely precisely defined and mean different things in various studies. In this chapter, we understand as “digitalization” the whole process of converting information into a digital, machine-readable format as well as the adoption of digital tools in the world of work. The first phase of “digitalization,” roughly from the 1960s to 2000, is mainly characterized by “computerization,” the adoption of computers and programs. The second phase of “digitalization,” over the last 20 to 25 years, can be mainly described by a large expansion of connectivity (the Internet), diffusion of mobile devices, availability of large datasets, and significant advances in AI. The only type of AI that appears to be available in the foreseeable future is “narrow AI,” i.e., AI that is capable of solving specific problems. Most experts agree that artificial general intelligence (AGI), which would allow for the creation of machines that can basically mimic or supersede human intelligence on a wide range of varying tasks, is currently out of reach and that it may still take hundreds of years or more to develop AGI, if it can ever be developed.³ Therefore, in this chapter, “digitalization” means computerization and adoption of (narrow) artificial intelligence.

3 The Traditional View: Adjustments to the Status Quo

Under the assumption that our economic system and our labor market institutions remain, by and large, the same as they are now and as they have been over the last decades, the economics literature has identified several channels through which digitalization and digital technologies affect labor markets. The two main ones are briefly discussed in this section. The first mechanism is that digital technologies are expected to raise productivity: this is to say output per worker is expected to increase. The second mechanism concerns the substitution or complementation of humans through digital machines (which would most likely also lead to productivity increases). The substitution of workers or human work activities through digital machines entails the risk of job losses and necessary organizational or personal adjustments to new activities, like transitioning of workers to new jobs and/or roles. This may require retraining of workers or the acquisition of new skills with possible changes of a country’s education and training systems. It follows from both channels, productivity increases and labor market transitions, that wages and incomes of workers and firms may change in the process. They may increase or decrease for particular workers or entities and hence lead to rising or declining inequality. Labor market transitions may also touch upon other aspects of work quality like job security, social security, or occupational health and safety.

3.1 Productivity

Productivity is expected to increase on the microlevel (in firms and organizations or at the individual workplace) as well as on the macrolevel (for the economy as a whole). The effects of productivity increases do not only affect labor markets but also other dimensions of the economy. For example, more productive firms tend to be more competitive, and they have growing market shares and are likely to be more successful in exporting their products and services.⁴ This process typically leads to a structural transformation of the economy in which new enterprises and sectors emerge and others shrink or even disappear. In order to compare productivity growth across countries or sectors, economists tend to measure productivity in monetary values, for example, in USD per worker or USD per working hour. For labor markets, one immediate effect is that productivity increases allow for and often translate into wage increases for workers. Indeed, on the macroeconomic level, over long time periods (decades), one can observe that real wages in most countries roughly grow in line with labor productivity. However, it needs to be emphasized that this is not a natural process but depends on institutional settings, for example, the bargaining power of workers and firms. Therefore, we can also find examples and time periods in which the relationship between productivity growth and real wage growth breaks, as, for example, currently in the USA where the two time series have decoupled since the 1980s.⁵ But for previous time periods and for most other countries, this relationship generally holds.

Surprisingly, for many policymakers and academics, labor productivity growth has been declining for decades (see, e.g., OECD (2015) or Dieppe (2021)). Starting in advanced economies in the 1970s and 1980s, the productivity growth slowdown is now a global phenomenon (ILO, 2023). It had and has been widely expected that the many applications of AI in combination with other digital technologies have an enormous potential for labor-saving automation, thereby increasing productivity [see OECD (2020), ILO (2022)]. Therefore, the lack of accelerating productivity growth rates has been dubbed the “modern productivity paradox.” It should be noted that productivity growth is not only determined by the availability of technology but also many other factors, and the factors behind this secular decline in growth rates are still being debated (ILO, 2023).

With respect to digital technologies, the economics literature offers three explanations why the recent technological advances have failed so far to translate into higher productivity growth and hence higher economic well-being for the average person: Firstly, productivity increases can be hard to measure, especially in the services sector or for activities for which there exist no market prices (e.g., private activities or public sector output). We may therefore simply mismeasure and underestimate the real productivity increases. Secondly, the implementation of general-purpose technologies takes time, and recent digital technologies (like AI) are still in their implementation or diffusion phase [e.g., Brynjolfssonet et al. (2019)]. The decline in productivity growth rates is basically explained by a time lag, and future productivity growth rates should accelerate once the technology can exert its full effect. Finally, it has been contemplated that the effect of digital technologies on productivity is simply overestimated. Technologies with large effects on productivity have already been discovered by mankind in the past (steam engine, electricity, and others). The “low-hanging fruits” have already been harvested, and low-productivity growth is the new normal (Gordon, 2013, 2017; Gordon & Sayed, 2019).

3.2 Labor Market Adjustments

On the labor market, the prevailing economics literature sees computers, and more recently AI, as forms of capital, i.e., an economic production factor which is competing with labor, as the other production factor. Usually, a certain degree of substitutability is assumed between both production factors and obviously also exists in practice. Hence, we observe a constant race between machines and humans to be employed (see Acemoglu and Restrepo (2018)), both having comparative advantages, depending on their current costs and their capabilities. With costs for computers, data storage declining, and with the capacity of machines to perform more and more human tasks, capital as a production factor becomes more attractive for firms to employ. Most of the economics literature about the future of work and about job losses and technological unemployment is based on this idea of substitution between capital and labor.

This idea has been formalized in economics with the so-called task approach to labor markets propagated by Acemoglu and Autor (2011), Autor (2013), and others. Economic output at the microlevel (e.g., in a firm) is produced by “tasks” (or “work activities”), and the boundary between what are “labor tasks” and what are “capital tasks” is fluid and is changing dynamically as technological capabilities evolve. Which task is carried out by which production factor depends – in each particular point in time – on the relative economic cost of the two factors. Based on the machine-task substitution in Autor et al. (2003), Autor (2013) suggests that the set of tasks most subject to machine displacement are those that are routine or codifiable.

A task is routine if it can be accomplished by computer-controlled machines following explicit rules that can be programmed. A task is non-routine if people do not sufficiently understand them and struggle to articulate rules that machines can execute—non-routine tasks require tacit knowledge (Autor et al., 2003). As the implementation of computers and other digital devices become relatively cheaper, more routine tasks are being carried out by machines. This has become known as the “routinization” hypothesis, or Routine Biased Technical Change (RBTC). Routine tasks appear to occur mainly in occupations in the middle of the wage distribution, and the RBTC has therefore been used to explain observed employment growth at the top and the bottom of the distribution for several advanced economies [e.g., Autor et al. (2006), Autor and Dorn (2013), Goos and Manning (2007), Goos et al. (2014)]. Biagi and Sebastian (2020) provide an overview of the RBTC literature as well as the definitions and data sources that have been used to calculate empirical measures for routine tasks.

The same idea is typically extended to the analysis of AI, in this new wave of digitalization, whereby AI is seen as a form of capital that can either be a complement or a substitute for (different types of) labor. This is echoed by Frey and Osbourne (2017), Brynjolfsson and McAfee (2014), Brynjolfsson et al. (2018), and others, who claim that the replacement of cognitive and manual routine tasks through capital is evident but that this potential for replacement needs to be extended to non-routine cognitive tasks in the context of AI. Frey and Osbourne (2017) predict that any (also non-routine) task can be carried out by capital as long as it is not subject to any engineering bottlenecks, which they roughly group into the three categories, perception and manipulation tasks (or unstructured problems), creative intelligence tasks, and social intelligence tasks. What clearly emerges from this literature is that routine tasks are most suitable for automation and the replacement of machines including through AI. AI even expands the set of tasks that can be automated; its automation potential is therefore assumed to be even bigger. In other words, AI can be expected to take over at least all routine tasks and probably even many more. Fossen and Sorgner (2019) use two empirical measures for digitalization that have been developed by Frey and Osbourne (2017) (“probability of computerization”) and by Brynjolfsson et al. (2018) (“suitable for machine learning”) to categorize potential future changes of US occupations into “human terrain,” “machine terrain,” “rising stars,” and “collapsing occupations.” The idea is that occupations that require tasks that machines are “good at” are likely to disappear but that many occupations are only being transformed: Humans will use machines in their occupations, but human activities remain essential for the occupation. In most cases, occupations will not disappear in their entirety, but the tasks that humans do within this occupation will change. For example, a radiologist may spend more time with his patients or learn more about how to interpret computer output, while an AI analyzes thousands of images, an activity that the radiologist previously did himself. In summary, digitalization is expected to trigger both replacement and change of human work. In both cases, the consequences are labor market transitions. People will transition to entirely new jobs, to new roles, or to new tasks within existing occupations. Such transitions should, for the benefit of workers and enterprises, be transitions toward jobs that are more productive, are better paid and offer better working conditions. A crucial factor to enable people to make such labor market transitions are the skills that they have or which they need to acquire or retrain.

3.3 Skills

The labor market adjustments discussed in the previous section are only possible if people have the right skills to make such adjustments. Skills can be distinguished from tasks or work activities. Tasks refer to a job or an occupation (in which the tasks have to be carried out—see the “task approach” in the previous section). Skills refer to workers (i.e. to people). A skill can be defined as the capacity of a person to use her abilities, her knowledge, her experience, and her training to carry out particular tasks in a certain context. The distinction between skills (the capabilities of a person) and tasks (generalized or specific work activities) is often not made properly in many discussions led by human resources (HR) specialists and economists about the future of work. The terms are often used as synonyms. Indeed, it may be the case that changing tasks also requires a change in peoples’ skillset, but this is not necessarily the case. We can imagine many situations in which a worker’s existing skills enable him or her to carry out new or different tasks. For example, imagine someone with a high degree of social perceptiveness, like an HR recruiter who is experienced in assessing potential candidates in a physical interview. The recruiter would often be immediately capable of carrying out interviews online in a video call, without or with very little (re-)training. Hence, despite frequent calls for major “re-skilling” our “upskilling” of our workforces in light of digitalization, it is not at all clear which new skills exactly are needed in the future and which types of skills are truly becoming obsolete. Furthermore, there exists also no common definition of “digital skills,” another frequently encountered term in the context of digitalization.

Various interpretations for this lacuna are possible: Digital skills could mean that people have the capability to develop digital tools and AI systems (e.g., IT specialists, programmers, mathematicians, etc. would have “digital skills”). The term could also refer to workers being skilled enough to utilize digital devices, for example, to collect data on the Internet and use certain applications or smartphones or tablets. (Would the majority of workers need massive re-skilling to carry out these tasks?) Digital skills could also refer to the proper use or the interpretation of machine output, e.g., an AI recruiting system rejects an applicant. How should the recruiter integrate this output from the AI in the overall recruiting process? Most likely, digitalization will raise demand for all types of these previously discussed tasks, but at what scale does this shift require re-skilling of our workforces? This is largely unknown. At the moment, for example, current labor shortages in North America and Europe exist in the construction sector, in health and elderly care, and the tourist industry, and none of these shortages are specifically linked to digitalization. McKinsey (2018) and others suggest that those skills become more valuable for which humans have a comparative advantage against machines. This is in line with the economics literature that views labor market effects of digitalization mainly through the lens of substitution between capital and labor. For example, McKinsey (2018) predicts for the United States that demand of physical and manual skills and basic cognitive skills declines by 2030, while higher cognitive skills and social and emotional skills gain in importance. With some common sense, a comparison between what computers are good at and what they are not good at provides us with examples of skills/tasks/situations in which machines and humans have comparative advantages respectively. Machines are usually more efficient than humans in computing, handling large amounts of data, solving same-context problems, non-personal communication, carrying out standardized transactions, categorizing and matching items, and detecting correlations. Humans tend to be better at personal communication, solving problems in changing contexts, detecting causality, tackling problems for which no previous data are available or only very small data points exist, and solving unstructured problems that require frequent switching of tasks.

4 The Historical Perspective: An Entirely New World of Work?

So far, the implicit assumption has been that the overall structure of our labor markets remains, by and large, unaffected by digitalization: Work content changes, and people adjust and transition to better or worse jobs. They may have to retrain or learn new skills, but in the end, our society and the world of work, our social security systems, and education remain as they are. But as indicated in the introduction, does it have to be this way, or could digitalization truly disrupt the world of work and transformation into “digital societies”? This final section is based on Samaan (2022)) and reflects on the potential of the “digital revolution” to transform the world of work on a similar scale as the industrial revolution.

Labor relations changed profoundly around 200 years ago during the IR, driven mainly by societies in Europe and North America. Facilitated by technological advances like the steam and combustion engines as well as expedited by regulatory changes, mass production and standardization of goods became the prevailing modes of production. This newly emerging factory system also entailed changes in the work organization: it has been characterized by a high physical concentration of labor in production facilities and a hitherto unseen division of labor, orchestrated by hierarchical organizations. Both changes, mechanization and standardization of production processes and the corresponding new work organization, have led to unprecedented productivity increases to which we owe much of today’s living standards. In his famous example of the pin factory, Adam Smith has illustrated the magnitude of these productivity increases more than two centuries ago, whereby output per worker could be increased to 4800 pins from less than 20 (Smith, 1776). These changes to the way we work and the role that work plays in our societies are still dominant to this day. Bergmann (2019) calls it the invention of the “job system”: We bundle the vast majority of our work activities (“tasks”) into “jobs.” We call standardized descriptions of such jobs “occupations.” These jobs are then bought and sold on the (labor) market for a supposedly fair share of society’s overall output (wage). Hence, the functioning of the industrial society is centered, not about work that we do for us but about obtaining and performing jobs for others.

The importance of the “job system” for our societies can hardly be underestimated. It is at the center of how we act and how we conduct our lives: We educate ourselves, predominantly, in order to “learn an occupation” and to “get a job.” We want to spend our lives being “employed” and not “unemployed.” Being “unemployed” or without a “real job” is social stigmata and leads to loss of income and social standing. Political competition in every Western democracy is critically concerned about creating new jobs or proposing suitable conditions for companies to crank out more jobs. We are prepared to accept all kinds of unpleasant trade-offs, like destroying our environment or heating up the climate, if only job creation is not harmed, because without jobs, we have no work, no income, no taxes, no public services, no social security systems, no more support for democratic votes, and finally no more society, as we currently know it. This way of thinking has not changed much since the industrial revolution. Today, we do reflect about the future of work, but our imagination of the future is restricted and dominated by the “job system” and by all the institutions and terminology that we created around it: “the labor market,” “re-skilling,” “unemployment,” “part-time work” or “contingent work,” etc. This list could be easily expanded and filled with the respective literature on the future of work. In other words, with some exceptions (e.g., Precht (2018)), most of the discussion on the future of work sees the job system as a given centerpiece of our societies. This was also the assumption in the previous sections about “labor market adjustments”: The job system remains intact.

What is now the role of digitalization in this debate? One can look at the changes during the industrial revolution as a solution to a societal coordination problem: Labor productivity, hence living standards, could only be raised significantly if the division of labor was increased to unprecedented levels. But such a high division of labor creates very complex societies with the necessity to coordinate and administer collective behavior. Who needs what and who should work on which (intermediate) product and which service? At the time, it was impossible to obtain information from each individual and to coordinate on this basis behavior from bottom up. So, the solution to this coordination problem has been a hierarchical society and a hierarchical world of work, physically amassing workers in space and time (in factories or offices, the invention of working time and shift work), the creation of the “job system,” and educating workers such that they can fill “jobs.” The output has been standardized goods from mass production for the “average need” of the population. A society might run out of jobs, but it can never run out of work. The real question that we face today is therefore whether or not digitalization and its powerful offspring, big data and artificial intelligence (AI), are going to eradicate the “job system” and, if so, how we can live without it.

There are three reasons why digitalization, understood as a technology, has the potential to destroy the job system. Firstly, artificial intelligence is a general-purpose technology (Brynjolfsson & McAfee, 2014). It is not an invention, like the radio or many others, which had a confined impact on certain economic sectors and societal domains, like the radio has had on mass media, the printing press, and perhaps the military sector. AI and digitalization are more comparable to electrification. We can find applications and devices in virtually all economic sectors for consumers and producers, workers, management, governments, and many other actors alike. This qualification as a general-purpose technology is a major ingredient for a revolutionary change. Secondly, big data together with the wide availability of smartphones and connectivity provide economic actors with information on the “states of the world” and facilitate decentralized decision-making and decentralized action. Production plants and workers do not need to be concentrated, neither spatially nor in time. Output does not have to be standardized but can be customized for a specific individual. We can think about the industrial economic world as a picture of islands of producers, customers, workers, and managers, whereby the middlemen are connecting the islands. The whole picture (“state of the world”) is not fully visible. Now big data is rapidly filling the empty spaces with many small dots and establishing direct connections among them. Thirdly, digitalization brings about an enormous potential to automate tasks. Such automation will not result in the end of human work but will lead to a big “re-shuffle” of work activities between humans and machines on a scale similar to the industrial revolution. This allows society also to reconsider which kind of work activities humans actually want to and should do and which ones we want to leave for the machines. There is potential for—literally—a “digital revolution” in the world of work rather than a slight and continuous adjustment of the status quo.

5 Conclusions

The world of work is changing through digitalization. The introduction of new technology into our working lives is nothing new but has already taken place for centuries. Even the process of digitalization has already changed the world of work for several decades. Job losses through automation and mechanization are possible but also the transformation of jobs through new technology, with positive effects on productivity and wages. Workers are likely to have to adjust their skills. Yet, as digital tools, especially in the form of AI, are general-purpose technologies and the changes through digitalization are profound and multifold, more disruptive change to the world of work is also a possibility. Whether the changes will be as transformative as the industrial revolution, and if they will lead us into a new “digital society,” is still an open question. Our labor market structures are still very much embossed by the changes imposed by the industrial revolution. We can see digitalization also as a chance to liberate us from the constraints that the industrial revolution has imposed on us, as a path toward more self-organizing, non-hierarchical work organizations that generate individualized work output.

Discussion Questions for Students and Their Teachers

• 1. Which differences do you see between “digitalization” and other technological changes that have been experienced throughout human history?

• 2. Which changes could digitalization trigger in the world of work, and through which channels could these changes be triggered? Reflect on a variety of possible labor market dimensions that could be affected by digitalization, for example, the number of jobs, wages, social security systems, employment security, working conditions, personal development opportunities, equality, and others.

• 3. Which risks and which opportunities do you see in this transformation for people (workers), for enterprises, and for society?

• 4. Where do you see a role for policymakers to implement new or to enforce better existing policies or regulations? Why?

Learning Resources for Students

• 1. Acemoglu, D.; Restrepo, P. 2019. “The wrong kind of AI? Artificial intelligence and the future of labour demand” Cambridge Journal of Regions, Economy and Society, Vol. 13, pp. 25–35.

  This paper discusses how AI is predominantly seen by economists as a tool to automate existing human tasks. This process typically leads to a reduction of labor demand. The authors claim, however, that AI can also be employed to create new and different activities for humans, if the right choices are made. This alternative use of AI would lead to more desirable social and economic outcomes.

• 2. Acemoglu, D.; Restrep, P. 2016. “The race between machines and humans: Implications for growth, factor shares and jobs”, voxEU.

  The article provides an overview of how economists have traditionally seen the relationship between technological progress and developments on the labor market. It relates this historic debate to recent advances in AI and the ongoing considerations about automation in economics.

• 3. Autor, David (2013), “The task approach to labor markets: an overview”, IZA Discussion Paper No. 7178.

  In this overview, the author explains important labor market terminology such as “jobs,” “skills,” and “tasks.” A conceptional, analytical framework is provided in which the two production factors “capital” and “labor” are combined to produce economic output. Through technological progress, the set of tasks that capital and human labor provide changes over time.

• 4. Bessen, J. 2016. “How computer automation affects occupations: Technology, jobs, and skills”, voxEU.

  The author explains how automation does not only lead to job losses through replacement of workers but also influences other economic variables, such as productivity and occupations. If one considers more complex economic interactions over time, it becomes much more difficult to predict whether automation accelerates or decelerates job growth.

• 5. Brynjolfsson, E.; Mitchell, T. 2017. “What Can Machine Learning Do? Workforce Implications.” Science 358 (6370): 1530–34.

  The authors analyze the capabilities of AI (machine learning) relative to the typical capabilities of the US American workforce and discuss possible implications for the latter.

• 6. Brynjolfsson, Erik, Daniel Rock, and Chad Syverson (2019a). “Artificial Intelligence and the Modern Productivity Paradox: A Clash of Expectations and Statistics.” In The Economics of Artificial Intelligence: An Agenda, edited by Ajay Agrawal, Joshua Gans, and Avi Goldfarb. University of Chicago Press.

  The paper reviews recent empirical evidence of the global productivity growth slowdown on the macroeconomic level. Different hypotheses are evaluated that can explain how the alleged technological potential of digitalization and AI can be reconciled with poor productivity growth in the national statistics.

• 7. Fossen, F.; Sorgner, A. 2019. “Mapping the Future of Occupations: Transformative and Destructive Effects of New Digital Technologies on Jobs.” Foresight and STI Governance 13 (2): 10–18.

  The authors distinguish between transformative and destructive effects of digital technologies whereby destructive effects increase the risk of occupations becoming obsolete and transformative effects tend to change occupations and the respective skill requirements. The theory is then applied to empirical data about existing US occupations.

• 8. MIT. 2020. “Study finds stronger links between automation and inequality”, (Cambridge, MA.)

  This short newspaper article discusses the links between automation and income inequality.

• 9. The Digital Humanism Initiative (2021): “Perspectives on Digital Humanism”.

  The book consists of short essays that provide perspectives on digitalization for humanity. It has been written by selected thinkers from a variety of disciplines, including computer science, philosophy, education, law, economics, history, anthropology, political science, and sociology.