A World Without Work: Technology, Automation, and How We Should Respond
Authors: Daniel Susskind, Daniel Susskind
 (Z-Library)_thumbnail.jpg)
Overview
A World Without Work examines the profound societal transformations arising from technological unemployment, arguing that a future with less work is not just a possibility, but a probability we must prepare for. It argues that the traditional response of “more education” will not suffice, and proposes new solutions, including a “conditional basic income” (CBI) and a significant expansion of the state’s role. Unlike past automation anxieties, the threat of technological unemployment is now real due to the nature of artificial intelligence (AI). The book explores the “pragmatist revolution” in AI, where machines no longer need to mimic human intelligence to outperform humans at various tasks. This shift, combined with task encroachment–machines taking on more and more human tasks–will lead to a future with insufficient work to sustain traditional employment structures. The book goes beyond narrow economic analyses, exploring the social and political implications of technological unemployment. It examines how technology is exacerbating inequality by increasing returns on unequally distributed capital, both human and traditional. As work becomes less central to how we distribute prosperity, the Big State will need to play a larger role in taxation and redistribution, including taxing workers whose skills remain valuable, taxing traditional capital, and taxing large technology companies (Big Tech). Furthermore, Big Tech’s growing political power requires new forms of oversight, advocating for a “Political Power Oversight Authority” to regulate their influence on our social and political lives. Finally, recognizing that work provides not only income but meaning and purpose, the book explores the challenge of finding meaning in a world without traditional work. It suggests a shift towards a ‘meaning-creating state’ that promotes leisure policies, provides opportunities for unpaid work, and encourages contributions to society outside of traditional employment. This book is intended for a broad audience, including policymakers, business leaders, and anyone interested in the future of work and society. Its arguments are especially relevant to those working in AI and technology, highlighting the societal responsibilities that come with technological progress.
Book Outline
1. Introduction
Technological progress has repeatedly led to anxieties about job displacement, but these anxieties have historically been misplaced. However, this time may be different due to the nature of AI.
Key concept: The ‘Great Manure Crisis’ of the 1890s, where cities struggled with horse manure, illustrates how seemingly insurmountable problems can be unexpectedly solved by new technologies (the car).
2. A History of Misplaced Anxiety
Economic growth has historically been driven by technological progress. While this has led to automation anxiety, this anxiety has so far been misplaced due to technology’s complementary role.
Key concept: The Age of Labor is a period where technological advancements generally benefited workers. This is in contrast to the potential future of widespread technological unemployment.
3. The Age of Labour
Economists traditionally believed technology was skill-biased. However, the recent phenomenon of “job polarization” indicates technology is also impacting medium-skilled jobs. This shift necessitates a new understanding of how technology affects the labor market.
Key concept: The ‘skill-biased’ view of technology suggests it primarily benefits high-skilled workers, but history shows that technology can benefit different skill levels at different times.
4. The Pragmatist Revolution
The development of AI has gone through two waves. The first, focused on replicating human intelligence, ultimately failed. The second wave, driven by a more pragmatic approach of designing machines to perform specific tasks effectively, has yielded significant success, ushering in a new era of capability.
Key concept: The pragmatist revolution in AI focuses on building machines to perform tasks effectively, regardless of whether they mimic human intelligence. This has unlocked capabilities unforeseen by previous approaches that prioritized replicating human thought processes.
5. Task Encroachment
Task encroachment, the gradual expansion of machine capabilities into tasks previously done by humans, is occurring across manual, cognitive, and affective domains. While the pace of adoption varies globally due to differing tasks, costs, and regulations, the general trend is consistent and likely to continue.
Key concept: Task encroachment describes the general trend of machines taking on more and more tasks previously performed by humans. It is a continuous process driven by advances in AI and automation, affecting manual, cognitive, and affective capabilities.
6. Frictional Technological Unemployment
Frictional technological unemployment occurs when available jobs and workers’ skills or locations don’t align, leading to displacement even when sufficient overall work exists. This issue will likely worsen with increasing technological advancement.
Key concept: Frictional technological unemployment arises from mismatches between workers and available jobs due to skills gaps, identity clashes (e.g., resistance to lower-status work), or geographical location.
7. Structural Technological Unemployment
Structural technological unemployment occurs when the complementing force of technology, which creates new work, fails to keep pace with the substituting force that displaces workers. This occurs due to declining productivity effects, bigger-pie effects, and changing-pie effects.
Key concept: Structural technological unemployment occurs when there simply isn’t enough work for everyone who wants it, in contrast to frictional unemployment where work exists but workers can’t access it.
8. Technology and Inequality
While not the sole cause, technology exacerbates existing economic inequalities by increasing returns on unequally distributed capital. The rise in income inequality, particularly at the top, alongside a falling labor share of income, highlights this growing disparity.
Key concept: Technological progress is not the root of all inequality, but often exacerbates existing disparities by rewarding holders of increasingly valuable and unequally distributed capital (both human and traditional).
9. Education and Its Limits
Education, while valuable, has limits as a solution to technological unemployment. Adapting education to focus on tasks machines struggle with, employing personalized learning, and emphasizing lifelong learning are crucial, but ultimately may not suffice if the overall demand for human work significantly declines.
Key concept: While education is currently our best response to technological unemployment, its limitations include unattainable skills for some, the time and effort required to retrain, and crucially, its inability to address insufficient demand in a structurally unemployed future.
10. The Big State
The Big State must play a more active role in redistribution. This includes not only taxing income and capital where it remains, but also addressing challenges like tax avoidance by large companies and rethinking our approach to existing social welfare systems.
Key concept: The ‘Big State’ will need to take on two key roles in a world with less work: taxing those with valuable capital and income, and redistributing it effectively to ensure everyone has enough.
11. Big Tech
The rise of Big Tech presents a unique challenge. While concerns about monopolies are valid, the growing political power of these companies to shape our lives and societies is a more profound concern, requiring new forms of oversight and regulation.
Key concept: The concern with Big Tech is not just their economic power, which is regulated by competition authorities, but their increasing political power to shape how we live together, requiring new oversight mechanisms.
12. Meaning and Purpose
The connection between work and meaning is complex, varying across history and cultures. As technology encroaches on work, a key challenge will be finding new sources of meaning and purpose, requiring societal adaptation and a potential role for the ‘meaning-creating state’ to help citizens navigate this transition.
Key concept: Work provides not only income but often meaning and purpose, presenting a significant challenge for individual and societal well-being in a future with less work. The “meaning-creating state” will play a vital role in helping people adapt to this shift.
13. Epilogue
Just as the ‘Golden Age of Security’ depicted in Stefan Zweig’s memoirs came to an end, so too is the era of work as the central pillar of society. The future brings both opportunities and challenges surrounding inequality, political power, and purpose, requiring proactive engagement and adaptation.
Key concept: The ‘Golden Age of Security,’ where work was a central organizing principle of life, is ending, and we must adapt to a future with new challenges around inequality, political power, and meaning.
Essential Questions
1. Why is the threat of ‘technological unemployment’ different this time?
Technological unemployment, unlike previous automation anxieties, is a real threat due to the nature of AI and its capacity to encroach on a widening range of tasks formerly done by humans. The “pragmatist revolution” in AI focuses on building machines to perform tasks effectively, irrespective of whether they mimic human thought processes. This has led to unforeseen advancements in AI’s ability to perform tasks across various domains, exceeding human capability in many cases. This progress, combined with decreasing costs and increasingly sophisticated algorithms, suggests that task encroachment will continue, likely leading to insufficient demand for human labor in the long run.
2. What are the limitations of ‘more education’ as a solution to technological unemployment?
Education, while important, will not solve technological unemployment entirely. While reforming education to emphasize non-routine skills, personalize learning, and promote lifelong learning can help, these measures are limited if there’s insufficient demand for human work. Furthermore, not all individuals possess the aptitude or capacity to acquire the skills needed in a rapidly evolving technological landscape.
3. What role should the ‘Big State’ play in a future with less work?
The “Big State” must adapt to a world with potentially less work by taking on a more active role in redistribution. This involves not only taxing income and capital where it remains, but also addressing tax avoidance by large companies and reforming existing welfare systems designed for a world where work is the norm. Furthermore, Big Tech’s growing political power requires new oversight mechanisms to ensure their economic dominance doesn’t translate into unchecked social and political influence.
4. How can we find meaning and purpose in a world with less work?
As work becomes less central to life, individuals and societies will need to find new sources of meaning and purpose. Historically, work has not always been viewed as meaningful, and even today, many jobs are far from fulfilling. The potential for widespread technological unemployment will require us to decouple meaning and identity from work, potentially through policies promoting leisure, opportunities for unpaid meaningful activities, and a focus on contributing to society outside of traditional employment.
1. Why is the threat of ‘technological unemployment’ different this time?
Technological unemployment, unlike previous automation anxieties, is a real threat due to the nature of AI and its capacity to encroach on a widening range of tasks formerly done by humans. The “pragmatist revolution” in AI focuses on building machines to perform tasks effectively, irrespective of whether they mimic human thought processes. This has led to unforeseen advancements in AI’s ability to perform tasks across various domains, exceeding human capability in many cases. This progress, combined with decreasing costs and increasingly sophisticated algorithms, suggests that task encroachment will continue, likely leading to insufficient demand for human labor in the long run.
2. What are the limitations of ‘more education’ as a solution to technological unemployment?
Education, while important, will not solve technological unemployment entirely. While reforming education to emphasize non-routine skills, personalize learning, and promote lifelong learning can help, these measures are limited if there’s insufficient demand for human work. Furthermore, not all individuals possess the aptitude or capacity to acquire the skills needed in a rapidly evolving technological landscape.
3. What role should the ‘Big State’ play in a future with less work?
The “Big State” must adapt to a world with potentially less work by taking on a more active role in redistribution. This involves not only taxing income and capital where it remains, but also addressing tax avoidance by large companies and reforming existing welfare systems designed for a world where work is the norm. Furthermore, Big Tech’s growing political power requires new oversight mechanisms to ensure their economic dominance doesn’t translate into unchecked social and political influence.
4. How can we find meaning and purpose in a world with less work?
As work becomes less central to life, individuals and societies will need to find new sources of meaning and purpose. Historically, work has not always been viewed as meaningful, and even today, many jobs are far from fulfilling. The potential for widespread technological unemployment will require us to decouple meaning and identity from work, potentially through policies promoting leisure, opportunities for unpaid meaningful activities, and a focus on contributing to society outside of traditional employment.
Key Takeaways
1. Task encroachment by AI is a continuous process, not a one-time event.
AI is rapidly changing the nature of work. “Task encroachment” is the trend of machines taking on ever more tasks previously performed by humans. This has implications for both the quantity and quality of work available to people. AI is no longer limited by needing to mimic human intelligence. The “pragmatist revolution” in AI allows systems to solve problems in entirely new ways, opening possibilities for machines to encroach on even “non-routine” tasks previously deemed safe from automation.
Practical Application:
AI product engineers can focus development on non-routine tasks, like creating tools that assist with complex problem-solving or creative endeavors. By identifying tasks where humans uniquely excel, AI can augment human capabilities instead of replacing them entirely.
2. The impact and adoption of AI will vary across countries and industries.
The adoption of AI will not be uniform across the globe. Factors like differences in wages, regulations, and cultural attitudes towards technology will influence how quickly and in what ways AI is adopted in different places. Understanding these differences is crucial for predicting the impact of AI on different labor markets and for designing effective strategies to manage that impact.
Practical Application:
AI product engineers should consider the varying costs of technology and labor when designing systems. A system that might be cost-effective in a high-wage country may not be so in a low-wage country where human labor remains relatively inexpensive.
3. A conditional basic income (CBI) may be a necessary response to widespread technological unemployment.
The traditional welfare state is ill-equipped to handle a world with significantly less work. A conditional basic income (CBI) could provide a solution by offering a regular payment conditioned on recipients contributing to society in some way, whether through work, education, training, or community service. This would not only help distribute wealth, but also address potential social and political unrest arising from mass unemployment.
Practical Application:
Policymakers can use a CBI to incentivize participation in activities deemed socially valuable, like education, community service, or care work. This could help address potential social unrest stemming from unemployment and inequality.
4. The growing political power of Big Tech needs careful oversight.
The economic and political power of Big Tech will likely increase in a world with less work. Traditional competition policy, focused primarily on consumer welfare, will not be enough to regulate these powerful companies. New oversight mechanisms, such as a Political Power Oversight Authority, are needed to consider the broad political implications of these companies’ power to shape our lives and societies, going beyond narrow economic concerns.
Practical Application:
AI product engineers need to consider the ethical implications of their work. For example, systems that collect and analyze user data should be designed with transparency and user control in mind, allowing users to understand how their data is being used and to make informed decisions about sharing it.
1. Task encroachment by AI is a continuous process, not a one-time event.
AI is rapidly changing the nature of work. “Task encroachment” is the trend of machines taking on ever more tasks previously performed by humans. This has implications for both the quantity and quality of work available to people. AI is no longer limited by needing to mimic human intelligence. The “pragmatist revolution” in AI allows systems to solve problems in entirely new ways, opening possibilities for machines to encroach on even “non-routine” tasks previously deemed safe from automation.
Practical Application:
AI product engineers can focus development on non-routine tasks, like creating tools that assist with complex problem-solving or creative endeavors. By identifying tasks where humans uniquely excel, AI can augment human capabilities instead of replacing them entirely.
2. The impact and adoption of AI will vary across countries and industries.
The adoption of AI will not be uniform across the globe. Factors like differences in wages, regulations, and cultural attitudes towards technology will influence how quickly and in what ways AI is adopted in different places. Understanding these differences is crucial for predicting the impact of AI on different labor markets and for designing effective strategies to manage that impact.
Practical Application:
AI product engineers should consider the varying costs of technology and labor when designing systems. A system that might be cost-effective in a high-wage country may not be so in a low-wage country where human labor remains relatively inexpensive.
3. A conditional basic income (CBI) may be a necessary response to widespread technological unemployment.
The traditional welfare state is ill-equipped to handle a world with significantly less work. A conditional basic income (CBI) could provide a solution by offering a regular payment conditioned on recipients contributing to society in some way, whether through work, education, training, or community service. This would not only help distribute wealth, but also address potential social and political unrest arising from mass unemployment.
Practical Application:
Policymakers can use a CBI to incentivize participation in activities deemed socially valuable, like education, community service, or care work. This could help address potential social unrest stemming from unemployment and inequality.
4. The growing political power of Big Tech needs careful oversight.
The economic and political power of Big Tech will likely increase in a world with less work. Traditional competition policy, focused primarily on consumer welfare, will not be enough to regulate these powerful companies. New oversight mechanisms, such as a Political Power Oversight Authority, are needed to consider the broad political implications of these companies’ power to shape our lives and societies, going beyond narrow economic concerns.
Practical Application:
AI product engineers need to consider the ethical implications of their work. For example, systems that collect and analyze user data should be designed with transparency and user control in mind, allowing users to understand how their data is being used and to make informed decisions about sharing it.
Suggested Deep Dive
Chapter: Part 3: The Response
This section focuses on solutions, particularly the concept of conditional basic income and its potential societal impact, offering valuable insights for crafting a positive future amidst technological disruption.
Memorable Quotes
Introduction. 9
What technological progress had done to horses, he said, it would eventually do to human beings as well: drive us out of work.
The Pragmatist Revolution. 43
Real life consists of bluffing, of little tactics of deception, of asking yourself what is the other man going to think I mean to do.
Underestimating Machines. 77
“AI is whatever hasn’t been done yet.”
Technology and Inequality. 139
“The problem with capitalism is that not everyone has capital.”
Big Tech. 211
“The digital is political.”
Introduction. 9
What technological progress had done to horses, he said, it would eventually do to human beings as well: drive us out of work.
The Pragmatist Revolution. 43
Real life consists of bluffing, of little tactics of deception, of asking yourself what is the other man going to think I mean to do.
Underestimating Machines. 77
“AI is whatever hasn’t been done yet.”
Technology and Inequality. 139
“The problem with capitalism is that not everyone has capital.”
Big Tech. 211
“The digital is political.”
Comparative Analysis
A World Without Work stands out for its comprehensive approach, linking technological advancements to economic theory, societal impact, and political philosophy. Unlike books focusing solely on the economic aspects of automation, such as Acemoglu and Restrepo’s “The Race Between Machine and Man”, Susskind delves into the multifaceted implications for meaning and purpose in life. In contrast to Frey and Osborne’s more statistically driven “The Future of Employment”, Susskind provides a more nuanced perspective on how technology transforms tasks rather than eliminating entire jobs. While agreeing with Brynjolfsson and McAfee’s “The Second Machine Age” on the transformative potential of AI, Susskind’s focus on the political implications and the rise of “Big Tech” sets it apart. Furthermore, Susskind offers more concrete policy proposals, such as the conditional basic income, compared to broader societal analyses like Harari’s “Homo Deus”.
Reflection
A World Without Work offers a compelling and thought-provoking vision of the future, but its arguments should be treated with a degree of caution. While the core premise of increasing technological unemployment is plausible, the exact pace and extent of job displacement remain uncertain. Susskind’s focus on the need for a “Big State” and large-scale redistribution may be seen as overly pessimistic by some and excessively optimistic by others, particularly given the political challenges associated with such policies. The book’s strength lies in its ability to connect the seemingly narrow field of AI with broader questions about inequality, political power, and the meaning of life, prompting valuable discussion about the future of work and society. However, its weakness lies in the potential to stoke anxieties about the future, possibly underestimating human adaptability and the potential emergence of new forms of work and meaning. Overall, A World Without Work is a significant contribution to the growing discourse on the future of work, raising critical questions and offering a compelling framework for understanding and responding to the transformative power of AI and automation.
Flashcards
What is task encroachment?
The trend of machines increasingly taking on tasks formerly performed by humans.
What is frictional technological unemployment?
Unemployment caused by mismatches between available jobs and workers’ skills, identities, or locations.
What is structural technological unemployment?
Unemployment caused by insufficient overall demand for human labor.
What is a Universal Basic Income (UBI)?
A regular, unconditional payment made by the government to all citizens.
What is a Conditional Basic Income (CBI)?
A regular payment made by the government with conditions attached, such as participation in education, training, or community service.
What is ‘Big Tech’?
The economic and political dominance of a few large technology companies.
What is the ‘AI fallacy’?
The mistaken belief that machines must replicate human intelligence to perform tasks effectively.
What is the “pragmatist” approach to AI?
Focusing on building systems that perform tasks effectively, regardless of how closely they mimic human thought processes.
What is task encroachment?
The trend of machines increasingly taking on tasks formerly performed by humans.
What is frictional technological unemployment?
Unemployment caused by mismatches between available jobs and workers’ skills, identities, or locations.
What is structural technological unemployment?
Unemployment caused by insufficient overall demand for human labor.
What is a Universal Basic Income (UBI)?
A regular, unconditional payment made by the government to all citizens.
What is a Conditional Basic Income (CBI)?
A regular payment made by the government with conditions attached, such as participation in education, training, or community service.
What is ‘Big Tech’?
The economic and political dominance of a few large technology companies.
What is the ‘AI fallacy’?
The mistaken belief that machines must replicate human intelligence to perform tasks effectively.
What is the “pragmatist” approach to AI?
Focusing on building systems that perform tasks effectively, regardless of how closely they mimic human thought processes.