Key Moments
Daron Acemoglu on Liberalism, Automation, and the Educated Elite
Conversations with TylerWant to know something specific about what's covered?
We've already dissected every moment. Ask and we will deliver (with timestamps).
Key Moments
Liberal democracy is failing not due to external threats, but because the educated elite has prioritized its own interests over broad-based prosperity, leading to societal divisions.
Key Insights
Social contract theories fail by avoiding the need to build consensus around shared moral values and priorities, instead positing abstract truths.
The current crisis in liberal democracy is partly due to the failure of left-liberalism, which became influential but misused its power and created internal weaknesses.
Automation, when prioritized without sufficient new task creation, significantly reduces the labor share of income and can lead to stagnant or negative wage growth.
High-skilled immigration is beneficial, but problematic when it leads the educated elite to neglect the education of their own country's citizens.
AI's productivity boost in the next 10 years is likely to be moderate, not the radical transformation predicted by some industry figures, with an estimated GDP/TFP increase akin to the 1995-1998 IT boom.
Pro-worker AI is defined not by regulatory approval, but by AI that enables workers to do new or more sophisticated things by expanding their information, problem-solving, and expertise.
Critique of social contract theories and the need for consensus
Daron Acemoglu begins by critiquing social contract theories, arguing they circumvent the crucial need to build consensus around shared moral values and priorities. Instead, theories like Rousseau's 'general will' are seen as placing abstract truths or predetermined outcomes above societal agreement, thereby avoiding difficult questions about how to handle intolerance or dissenting views. Acemoglu emphasizes that liberalism must recognize the necessity of fostering community-level agreements and cannot rely on a static set of 'absolute truths.' This perspective forms the bedrock of his argument for revitalizing liberalism by grounding it in the continuous, albeit uncertain, evolution of shared values rather than immutable philosophical tenets. This approach highlights the inherent tension in liberal thought between individual freedoms and the collective need for consensus-building, suggesting that the former cannot fully thrive without the latter.
The failure of left-liberalism and the rise of the educated elite
Acemoglu posits that the current crisis in liberal democracy stems significantly from the failures of left-liberalism itself. While influential for decades, its establishment power was not always used constructively, leading to internal weaknesses that made it vulnerable to external challenges. He likens this to the internal decay that weakened Rome, making it susceptible to external threats. This internal vulnerability is exacerbated by the rise of a post-industrial economy, which has empowered and increased the number of an 'educated elite.' This elite, Acemoglu argues, has increasingly prioritized its own interests and perspectives, leading to a philosophical shift that grants them undue power and influence. This concentration of power, whether based on education, income, or other hierarchies, is seen as problematic for achieving true freedom and flourishing for the broader population.
Automation's impact on labor share and shared prosperity
Acemoglu expresses concern about the role of automation in severing the link between mass production and shared prosperity. Contrary to some views, he argues that automation, at the firm or sectoral level, demonstrably reduces the labor share of income because tasks previously done by labor are transformed into capital, algorithms, or machinery. While this doesn't necessarily mean mass unemployment, it hinders wage growth sufficiently to restore the labor share. He distinguishes this from other technological changes that create new tasks and opportunities. Acemoglu clarifies that he is not against automation itself, which has historically freed humans from difficult tasks, but against prioritizing it without a concurrent focus on creating new roles and opportunities for labor. He points to historical periods of high wage growth, like the mid-20th century in the US, as being driven by a combination of automation and substantial new task creation, a balance he fears is threatened by the current trajectory of AI and robotics.
The complexities of high-skilled immigration and domestic education
While acknowledging the benefits of high-skilled immigration, such as the H1B visa program, Acemoglu voices a significant concern: the educated elite's apparent abandonment of investing in the education of their own citizens. He argues that when the elite can easily fill skill gaps by importing talent (nurses, doctors, programmers), there's less incentive to address the deficiencies in domestic education systems. This creates a problematic reliance on foreign expertise while neglecting the potential and development of the native workforce. Acemoglu advocates for a renewed 'social compact' where influential figures are committed to improving American education, suggesting that current levels of investment, even if high in per capita spending, are not addressing root causes and may be hampered by issues like ideological dysfunction within educational institutions.
AI's projected productivity impact and the centralizing trend
Acemoglu revisits his earlier assessment of AI's productivity impact, acknowledging that while models advance rapidly, the development of practical applications has been slower than anticipated. He expects a moderate productivity boost over the next decade, comparable to the IT boom of 1995-1998, rather than the radical transformation some predict. He attributes this to the need for new task creation alongside automation, a balance that hasn't yet been fully realized. A significant concern is the centralizing nature of large language models, which aggregate vast amounts of data and tend toward a 'one model rules all' philosophy. This contrasts with his preference for decentralized, domain-specific AI models that could better foster competition and avoid excessive power concentration, a trend he observes more in China than in the US.
Defining 'pro-worker' AI and the role of regulation
Acemoglu clarifies his concept of 'pro-worker AI.' He does not advocate for a regulatory body to approve or deny AI based on whether it is 'pro-worker.' Instead, he defines pro-worker AI as technology that enhances workers' capabilities by providing information, expanding problem-solving skills, and increasing expertise. This vision contrasts with narrow automation, which displaces labor. He believes that fostering pro-worker AI requires a shift in the narrative and aspirations of tech entrepreneurs and companies, supported by public policy rather than dictated by it. Examples include AI tools that enable nurses, journalists, or electricians to perform more sophisticated tasks, thus improving their roles and contributions.
Fertility decline and the long-term economic outlook
Acemoglu expresses less worry about the fertility crisis than some, particularly in light of recent research suggesting that induced innovation can offset potential negative impacts on aggregate demand and supply. His models indicate that declining birth rates, while slow-acting, can be compensated by innovations driven by labor scarcity. He acknowledges the surprising nature of these findings, which have been repeatedly tested. While recognizing the potential for different global equilibrium effects due to widespread fertility decline (except in Sub-Saharan Africa), he also points to increased life expectancy and health advancements, enabled partly by computational biology, as factors that will allow for greater human capital accumulation throughout longer lives. This perspective shifts focus from potential demand shortages to the ongoing capacity for innovation and human development.
The future of academic research and AI's role
Reflecting on his own prolific output, Acemoglu notes that while AI can assist in labor-saving tasks like data cleaning and meta-analysis, it hasn't yet fundamentally changed his process of generating groundbreaking ideas. He uses AI more as a tool for efficiency rather than a source of novel conceptual breakthroughs. While acknowledging the potential for AI to increase the quantity of research, he remains focused on the quality and impact of papers. He also touches on the trend of AI companies hiring economists and philosophers, viewing the latter as essential for ethical considerations, but cautioning economists against using their expertise solely for 'whitewashing' corporate impacts. He emphasizes the need for economists joining these firms to maintain strong ethical standards and a critical perspective.
Mentioned in This Episode
●Companies
●Organizations
●Books
●Drugs & Medications
●Concepts
●People Referenced
Common Questions
Acemoglu believes social contract theories circumvent the need to build consensus around shared moral values and priorities, instead relying on abstract, absolute truths. He argues this approach avoids confronting difficult questions like how to handle intolerance or dissenting views.
Topics
Mentioned in this video
Nobel laureate in economics, author of 'What Happened to Liberal Democracy?', and guest on the podcast.
Host of the podcast 'Conversations with Tyler' and interviewer.
Mentioned for his consensus-based defense of liberal democracy, contrasted with Acemoglu's views.
Mentioned in the context of social contract theory and utilitarianism, specifically regarding his approach to fairness.
Co-author of papers with Acemoglu on automation and robots.
Co-author of a paper with Acemoglu on induced innovation and its impact on per capita income, offering a more optimistic vision.
Quoted at the beginning of Acemoglu's book regarding the connection between free speech and the abolition of slavery.
Mentioned regarding India's changing stance on population and fertility rates.
Co-author of a paper with Acemoglu and Autor on induced innovation and population dynamics.
Mentioned in the context of potentially influencing the education system, though the transcript refers to 'President Trump' in relation to higher education crises.
Mentioned as having created a crisis in higher education, though Acemoglu notes longer-term improvements.
Mentioned in the context of concerns about shrinking market sizes due to fertility decline.
Mentioned as an example of an Armenian individual known for intelligence and expertise.
Mentioned as a company making significant advancements in AI, with its economic impact being debated.
Mentioned as a major AI company with rapid revenue growth, and its potential impact on productivity.
Mentioned as a company that could benefit from AI advancements in the supply chain.
Mentioned as an example of a company using AI in its operations.
Discussed in terms of its geopolitical situation, historical influences, economic performance, and potential in the AI age.
Mentioned as a key geopolitical influencer and neighbor of Armenia.
Mentioned as a neighbor of Armenia with geopolitical significance and influence on Armenian culture and food.
Mentioned as a neighboring country to Armenia with geopolitical implications.
Mentioned as having the second-largest Armenian population globally.
More from Conversations with Tyler
View all 53 summaries
68 minJulia Ioffe on Russian Women, Missing Men, and the Roots of Autocracy
64 minAndrew Graham-Dixon on Vermeer, a Forgotten Sect, and Underrated Artists
63 minChase Koch on Principles, Music, and Overcoming Entropy | Conversations with Tyler
52 minJoel Mokyr on Clans, Corporations, and a Culture of Growth
Ask anything from this episode.
Save it, chat with it, and connect it to Claude or ChatGPT. Get cited answers from the actual content — and build your own knowledge base of every podcast and video you care about.
Get Started Free