Nobel laureate Daron Acemoglu has become a distinctive voice in the growing conversation about artificial intelligence, not because he dismisses the technology’s risks, but because he finds the current discourse trapped in unproductive extremes. In a recent interview with Fortune, he described his frustration with both the uncritical enthusiasts who proclaim an AI‑driven utopia for everyone and the hardened skeptics who reduce sophisticated models to mere random chatter. According to Acemoglu, yielding to either fear or blind optimism is the worst possible reaction at this juncture. He likens his position to that of Groucho Marx, refusing to join any club that would have him as a member, underscoring his insistence on maintaining intellectual independence. The economist argues that the debate has become a mirror of broader societal tensions, where nuance is sacrificed for attention‑grabbing headlines. By rejecting the binary framing, Acemoglu opens space for a more measured examination of what AI can truly deliver and where it may falter. His call is for stakeholders to step back from the echo chambers, recognize the genuine advances in areas like code generation and mathematical reasoning, while simultaneously acknowledging legitimate concerns about workforce disruption and social inequality. Only through this balanced lens can we begin to craft policies that harness AI’s potential without undermining the foundations of liberal democracy.
He characterizes the two dominant camps in the AI discussion with vivid labels that reveal their ideological rigidity. On one side are what he terms the quasi‑moderate‑friendly believers, individuals so convinced that artificial intelligence will inevitably benefit every segment of society that any challenge to this worldview provokes an almost visceral reaction. On the opposite flank sit the skeptics who treat large language models as stochastic parrots, dismissing their capacity for genuine comprehension and reducing their output to meaningless noise. Acemoglu notes that there are segments of the political left that fall into this latter camp, refusing to entertain evidence that the models can perform complex tasks, a stance he finds profoundly unproductive. He insists that anyone who has interacted with frontier systems will see that the parrot analogy fails to capture the subtle progress in pattern recognition, logical inference, and domain‑specific expertise. At the same time, he warns against embracing an unfounded belief that technological progress will automatically resolve socioeconomic challenges. The economist’s refusal to align with either extreme stems from his conviction that productive dialogue requires acknowledging both the tangible capabilities of AI and the legitimate anxieties it provokes about job quality, wage stagnation, and power concentration. By holding these seemingly contradictory ideas in tension, we can move toward a more realistic assessment of the technology’s trajectory.
Acemoglu points to concrete examples where cutting‑edge AI models are already delivering measurable advances, particularly in fields that demand rigorous logical reasoning. He highlights the way modern systems are assisting researchers in constructing mathematical proofs, suggesting that the technology is beginning to augment human creativity rather than merely replicate it. In software engineering, AI‑driven code completion tools have become commonplace, accelerating development cycles and lowering the barrier to entry for complex projects. These successes, he argues, deserve recognition and should inform public policy that encourages responsible innovation. Simultaneously, he stresses that anxieties about future job losses are not unfounded; historical episodes of automation have shown that while new occupations can emerge, the transition period often leaves workers vulnerable. The economist therefore advocates celebrating genuine technical achievements while maintaining a vigilant eye on labor market signals. He notes that the ability to hold two apparently conflicting ideas—enthusiasm for AI’s potential and caution about its disruptive effects—has become a radical act in today’s polarized environment. Cultivating this mental flexibility is essential for designing interventions that upskill workers, redirect productivity gains toward broad‑based prosperity, and prevent the consolidation of gains in the hands of a few tech giants.
The economist draws a striking parallel between the current AI debate and the crises afflicting liberal democratic systems worldwide. He observes that both arenas have become increasingly susceptible to escalation, where outrage and sensationalism attract attention more readily than measured deliberation. Social media platforms, he argues, have amplified this tendency by rewarding extreme positions and discouraging nuanced exchange. As a result, political discourse mirrors the AI conversation: each side retreats into echo chambers, unwilling to concede any ground to the opposition. Acemoglu contends that this dynamic erodes the foundational capacity of liberal democracies to negotiate trade‑offs, build consensus, and steer economic change toward shared objectives. In healthier periods, democratic institutions facilitated compromise through mechanisms such as referenda, citizen assemblies, and cross‑party negotiations, allowing societies to adapt to technological shifts without fracturing along ideological lines. Today, however, the same institutions appear weakened, unable to mediate between competing narratives about artificial intelligence. The economist warns that unless we restore spaces for genuine dialogue—where listeners can entertain opposing viewpoints without immediate rebuttal—the spiral of polarization will intensify, jeopardizing both the legitimacy of democratic governance and the equitable deployment of emerging technologies.
To understand why liberal democracy is faltering, Acemoglu returns to its historical core, emphasizing that the system’s stability once rested on more than periodic elections and constitutional safeguards. He argues that the true glue holding democratic societies together was the promise of shared prosperity—a belief that economic growth would translate into rising living standards for broad segments of the population. This implicit contract fostered a sense of common purpose and reduced the temptation to view politics as a zero‑sum struggle. However, the transition to what he calls the postindustrial economy disrupted that arrangement. Technological change, globalization, and shifts in labor demand began to concentrate wealth and opportunity among those with advanced education and specialized skills, while large portions of the workforce experienced wage stagnation, precarious employment, and diminished prospects for upward mobility. As the material benefits of growth became unevenly distributed, the social contract frayed, giving rise to resentment, mistrust, and the search for alternative identities rooted in culture or nationalism. Acemoglu suggests that the breakdown of this prosperity‑based consensus is a prerequisite for the current climate of hostility, where debates over issues such as artificial intelligence are fought not on the basis of evidence but as proxies for deeper anxieties about status, dignity, and belonging.
The resulting societal split mirrors the fault lines observable in the contemporary AI discourse. On one side stand the more educated, who have gravitated toward the center‑left and tend to emphasize technological optimism, cultural progressivism, and faith in expert‑driven solutions. On the other side are the less educated, many of whom have drifted toward the center‑right or hard‑right, drawn by promises of economic nationalism, skepticism toward elite institutions, and a desire to reclaim a sense of agency. Acemoglu notes that this divergence extends beyond mere policy preferences; it encompasses a significant gap in values, lived experiences, and mutual empathy. The center‑left, he observes, often remains silent as inequality widens, focusing instead on cultural issues that fail to address the material anxieties of working‑class communities. Simultaneously, cultural politics have become a weapon that deepens the chasm, framing debates around identity, morality, and lifestyle choices in ways that alienate large swaths of the populace. This erosion of shared roots, he argues, has stripped liberal democracy of its capacity to forge compromises that cut across class and cultural divides. Without a basis of mutual understanding and a willingness to listen, the system becomes prone to gridlock, populist backlash, and the kind of zero‑sum thinking that makes constructive dialogue about AI—or any other complex challenge—exceedingly difficult.
Acemoglu’s framework illuminates why the AI debate has become so dysfunctional: the public’s inability to simultaneously recognize the technology’s real capabilities and its potential social harms reflects a broader decay in democratic deliberation. When citizens lose the habit of weighing trade‑offs, they gravitate toward simplistic narratives that either glorify innovation as an unqualified good or condemn it as an existential threat. This binary thinking obstructs the formation of policies that could, for example, incentivize AI applications that augment worker productivity while imposing safeguards against mass displacement. The economist warns that without a renewed capacity for nuanced discussion, societies risk repeating past mistakes where technological breakthroughs exacerbated inequality rather than alleviating it. He points to historical episodes—such as the introduction of electricity or the internal combustion engine—where inclusive institutions managed to diffuse benefits widely, creating new middle‑class jobs and raising overall welfare. By contrast, when political capture and institutional weakness prevailed, gains tended to accrue to a narrow elite, leaving large portions of the population behind. Acemoglu contends that rekindling the democratic habit of deliberative compromise is therefore not merely an abstract ideal; it is a practical prerequisite for ensuring that artificial intelligence serves as a force for broad‑based improvement rather than a catalyst for deeper division.
To make his point tangible, Acemoglu invokes the notion of the ‘jagged frontier’ coined by Wharton professor Ethan Mollick, which captures the uneven nature of current AI performance. According to this view, cutting‑edge models excel at narrowly defined tasks—such as generating functional code, answering factual queries, or assisting in statistical analysis—but falter when confronted with ambiguous, context‑rich situations that require common‑sense reasoning or ethical judgment. This unevenness means that deploying AI in real‑world settings demands meticulous oversight, continuous monitoring, and a willingness to intervene when the system produces erroneous or biased outputs. The economist notes that, beyond the realm of software development, evidence of widespread AI adoption remains thin. Surveys of industries such as customer service and manufacturing show only marginal shifts in employment levels, suggesting that many organizations are still in the experimental phase or are applying the technology in limited, pilot‑scale projects. He cautions against interpreting this nascent adoption as proof of inevitability; instead, he urges decision‑makers to treat AI as a tool whose impact hinges on institutional design, regulatory frameworks, and the extent to which workers are involved in shaping its implementation. By recognizing the jagged frontier, firms can avoid overreliance on untested capabilities while still capturing productivity gains where the technology demonstrably adds value.
Despite his sober assessment, Acemoglu resists labeling himself an optimist. Instead, he describes his stance as a resolute refusal to surrender hope, a sentiment he delivers with a wink to Groucho Marx’s comedic sensibility. This perspective acknowledges the serious risks posed by uncontrolled AI deployment—ranging from labor market dislocation to the erosion of democratic norms—while maintaining confidence in humanity’s capacity to redirect technological trajectories through collective action. He argues that hope, in this context, is not naïve optimism but a disciplined commitment to experimenting with policies that align innovation with inclusive growth. Examples include investing in lifelong learning programs, strengthening social safety nets, and encouraging corporate governance models that give employees a meaningful voice in how AI is integrated into the workplace. By coupling realistic appraisal with proactive stewardship, societies can navigate the turbulence of technological change without sacrificing the democratic ideals that have historically enabled peaceful conflict resolution and broad‑based prosperity. Acemoglu’s refusal to capitulate to despair thus serves as a rallying call for stakeholders across the political spectrum to engage in the hard work of institution‑building, rather than retreating into cynicism or utopianism.
The economist points to his long‑standing professional relationship with fellow scholar Erik Brynjolfsson of Stanford as a living illustration of how disagreeing parties can still find common ground. Although Acemoglu and Brynjolfsson have publicly clashed over the magnitude of AI’s impact on productivity—Brynjolfsson advocating for substantial gains while Acemoglu projects more modest outcomes—they share a deep agreement on numerous ancillary issues, such as the importance of worker‑centric design and the need for policies that mitigate adverse distributional effects. Their capacity to debate sharply yet respectfully, he suggests, offers a template for policymakers, industry leaders, and academics who seek to move beyond partisan shouting matches. Acemoglu expresses particular satisfaction that Brynjolfsson has increasingly advocated for steering AI toward human‑complementary applications, a stance that aligns with his own longstanding concern about preventing technology from undermining labor dignity. He also highlights Brynjolfsson’s role in documenting potential job losses through initiatives like the Canaries dashboard, which aggregates ADP payroll data to reveal early signals of disruption. By modeling this kind of principled disagreement, the two economists demonstrate that progress does not require unanimity; it requires a willingness to listen, to adjust beliefs in light of evidence, and to pursue solutions that balance efficiency with equity.
Turning to policy, Acemoglu’s recent book outlines a concrete agenda for ensuring that artificial intelligence reinforces rather than erodes liberal democratic values. Central to this agenda is the promotion of ‘pro‑worker AI’—tools that enhance employee effectiveness, improve job quality, and create new tasks that raise the value of human labor, rather than simply substituting it. To support this shift, he recommends revisiting tax structures that currently favor capital gains over wages, arguing that a more balanced approach would incentivize firms to invest in their workforce alongside automation. Strengthening the social safety net and expanding redistributive measures are likewise presented as essential buffers against the dislocation that can accompany rapid technological change. Additionally, Acemoglu calls for curbing the disproportionate influence of large technology conglomerates through antitrust enforcement and data governance rules that prevent monopolistic control over essential AI infrastructure. Finally, he stresses the necessity of embedding worker representation in decision‑making bodies that shape AI deployment, ensuring that those most affected by the technology have a seat at the table. Above all, he warns that allowing a scenario where fifty, sixty, or seventy percent of the population becomes jobless, hopeless, and stripped of workplace dignity would render liberal democracy untenable, as mass alienation undermines the very legitimacy of self‑governance.
In closing, Acemoglu reminds us that the project of liberal democracy has been a remarkable achievement in human history—a system that has enabled diverse societies to manage conflict, negotiate compromises, and avoid routine descent into violence despite our innate cantankerous tendencies. He urges citizens, business leaders, and policymakers to draw inspiration from this legacy by actively cultivating the habits of listening, perspective‑taking, and compromise that lie at the heart of democratic resilience. Practical steps include supporting public forums where AI’s benefits and risks can be examined transparently, encouraging corporate adoption of human‑impact assessments before rolling out autonomous systems, and advocating for education reforms that equip workers with the skills needed to thrive alongside intelligent machines. Moreover, reinvigorating independent media and strengthening fact‑checking ecosystems can help counteract the sensationalism that fuels escalation in both political and technological debates. By treating AI not as an inevitable force but as a malleable instrument shaped by collective choices, societies can harness its productive potential while safeguarding the inclusive, dignified future that liberal democracy promises. The path forward demands courage, patience, and a steadfast belief that, even in turbulent times, reasoned dialogue can steer innovation toward the common good.