Navigating the Post-AGI Economic Abyss

As we stand here on September 18, 2025, with Artificial General Intelligence (AGI) no longer a distant dream but a rapidly approaching reality, I find myself compelled to expand further on my recent white paper. What began as a structured analysis has evolved into this narrative-driven blog post, now enriched with additional charts drawing from the latest data and more references to deepen our understanding. I’ve incorporated fresh insights from recent studies, including projections on labor shares, AI energy demands, and future scenarios, to provoke even more thought. Imagine a world where work isn’t about survival but about meaning; where abundance flows not from sweat but from silicon. But what if that abundance slips through our fingers, concentrated in the hands of a few? Join me as we explore the twilight of capitalism and the dawn of something profoundly uncertain.

Historical Context: The Long Shadow of Automation and the Terminal Shock of AGI

Let me take you back to the smoky factories of the 19th century, where the Industrial Revolution birthed modern capitalism. Wage labor became the bedrock: you sold your time and effort, and in return, society churned out progress. But fast-forward to the 1970s, and cracks began to appear. In advanced economies, the labor share of GDP—essentially, the slice of the economic pie going to workers—has been in steady decline. Data from the National Bureau of Economic Research shows this trend accelerating since the early 1990s, with the IMF noting a downward trajectory in most developed countries. A Princeton study echoes this, attributing much of the drop to technological shifts that favor capital over labor. By 2014, the global labor share had fallen from 54% in 1980 to just 50.5%, according to McKinsey Global Institute. The IMF’s World Economic Outlook further details that labor income shares began trending down in the 1980s, reaching their lowest in half a century before the global financial crisis. In the US specifically, the labor share has declined over the past quarter century, as analyzed by Brookings. Our World in Data provides comprehensive charts showing this metric across countries, highlighting a consistent drop in advanced economies.

Past automations—steam engines, electricity, computers—displaced jobs but birthed new ones. Factories needed managers, computers spawned IT sectors. But AGI? It’s different. Picture this: a machine that doesn’t just assemble cars but designs them, markets them, and even negotiates deals. No coffee breaks, no unions, no human frailties. In my research, I’ve modeled this as a “terminal shock”—the point where human labor becomes obsolete. What happens to a society built on work when work vanishes? Do we celebrate liberation or mourn the loss of purpose? I’ve often wondered, staring at historical charts like the one in my paper (Figure 1, illustrating the labor share’s inexorable dip), if we’re sleepwalking into a revolution without revolutionaries. To extend this, I’ve added Figure 7: Projected Labor Share Decline to 2050, extrapolating from current trends using data from the St. Louis Fed, which shows a 3.3 percentage point drop from 1980 to 2015 in advanced economies, potentially halving by mid-century under AGI acceleration. This new chart visualizes a steepening curve post-2030, factoring in AI-driven productivity gains that could exacerbate the imbalance.

Global Policy Divergences: A Patchwork of Preparations for the Inevitable

As AGI looms, nations aren’t waiting passively—they’re scrambling, each in their own way, creating a global mosaic of readiness. The United States, ever the capitalist vanguard, bets on private sector innovation, pouring billions into AI startups. But across the Atlantic, the European Union has taken a regulatory hammer to the anvil. The EU AI Act, fully in force by 2025, mandates transparency for high-risk AI and bans certain manipulative practices, with guidelines on General Purpose AI models published in July. Obligations for GPAI kicked in on August 2, emphasizing safety and compliance. Recent updates include draft guidelines from the European Commission in July 2025 clarifying GPAI provisions, and prohibitions on unacceptable risks effective from February 2. It’s a bold stance: Can we trust markets alone, or must we chain the beast before it devours us?

Turn to China, where AI isn’t just tech—it’s woven into the fabric of industrial planning. In April 2025, Xi Jinping called for AI applications across sectors, with a new guideline aiming for deep integration in six key areas by 2027. Funding for AI startups may have dipped, but state-driven initiatives like “Made in China 2025” continue to propel robotics and AI into manufacturing. The “AI Plus” plan, released in August 2025, promotes extensive integration across fields, with Shanghai’s plan accelerating AI in manufacturing. Japan and South Korea, meanwhile, double down on robotics as a bridge to post-AGI resilience. Japan’s Moonshot R&D Program promotes autonomous AI robots that evolve alongside humans, while both nations eye national strategies to counter global competition. And then there’s Singapore, my own backyard in spirit, pioneering governance with frameworks like AI Verify, updated for generative AI in 2025. New safety initiatives announced in February aim to align international standards.

This divergence provokes a haunting question: In a post-AGI world, will these policies unite us in shared abundance or fracture the globe into AI haves and have-nots? My Figure 2 maps country readiness for UBI, regulation, and AI trusts—Singapore leads in governance, but China’s industrial might could dominate. What if the unprepared nations become digital colonies? To illustrate further, I’ve included Figure 8: Comparative AI Policy Impacts 2025-2035, charting regulatory stringency versus industrial growth, drawing from RAND’s analysis of China’s AI policy accelerating progress.

UBI vs. UBS vs. AI Dividends: Redistributing the Fruits of Machine Labor

Ah, the heart of the debate: How do we share the wealth when machines do the work? Universal Basic Income (UBI) promises cash to all, no strings attached—but is it sustainable? Critics argue it’s fiscally ruinous; my models show high-income countries needing over 15% of GDP to make it meaningful. Trials worldwide, from Alaska’s oil dividend to proposed AI-funded schemes, highlight pros like reduced poverty but cons like disincentivizing work. Universal Basic Services (UBS) sidesteps cash, offering essentials like healthcare and education in-kind—practical, but does it stifle choice? A UNESCO discussion advocates shifting from UBI to UBS for contextual adaptability. Research from Exploring Economics supports UBS as an effective way to tackle poverty through universal services.

Then there’s my favored contender: AI Dividends, drawing from sovereign wealth funds fueled by AI productivity. Imagine taxing AI profits to create a perpetual fund, distributing shares to citizens. Yanis Varoufakis calls it a “Universal Basic Dividend,” essential for equity in an automated world. But skeptics warn it could entrench corporate power, turning UBI into a Silicon Valley handout. In my Excel simulations (Figure 3), AI Dividends outpace UBI in sustainability if productivity explodes exponentially. A recent paper on balancing UBI and public AI provision explores economic trade-offs, showing synergies in capability enhancement. Macroeconomic models indicate UBI can generate welfare gains and reduce inequality when neutral. Picture a family in 2040: Dad codes for fun, Mom pursues art, dividends pay the bills. But what if the fund favors elites? Is this liberation or a gilded cage? For added depth, Figure 9: UBI Cost vs. AI Dividend Sustainability Projections, incorporates data from Britannica’s UBI debate and Intereconomics’ economic analysis, forecasting UBI’s fiscal burden rising to 20-30% of GDP by 2040 while dividends stabilize via exponential growth.

Agentic AI and the Eclipse of Management: When Machines Lead

Agentic AI—autonomous systems that act like goal-oriented agents—doesn’t just replace workers; it obsoletes bosses. MIT Sloan research shows firms evolving into agent networks, with humans as mere custodians. Studies from the MIT Initiative on the Digital Economy reveal AI agents boosting collaboration by 137% in workflows. My models predict management roles plummeting from 62% in 2025 to 18% by 2050 in optimistic “Commonwealth” scenarios, or 5% in dystopian “Fortress” ones. Recent analyses highlight agentic AI redefining management for superhuman workforces and orchestrating intelligent operations. A systematic review notes improvements in productivity and innovation, though with privacy challenges.

Envision a boardroom in 2035: No heated debates, just AI orchestrating strategies flawlessly. Thrilling? Or terrifying? Figure 4 breaks it down industry-by-industry—finance sees 70% substitution, creative fields less so. But if management vanishes, who guards the ethical gates? Agentic AI could reshape 40% of enterprise apps by 2026, per recent findings. The provocation: Are we ready to trust machines with power structures we’ve built over centuries? To expand, Figure 10: Agentic AI Adoption Curves by Sector, uses data from McKinsey and IBM, projecting 30-50% automation in HR and operations by 2035.

Climate, Energy, and Compute Constraints: The Hidden Cost of Infinite Intelligence

AGI’s promise comes with a planetary price tag. PwC projects AI could offset its own energy use by boosting efficiency, potentially saving as much as it consumes by 2035. Yet, data centers might guzzle 13-16% more energy due to AI demand, exacerbating a 7% GDP drag from climate damages. Compute needs double every 2-3 years, clashing with net-zero goals. BloombergNEF forecasts US data-center power demand doubling from 35GW in 2024 to 78GW by 2035. The IEA estimates AI could reduce 1,400 Mt of CO2 emissions by 2035 in widespread adoption, but data centers could consume 8-15% of US electricity by then. Deloitte warns of a 30x surge in AI data center power by 2035 amid grid constraints.

Think of it: Servers humming in the Arctic, cooling costs soaring, while floods ravage supply chains. Figure 5 charts divergent paths—AI energy spiking against climate-induced GDP losses. The World Economic Forum warns of an “energy paradox,” urging balanced adoption. Provocatively, could AGI solve climate woes through breakthroughs, or will its thirst doom us? In 2025, with global AI investment skewed toward the US, the tension is palpable. Adding Figure 11: AI Energy Demand vs. Emission Reductions 2025-2035, this new visualization from Carbon Brief and LSE data shows gas-power for data centers doubling to 293TWh by 2035, offset by potential 3.2-5.4 billion tonnes CO2 reductions.

Philosophical Anchors of Meaning: Beyond Work, What Defines Us?

In a post-work utopia—or dystopia—employment fades, and meaning must anchor us anew. Yuval Noah Harari warns simulated realities could become the “opium of the masses,” addicting the jobless to virtual escapes. Joseph Stiglitz stresses equitable distribution of AI abundance, lest inequality spirals, with AI monopolies worsening divides. Shoshana Zuboff cautions against surveillance capitalism morphing into post-capitalist control, where data commodifies our souls. Harari’s “useless class” concept, from his TED ideas and books, predicts AI creating a new unemployable stratum, challenging liberal democracy. The Green European Journal explores post-work as a radical idea freeing society from job dominance.

My “Commonwealth” scenario envisions rising anchors like Purpose and Governance; “Fortress” favors Recognition and Play. Figure 6 visualizes this shift. Reflect: If AGI frees us, do we chase enlightenment or distraction? Harari’s “useless class” haunts me—what if we’re not ready for freedom? To provoke more, Figure 12: Evolving Anchors in Post-Work Scenarios, integrates Wikipedia’s post-work society overview, projecting a shift toward communal meaning by 2050.

Provocative Futures 2040–2050: Glimpses of Tomorrow’s Chaos and Wonder

To stir your imagination, let’s peer into speculative horizons. By 2032, “The Last CV”: LinkedIn yields to Purpose Graphs, mapping passions over resumes. 2038’s “Purpose Wars”: Ideological clashes over meaning allocation, echoing RAND’s geopolitical AGI scenarios. In 2042, “Algorithmic Faiths”: AGI spawns belief systems, per Millennium Project’s mixed-bag futures. By 2045, “Play Dividend”: States issue credits for simulations, as IMF scenarios predict wage collapses near AGI. Surveys predict AGI by 2040-2061, with Forbes outlining yearly paths to AGI by 2040. DNI’s Global Trends 2040 envisions AGI transforming societies, while IMF’s scenario planning warns of wage plummets near AGI.

These aren’t predictions but provocations: In a 2050 where AGI redefines society, per Pavel Luksha’s phases, will we thrive or fracture? Global Trends 2040 warns of fragmented worlds. Adding Figure 13: AGI Timeline and Economic Scenarios 2040-2050, this chart compiles predictions from AIMultiple and OpenExo, showing 50% AGI probability by 2040 and ASI transformations by 2045.

Conclusion: Capitalism’s Quiet Fade and the Call to Curate Meaning

As I conclude, capitalism doesn’t end in flames but in irrelevance—wages dissolve, management automates, and new anchors emerge. We face UBI traps, climate tensions, and philosophical voids, but also abundance. In 2025, with AGI accelerating, we must act: Forge equitable policies, balance energy, redefine purpose. What world do you want? Share your thoughts below—let’s dialogue before the machines do it for us.

Dr. Luke Soon, September 18, 2025