(From Perplexity…)

The Post-AGI World: Understanding the Economic Future in Simple Terms

Artificial General Intelligence (AGI) refers to AI systems that can perform all tasks humans can do—from complex problem-solving to physical work. When AGI arrives, it will fundamentally transform how our economy works. Let me explain this transformation in straightforward terms.

How the Economy Will Change

The Shift from Human Labor to Computational Power

Today, our economy runs on human work—people earn wages, spend money, and drive economic growth. In a post-AGI world, this changes dramatically. The economy will be driven primarily by computational resources (computing power) rather than human labor. Think of it this way: just as factories replaced human muscle power during the Industrial Revolution, AGI will replace human brain power across virtually all economic activities.arxiv+2

Economic output will grow based on how much computing power society can deploy, not how many people work. If we double computing power, we can double economic output—but doubling the human workforce won’t matter much anymore.nber+1

What Happens to Jobs

The outlook for human employment is stark: AGI systems will automate almost all work. Here’s why:80000hours+2

• AGI can work 24/7 without sleep, vacation, or retirement[80000hours]​
• It costs pennies in electricity versus hundreds of dollars per hour for human professionals[80000hours]​
• It can be copied infinitely—turning 10,000 AI researchers into 10 million overnight[80000hours]​
• It performs both cognitive work (like management, medicine, research) and physical work (through advanced robotics)nber+1

Think of jobs in two categories:

1. Bottleneck work (essential for growth): energy production, infrastructure, scientific research, security. AGI will automate all of these.campuspress.yale+1
2. Supplementary work (nice but not essential): arts, hospitality, entertainment, crafts. Some of this might remain with humans, not because AGI can’t do it, but because it’s not worth using scarce computing resources on it.campuspress.yale+1

How Income and Wealth Work

Here’s the surprising part: even though humans can’t compete with AGI for jobs, most people won’t be poor. Here’s why:nber+1

When AGI makes the economy 100 times or 1,000 times larger, governments can maintain their current level of taxation (say, 30-50% of GDP) and redistribute that wealth. Since the total pie is so much bigger, everyone can receive what would be billionaire-level wealth today through mechanisms like:[80000hours]​

• Universal Basic Income (UBI): Regular payments to all citizens regardless of employmentarxiv+2
• Public ownership of AGI: Treating computing resources like public utilitiesarxiv+1
• Progressive taxation: Taxing AGI capital (computing resources) to fund social programsarxiv+1

The math is simple: if the economy grows 1,000-fold and we redistribute even 10% of it equally, everyone becomes far richer than they are today—without working.[80000hours]​

Understanding Wages in Simple Terms

Why Wages Fall to Zero (or Nearly Zero)

In today’s economy, your wage reflects the value you create. In an AGI economy, your wage is capped by the cost of replicating your work with computers.campuspress.yale+1

Example: If a computer can do your job using $5 worth of electricity, your wage can’t be much higher than that—employers would just use the computer instead. Since AGI operates at incredibly low cost (think pennies per hour of work that costs $100/hour today), human wages become negligible.epoch+1

The Income Share Problem

Currently, workers earn about 50-70% of all economic output. In a post-AGI world, labor’s share approaches zero—almost all income goes to owners of computing resources. This happens because:epoch+3

• Human brainpower: ~10^18 computations per second for all 8 billion people
• Potential computing power: could reach 10^54 computations per second[nber]​

Human computational contribution becomes a rounding error—like 0.0000001% of total economic capacity.[nber]​

The Critical Challenge: How to Pay for UBI

The Funding Problem

This is where things get tricky. Many people suggest UBI as the solution, but there’s a catch: traditional UBI proposals rely on taxing labor income. In a post-AGI world, there’s little labor income to tax.linkedin+1

Example: Switzerland’s UBI proposal would need about 30% of GDP. They planned to get two-thirds of that from taxing wages—but if AGI eliminates jobs, those wages disappear, creating a massive funding gap.lesswrong+1

Potential Solutions

To make UBI work, society needs new tax sources:

1. Tax computing resources directly: A small tax (0.5-1%) on AGI operations could fund UBI for everyone[aicompetence]​
2. Tax AI capital profits: When AGI becomes 3-5 times more productive than humans, taxing 33% of profits could fund UBI equal to 11% of GDParxiv+1
3. Treat compute as public property: Like natural resources, distribute ownership of computing capacity to all citizensnber+1

The key insight: UBI in an AGI economy must be funded by taxing the productive assets (computing power), not the labor that’s been replaced.aicompetence+1

What Daily Life Looks Like

Radical Abundance

Imagine having personal teams of hundreds of AI assistants available 24/7:[80000hours]​

• Healthcare: AI doctors with knowledge of millions of cases, available instantly, perfect diagnosis[80000hours]​
• Education: Personal AI tutors for you or your children, optimized for individual learning styles[80000hours]​
• Daily tasks: Robot assistants handle cooking, cleaning, home maintenance perfectly[80000hours]​
• Creative pursuits: AI collaborators for any hobby—music, art, writing, invention[80000hours]​

The cost of all services plummets because AI labor is nearly free. A medical consultation that costs $200 today might cost 1 cent.[80000hours]​

The “Human Touch” Question

You might wonder: won’t people still want human teachers for their kids, or human doctors?

Research suggests not really. Consider robot nannies: they never get tired, never abuse children, have knowledge from raising millions of kids, available 24/7, and can be programmed with parents’ values. Most parents would choose this over even the best human nanny who can only work 8 hours a day and costs $50,000/year.[80000hours]​

Similarly with other services—when AI versions are dramatically better, cheaper, and more available, human preference shifts.[80000hours]​

Loss of Purpose

Here’s the psychological challenge: historically, work provided not just income but also meaning and social recognition. In an AGI world, “we would not be missed”—human effort is no longer needed for society to function.[nber]​

People will need to find purpose through:

• Creative pursuits and hobbies
• Relationships and community
• Learning and self-improvement
• Activities chosen for enjoyment, not economic necessity

This is a profound shift in human identity.

Timeline and Transition

How Fast Could This Happen?

Leading AI researchers and economists estimate:

• 50% chance AGI arrives by 2030[aicerts]​
• Economic doubling time: Could drop from 15 years today to under 1 year after AGI[80000hours]​
• Job displacement: 30-50% of entry-level white-collar jobs eliminated within 1-5 years of AGI[freethink]​
• Unemployment spike: Could reach 10-20% as transition occurs[freethink]​

Two Possible Transition Paths

1. Gradual (compute-limited): Jobs disappear slowly as computing power expands. Workers can retrain and adjust. Wages decline smoothly.campuspress.yale+1
2. Sudden (algorithm-limited): Breakthroughs happen unpredictably. One day radiologists are in high demand; the next day, AI can read X-rays better. Sudden job losses and wage crashes.campuspress.yale+1

Most experts expect something between these extremes, but closer to sudden—making preparation urgent.freethink+1

Key Economic Principles Explained Simply

Let me clarify some technical concepts in simple language:

Compute-Equivalent Units (CEU)

Think of this as “how much computing power would replicate your job?” If you’re a lawyer earning $200/hour, and AI can do your job using $2 of electricity, your CEU is $2. Your future wage can’t exceed your CEU—otherwise employers use AI instead.campuspress.yale+1

Bottleneck vs. Supplementary Work

• Bottleneck: Must grow for economy to grow (energy, food, infrastructure, science)
• Supplementary: Economy can grow without it (poetry, artisan crafts, live theater)

AGI automates all bottlenecks first because that’s where the economic value is. Supplementary work might stay human, but wages stay low because it’s not essential.campuspress.yale+1

The Decoupling of Wages and GDP

Today: GDP ↑ = Wages ↑ (economic growth lifts all boats)

Post-AGI: GDP ↑ dramatically, but wages → 0 (growth happens without human contribution)epoch+1

This is why redistribution becomes essential—the connection between economic output and human income breaks.epoch+1

Major Risks and Challenges

Inequality Explosion

Without intervention, wealth concentrates in the hands of those who own computing resources. A few companies (think Google, Microsoft, OpenAI) could own all productive capacity, while billions have no income.arxiv+2

This creates what some call “techno-feudalism”—a new aristocracy of AI owners, and a powerless majority.[arxiv]​

Aggregate Demand Collapse

Here’s the paradox: AGI produces enormous quantities of goods, but if people have no income, who buys them? This could cause:arxiv+1

• Economic instability and deflation
• Overproduction crises
• Social unrest when people see abundance but can’t access it

This is why economists stress that redistribution isn’t charity—it’s necessary for market stability.arxiv+1

Political and Social Instability

When large populations lose economic purpose and livelihoods, societies become unstable:

• Increased political extremism and populismarxiv+1
• Potential for violent revolution if wealth isn’t shared[epoch]​
• Breakdown of social cohesion built around work[nber]​

Geopolitical Competition

Countries that adopt AGI first gain massive advantages. A nation with AGI growing its economy 100%/year will dwarf competitors growing at 3%/year within a decade. This creates pressure for:[80000hours]​

• Racing to AGI without adequate safety measures
• International conflict over computing resources
• Arms races in AI-powered weapons[80000hours]​

What Needs to Happen (Policy Solutions)

Based on economic research, here’s what experts recommend:

1. Establish New Revenue Sources

• Tax computing power used for production (e.g., 0.5% “compute excise tax”)[freethink]​
• Tax AI capital profits progressively (30-50% on profits over certain thresholds)aicompetence+1
• Charge for access to public data used to train AI[aicompetence]​

2. Create Distribution Mechanisms

• Universal Basic Income sufficient for comfortable living (15-30% of GDP)linkedin+1
• Public ownership stakes in AI companiesarxiv+1
• Sovereign wealth funds investing in computing resources, distributing dividends[80000hours]​

3. Manage the Transition

• Retraining programs for displaced workersarxiv+1
• Lifelong learning accounts funded by AI taxes[freethink]​
• Social support for people whose identities are tied to work[nber]​

4. International Coordination

• Treaties to prevent AI arms races[80000hours]​
• Sharing of natural resources needed for computing (energy, rare materials)freethink+1
• Global redistribution mechanisms for extreme inequality[80000hours]​

The Optimistic Case

Despite the challenges, there’s a genuinely positive scenario:

Material Abundance

Post-AGI humanity could have:

• Extreme longevity through advanced medical AI[80000hours]​
• Elimination of poverty and hunger[80000hours]​
• Nearly unlimited access to education, entertainment, and cultural goods[80000hours]​
• Freedom to pursue passions without economic pressurenber+1

Scientific Acceleration

AGI conducting science means:

• Cures for all diseases discovered rapidly[80000hours]​
• Climate change solutions developed and implemented[80000hours]​
• Space exploration and off-planet resources[80000hours]​
• Understanding of fundamental physics advancing exponentially[80000hours]​

Human Flourishing

If managed well, humans could:

• Spend time on relationships, creativity, and personal growth[nber]​
• Have AI collaborators for any endeavor[80000hours]​
• Live in a post-scarcity society where basic needs are trivially metfreethink+1

The total wages humans earn collectively will likely be higher than today (not lower), even though individual wages fall—because the economy is so much larger.campuspress.yale+1

The Pessimistic Case

If managed poorly:

• Extreme concentration of wealth in a few tech companies/nationssyzygy-group+1
• Mass unemployment without adequate social support[epoch]​
• Loss of human agency and political power[arxiv]​
• Potential for authoritarian control using AGI-powered surveillance[arxiv]​
• Environmental damage from energy demands of computing[freethink]​
• Social breakdown and violence[epoch]​

---

Simple Summary

Before AGI: Humans work → earn wages → buy goods → economy grows

After AGI:

• Computers do all work → produce massive output → humans get paid from taxes on computing → everyone lives in abundance
• Human labor becomes irrelevant to production
• Wages fall to near zero (can’t compete with AI costing pennies)
• BUT total wealth explodes 100-1000x
• Society must redistribute this wealth (via UBI or similar) or face crisis
• Daily life becomes one of radical leisure and abundance—or poverty and instability, depending on policy choices

The bottom line: AGI could create unprecedented prosperity, but only if we redesign our economic systems to distribute wealth based on citizenship rather than labor. Without this shift, we risk a future where AI generates enormous value that humans cannot access.

---

Key Technical Terms Explained

Since you asked for technical terms to be explained simply, here are the main ones:

AGI (Artificial General Intelligence): AI that can do any intellectual task a human can do, at human level or better. Different from today’s “narrow AI” that only does specific tasks.arxiv+1

Compute/Computational Resources: The raw processing power of computers, measured in operations per second. Think of it like electricity—a fundamental resource that powers the digital economy.campuspress.yale+1

Universal Basic Income (UBI): Regular cash payments to all citizens, no strings attached, enough to cover basic living expenses. Like Social Security for everyone, regardless of age or employment.wikipedia+1

Labor Share of GDP: The percentage of economic output that goes to workers as wages. Currently ~60% in most countries; post-AGI, approaches 0%.campuspress.yale+1

Marginal Productivity: The additional value created by one more unit of input (one more worker, one more computer). Your wage equals your marginal productivity—what you add to output.[nber]​

Wealth Redistribution: Transferring money from high-income individuals/corporations to lower-income individuals through taxes and social programs. Essential in post-AGI economy because production and consumption become disconnected.arxiv+1

Total Factor Productivity (TFP): How efficiently an economy converts inputs (labor, capital) into outputs. AGI massively increases TFP.[arxiv]​

Capital vs. Labor: Capital = equipment, buildings, computers (things you own). Labor = human work. Today’s economy needs both; post-AGI economy needs only capital.arxiv+1

Progressive Taxation: Tax rate increases as income/wealth increases. Proposed as a way to fund UBI from AI profits without crushing innovation.arxiv+1

Social Contract: The implicit agreement between citizens and government about rights, responsibilities, and mutual support. AGI requires “renegotiating” this because the assumption that you work for income breaks down.arxiv+1

This transformation represents one of the most profound shifts in human history—comparable to the Agricultural and Industrial Revolutions, but happening far more quickly. The key is preparing now while we still have time to shape the outcome.