Hello! If you've been following AI news lately, you've probably felt a mix of fascination and unease. Watching ChatGPT pass the bar exam, AI generate artwork, and even write code, you've likely wondered: "Is my job safe?"
I've been deeply curious about this myself, so I had extensive conversations with various AI systems and organized those insights into this piece.
Today, I want to explore the inequality AI might bring and Universal Basic Income (UBI) as a proposed solution. We'll look at historical precedents and even touch on practical investment perspectives.
While this topic has been widely discussed, my focus here is: Why are AI leaders advocating so strongly for universal basic income?
As of 2025, the AI market has surpassed $7.9 billion. But behind this growth, inequality is quietly intensifying.
What the Industrial Revolution Taught Us
Looking back at history, technological revolutions have always been double-edged swords. Consider the 19th-century Industrial Revolution. Steam engines and factories sparked explosive economic growth, but they also created urban slums and widespread child labor.
IMF reports reveal a striking pattern: whenever the Gini coefficient (a measure of inequality) rises, property crime rates rise with it. In 1800s Britain, violent crime actually increased by over 20%.
The key point isn't simply that "technology is bad." The problem is that technological benefits weren't distributed equitably. Factory owners accumulated wealth while workers toiled 14-hour days in brutal conditions.
Will the AI Era Be Different?
Honestly, I think we're likely to see similar patterns repeat.
According to the Stanford AI Index 2025 report, AI development and usage are overwhelmingly concentrated in developed nations like the US and China. Even more surprising: AI threatens high-income jobs first. (Though I say "surprising," most of you probably already knew this.)
Goldman Sachs analysis suggests AI will impact 60% of jobs in developed countries, with white-collar professions—software developers, lawyers, accountants—being particularly vulnerable. This contradicts the common assumption that "AI will eliminate low-income jobs first."
Why? AI software, once developed, can be infinitely replicated. But cleaning robots or delivery robots have far higher physical production costs.
A New Problem: Time Inequality
Here's where an intriguing concept emerges: "time inequality."
The wealthy can leverage AI to boost productivity 10x or 100x while gaining more leisure time. Tasks that took an hour now take 10 minutes. Meanwhile, those with limited AI access must still work manually, investing more time for the same results.
According to the Anthropic Economic Index 2025, 40% of US workers use AI in their jobs, but they're predominantly high earners. Inequality is expanding beyond money into time itself.
If even time becomes unequal, I believe the acceleration of the rich-getting-richer, poor-getting-poorer dynamic will be staggering.
Is Universal Basic Income Really the Solution?
In this context, AI leaders like Elon Musk, Sam Altman, and Bill Gates are unanimously advocating for Universal Basic Income (UBI).
Musk has repeatedly emphasized that "universal high income will be essential when AI replaces all labor." Bill Gates has indirectly supported UBI by suggesting "AI could enable a two-day work week."
The logic goes: If AI dramatically increases productivity, wealth could be generated to support everyone's basic needs without requiring work.
In my view, their UBI advocacy seems aimed at system stabilization. They're concerned that AI-driven mass unemployment could trigger social unrest.
But History Warns Us
Did you know something similar to UBI existed in the past?
In 123 BCE, the Roman politician Gaius Gracchus introduced a revolutionary policy: the Annona, a free grain distribution program. It provided 20,000-30,000 Roman citizens with 33kg of grain monthly at no cost.
The intention was noble: ensuring food security and reducing social unrest. But what was the result?
Later historians criticized it as "panem et circenses" (bread and circuses). Citizens lost their work motivation, becoming absorbed in leisure and entertainment. Over time, the program expanded to include olive oil and wine, and Rome's fiscal burden grew enormous.
Of course, modern UBI differs from Rome's grain dole. Rome distributed goods; UBI provides cash or tokens. But the pattern of "unconditional support potentially fostering dependency" seems eerily similar.
The Dangers of Big Tech-Led UBI
Here's what worries me more: Past UBI experiments (Finland, Stockton, California) were government-led. But the AI-era UBI being discussed now is proposed by Big Tech companies.
This raises a question: What happens when corporate interests merge with UBI?
Imagine this scenario: Major tech companies say, "Sign up for our app and provide your data, and we'll give you monthly basic income." Formally it's UBI, but in reality, it becomes conditional support in exchange for personal information.
Research from the London School of Economics warns this could be "a new form of exploitation." Just as we traded personal data for free apps during the mobile revolution, similar patterns might repeat in the UBI era.
Perhaps in this age, we'll have to provide every life experience as data?
So What Should We Do?
I haven't only shared doom and gloom—there's hope too.
I believe right now, with AI tools available for free or reasonable subscription fees, might be the most equal period we'll see.
The World Economic Forum predicts that by 2025, AI will eliminate 85 million jobs but create 97 million new ones. The challenge? These new jobs require AI-related skills.
The key, then, is education and retraining. To keep pace with the AI era, we must learn continuously. Fortunately, AI itself can lower educational barriers. Personalized learning, free courses, real-time feedback—AI has tremendous potential to democratize education.
Policy Approaches Are Also Essential
Individual effort alone isn't enough. Government and society must act together:
- Transparent AI Governance: We need independent bodies to oversee AI development and use
- Enhanced Educational Access: Public investment to ensure everyone can receive AI education
- Proper UBI Design: Truly "unconditional" basic income, not conditional, with phased implementation
- Diverse Work Arrangements: Expanding various job formats and schedules (labor contract flexibility will become crucial)
An Investor's Perspective on AI
Now let's get practical. "How should I invest in this environment?"
⚠️ Disclaimer First: The following is not investment advice. All investment decisions should be made based on your own judgment and at your own risk. I recommend consulting with professional financial advisors.
Sectors Worth Watching
1. AI Education Platforms
- Examples: Duolingo (DUOL), Coursera (COUR)
- Why: As AI transforms jobs, retraining demand will explode
2. AI Infrastructure ETFs
- Examples: Global X Robotics & AI ETF (BOTZ), iShares Future AI & Tech ETF (ARTY)
- Why: Diversified exposure across the industry rather than individual companies
- AI investment is growing to $252.3 billion in 2025
3. Insurance/Healthcare
- Example: UnitedHealth Group (UNH)
- Why: The insurance industry could gain attention as social safety net needs intensify
4. Blockchain/Crypto
- Example: Coinbase (COIN)
- Why: Transparent governance and decentralization are emerging as AI-era alternatives
Investment caution: The AI sector is highly volatile with persistent regulatory risks. Approach with a long-term perspective rather than seeking short-term gains. And ultimately, it's your choice.
Conclusion: Asking AI About Solutions to Inequality
A Pure Analysis from AI's Perspective
Finally, I posed this question to AI:
"Humanity hasn't solved inequality for centuries. Looking at it through AI's learning methods, could we find a solution?"
I asked for an answer based purely on reasoning, without searching. I requested that AI apply its data-learning process—pattern recognition, causal analysis, datafication, and evolution—to the inequality problem.
Pattern Recognition: The Recurring Cycle of Inequality
The pattern AI discovered in historical data was clear.
Analyzing thousands of years of human history, it identified an intriguing cycle:
- Equality Movements → 2. Temporary Improvements → 3. Power Reconcentration → 4. Inequality Intensification
This pattern repeats from ancient Rome through the French Revolution, Russian Revolution, and modern welfare states. No society has ever achieved complete equality.
More interestingly, this pattern accelerates with each technological revolution. The Agricultural Revolution created landowners and serfs. The Industrial Revolution created capitalists and laborers. The AI revolution will likely divide us into data owners and data providers.
Causal Analysis: Why Hasn't It Been Solved?
AI identified three root causes of this pattern.
First Cause: Human Nature's Self-Interest
From an evolutionary psychology perspective, humans have an instinct to "accumulate more." This was a survival strategy. In tribal times, leaders who secured more resources improved their group's survival rates. The problem is this instinct still operates in modern society.
But here's a crucial insight: Self-interest itself isn't the problem—the system in which self-interest operates is. If we can create positive-sum game structures rather than zero-sum ones, the same self-interest could actually promote equality.
Second Cause: Self-Reinforcing Feedback Loops
Inequality reinforces itself, much like how biased data in AI training produces biased results.
The wealthy → Better education → Better jobs → More wealth → Better education for children...
Without breaking this loop, inequality calcifies across generations. This is why the "American Dream" becomes increasingly unattainable.
Third Cause: Social Justification Mechanisms
More insidiously, inequality gets packaged in ideology. "Work hard and you'll succeed," "The market rewards fairly," "Merit-based differences are natural"—these narratives shift structural inequality onto individuals.
From an AI perspective, this is a "labeling error." If you misclassify a problem's cause, you'll generate wrong solutions.
Datafication and Evolution: AI's Proposed Solutions
Now for the most crucial part. What solutions emerge when we apply AI's learning process to solving inequality?
Solution 1: Dynamic UBI System
Past UBI experiments were "static"—giving everyone the same amount, which could foster dependency like Rome's grain dole.
AI proposes adaptive UBI:
- Analyze individuals' real-time situations (unemployment, health, education needs)
- Automatically adjust support levels based on need
- Provide customized support (education credits, healthcare vouchers) rather than just cash
- Design gradual reduction to avoid discouraging economic participation
This is like AI adjusting learning rates to optimize social support.
Solution 2: AI-Based Educational Equity
AI analysis suggests approximately 70% of inequality stems from educational gaps. This is where AI's true potential could shine.
Imagine:
- Every student having a personal AI tutor
- One-on-one personalized education—once exclusive to wealthy families—becoming universal
- Optimal learning experiences regardless of language, location, or disability
But the key isn't just creating AI education tools—it's ensuring access through policy. The best AI tutor is useless without internet access. Ultimately, this speaks to equalizing opportunity.
Solution 3: Transparent AI Governance
Much inequality stems from "information asymmetry." Those in power monopolize information and make opaque decisions.
AI could change this:
- Real-time transparency in government budget execution
- Automatic monitoring of corporate wage gaps
- Independent AI auditors verifying fairness in resource allocation
Just as blockchain made financial transactions transparent, AI could make power exercise transparent.
Solution 4: Hybrid Economic Model
Pure capitalism breeds inequality; pure socialism loses efficiency. History has proven both.
AI proposes dynamic balance:
- Maintain market innovation incentives
- Trigger automatic redistribution mechanisms when inequality exceeds certain thresholds
- AI monitors economic indicators in real-time, acting as a "thermostat"
This resembles how AI uses regularization to prevent overfitting during training.
Honest Limitations: No Perfect Solution Exists
But AI was also honest: It hasn't found a "perfect solution."
Because inequality is a chaotic system. Small changes produce unpredictable results. Moreover, human free will and emotions can't be perfectly predicted by algorithms.
The proposed solutions could create new problems:
- Dynamic UBI could lead to excessive surveillance
- AI education might produce homogenized thinking
- Transparent governance could violate privacy
- Hybrid models might ultimately collapse under complexity
The most important limitation: AI is just a tool. No matter how good the algorithm, it's useless without human will and ethics to deploy it properly.
Why There's Still Hope
I asked AI to think long and carefully. It offered a cautiously optimistic conclusion.
Because this time could be different.
There's a crucial difference between past technological revolutions and the AI revolution: AI recognizes patterns and makes predictions. For the first time, we have tools to measure inequality intensification in real-time and intervene early.
After the Industrial Revolution, labor laws took 100 years to develop. But in the AI era, we could detect and respond to problems in years—even months.
More importantly, humanity has learned. We now understand how inequality destroys societies. The French Revolution, Russian Revolution, Great Depression—we've learned from failed patterns.
Solving Inequality Is a Journey
From the Industrial Revolution to AI, inequality patterns are clearly repeating. But history isn't simple repetition—it's a spiral of progress. Similar problems return, but we gradually find wiser solutions.
As the memorable line from Interstellar goes: humanity will find a way, as we always have.
Analyzing inequality through AI's lens revealed it's not an unsolvable "fate" but a "dynamic system" requiring continuous management. Perfect equality is impossible and perhaps undesirable. But preventing extreme inequality and ensuring basic opportunities for everyone is entirely achievable.
So shouldn't we start with what's possible?
AI concluded with these words:
"Solving inequality isn't a destination—it's a journey. We must keep adjusting, learning, and adapting."
I deeply resonated with this message. Rather than waiting for perfect answers and doing nothing, starting with small changes matters more.
What do you think about inequality in the AI age? Can we break the pattern? Share your thoughts in the comments. I'd love to learn through our conversation together.
Key References
- IMF AI Adoption and Inequality Report (2025)
- Stanford AI Index Report (2025)
- Goldman Sachs: The Potentially Large Effects of AI (2023)
- WEF Future of Jobs Report (2025)
- Anthropic Economic Index (2025)
Comments
Post a Comment