When AI Starts to Take Jobs: The Future of the US Economy – A Time of Celebration or Divisions?
Hello everyone, I’m your financial journalist. Recently, a report from the AI community has caused quite a stir.
This report was produced by the economic research team at Anthropic, the company behind the Claude AI model, and it’s titled “The Economic Implications of Transformative AI.” In simple terms, they did some calculations to ask: If AI really becomes as powerful as people imagine, what will the US economy look like in the future? Where will the money flow? Will we still have our jobs?
The conclusions are somewhat disheartening: The overall economy will grow significantly, but how and to whom that growth will be distributed could be extremely unfair. Moreover, many white-collar workers might have to learn how to fix pipes or become nurses.
Below, I’ll break down this comprehensive report into five parts that everyone can understand, and we’ll discuss the logic behind these findings.
---
1. Three Possible Futures: From a Steady Growth to Explosive Expansion
The report doesn’t offer a single outcome but outlines three scenarios based on different rates of AI development. Imagine driving a car:
- Scenario One: Moderate Growth (Similar to the Internet)
- Description: AI gradually improves efficiency, but it doesn’t completely disrupt daily life.
- Result: By 2030, the US GDP will increase from $30 trillion to $34.1 trillion, an increase of just 1.6% compared to without AI.
- Impact: Life will be relatively unchanged, with stable employment and normal unemployment rates. This is the most reassuring scenario.
- Scenario Two: Substantial Impact (AI Handles 50% of Work)
- Description: AI becomes so advanced that it can perform 50% of knowledge-based tasks independently (such as coding, data analysis, customer service).
- Result: GDP will double, reaching $36.3 trillion, an 8.3% increase.
- Impact: The economy grows, but white-collar salaries may stagnate because AI replaces some jobs. However, those in manual or non-knowledge-based roles might see a slight increase in income due to increased productivity.
- Scenario Three: Extreme Impact (AI Outperforms Human Intelligence)
- Description: AI not only performs tasks but also improves itself, surpassing humans in most mental work and creating few new jobs for white-collar workers.
- Result: GDP will grow by an astonishing 15%, doubling by 2030 to $44.4 trillion, a 32.4% increase.
- Impact: This is a world of extreme wealth disparity; while some become very rich, many lose their jobs and incomes.
Journalist’s Note: Most people currently expect a scenario similar to Scenario Two, but technological breakthroughs are often non-linear, and once a critical point is reached, the situation could quickly shift to Scenario Three.
---
2. Where Does the Money Go? Bosses Earn More, Workers Less
A key finding of the report is that AI will significantly shift wealth towards capital:
- Current Distribution: In the US, 60% of national income goes to workers, and 40% to capital owners.
- With AI:
- Scenario One: Capital owners get a slightly higher share (40.6%), and workers 59.4% (slight change).
- Scenario Two: Capital owners get 43.9%, and workers 56.1% (increasing gap).
- Scenario Three: Complete Reversal: Capital owners take 54.8%, and workers only 45.2%.
What Does This Mean?
In the extreme scenario, although GDP grows by 32%, workers’ total wages may not increase much because AI is a form of capital. The more AI is used, the lower the costs for employers, and the higher their profits, with fewer jobs created for workers.
In Simple Terms: In the past, more people meant more power and a larger share of the wealth; now, machines (capital) take the larger portion, leaving workers with less.
---
3. The White-Collar Nightmare: From Coding to Pipe Repairing
A particularly striking statement in the report is: “Programmers and call center staff may have to switch to jobs less affected by AI, such as electricians or nurses.”
Why? Because AI excels at processing information, text, and data—these are core skills of knowledge workers.
- What AI Can Do: Code, write content, analyze data, answer customer questions, even perform legal research.
- What AI Can’t Do: Cut hair, perform surgeries, repair pipes, care for the elderly, lift heavy objects.
In the extreme scenario, the proportion of knowledge workers in the workforce will drop from 62.2% to 48.7%. This means that 1-2 out of every 10 white-collar workers could lose their jobs or be forced to switch careers.
The report predicts that only 5.2% of workers will successfully transition to non-knowledge-based roles, and 8.3% will face “structural unemployment”—they want jobs but lack the necessary skills, while new roles (like electricians) require specialized training.
Journalist’s Note: This doesn’t mean white-collar workers are useless; it means the barriers to high-paying jobs based on pure mental skills are disappearing. The days of high salaries based on education and experience might be over. Future high-paying roles might require knowledge of AI or skills using AI tools, not just advanced programming.
---
4. Why Structural Unemployment? Switching Careers Is Difficult
Many ask: If AI takes white-collar jobs, can’t white-collar workers switch to blue-collar roles? Theoretically, yes, but in reality, it’s hard. This is “structural unemployment”:
1. Skill Mismatch: A programmer who has been coding for 10 years would have to learn new skills from scratch, which is time-consuming and costly.
2. Psychological Barrier: Moving from a comfortable office to a physically demanding job is challenging for many.
3. Training Costs: Society needs to invest heavily in retraining. If governments and companies don’t provide support, individuals will struggle.
The report suggests that in the extreme scenario, many white-collar workers will be in a “transition period” with no job and no immediate ability to switch to manual work.
Journalist’s Note: The future challenge is not just about having a job but about how to help people transition quickly. Without proper vocational training and social support, this transition could lead to significant social unrest.
---
5. An Uncertain Future: What Should We Worry About?
The report also reminds us not to panic, as the future is still uncertain:
- Public Perception: An Anthropic survey of over 10,000 Americans shows that most expect a Scenario Two, with a 10% GDP increase and a 5% unemployment rate. People realize the challenges, but they don’t see it as the end of the world.
- Key Variables:
- AI’s Limits: How intelligent will AI become? Will it hit a bottleneck?
- Policy Responses: Will governments introduce taxes on capital, provide universal basic income (UBI), or strengthen vocational training?
- Business Decisions: Will companies use AI to cut costs or to improve efficiency and pay higher wages?
Report’s Limitations:
- It doesn’t account for the impact of robots on manual labor; if robots can also perform physical tasks, even more jobs could be lost.
- It doesn’t consider policy interventions; strong government actions could change the outcome significantly.
- It doesn’t consider the massive investment needed for data center construction, which could create new jobs.
Journalist’s Summary: This report isn’t meant to scare people but to raise awareness. AI will drive economic growth, but the benefits will be unevenly distributed.
For individuals, the rules for survival in the future are changing:
1. Diversify Skills: Combine mental, physical, emotional, and creative skills—these are harder for AI to replace.
2. Continuous Learning: Skill updates will be faster; lifelong learning is essential.
3. Pay Attention to Policy: Future social security and tax policies will determine how much of the wealth you can access.
In One Sentence: AI will make the country wealthier, but not necessarily everyone. At this critical juncture, how the wealth is distributed is more important than how much wealth is created.