Will AI cause mass unemployment? Maybe not.

Current Automation Reality and Trends

• Over 50% of people fear job displacement due to AI; experts believe AI will eventually perform every economically important task more efficiently than humans.
• AI currently generates photorealistic videos, drives taxis more safely than humans, and performs accurate medical diagnoses.
• Historical patterns (e.g., ATMs, textile production) show automation often initially increases wages and employment by lowering costs and expanding markets before potentially declining.
• Economic models suggest a "rising then falling" wage trajectory: wages could rise tenfold until ~2037 before crashing if 100% task automation is achieved, whereas retaining just 1% human involvement allows indefinite wage growth.
• The transition from human labor to automation in high-income countries (e.g., agriculture) historically led to incomes 100 times higher, though the immediate transition period will be disruptive.
• Current data indicates routine white-collar jobs (earning $100k–$200k) are more vulnerable to automation than many blue-collar roles, as digital data for training is abundant while physical movement data is scarce.

Four Categories of Skills Increasing in Value

• Hard to Automate Skills: Physical tasks in unpredictable environments (e.g., complex surgery, plumbing) resist automation due to data scarcity; social tasks with long horizons (e.g., organizational strategy, conceptual insight) remain difficult for AI to replicate.
• AI Deployment and Complementary Skills: Roles requiring the direction of AI agents (e.g., spotting problems, writing specifications, managing AI teams) will become more valuable as AI capability increases.
• Skills Producing Scalable Goods: Jobs where cheaper/better AI supply drives higher demand (e.g., healthcare, luxury experiences, ride-sharing) will see employment growth, unlike compliance-based roles where efficiency reduces headcount.
• Specialized, Hard-to-Learn Skills: Trades requiring unique physical abilities or deep specialized knowledge (e.g., data center electricians in Virginia, where wages are 37% above average) will see faster value growth due to labor market bottlenecks.

Specific High-Value Skill Recommendations

• Universal Skills:
  • Deploying AI to solve real-world problems and navigating human-in-the-loop gaps.
  • Personal health and resilience management (e.g., sleep, exercise, CBT therapy for mental health).
  • Personal effectiveness (e.g., goal setting, meeting management) and the ability to work on ill-defined problems over long horizons.
  • Prioritization, forecasting, and decision-making to determine high-impact activities.
  • Social skills for leadership, coordination of small AI-managed teams, and building trust.
  • Rapid learning capabilities, facilitated by AI-powered tutoring.
• Core Professional Skills:
  • Leadership and management, particularly for organizing small, high-leverage teams.
  • Operations management (recruitment, financial systems, office management) to handle day-to-day organizational complexities.
  • Policy and political skills to navigate government systems where human decision-making remains a social demand.
  • Communications focused on taste, strategy, and authentic relationship building rather than volume content creation.
• Specialist Expertise Areas:
  • Machine Learning research and engineering, specifically applied to AI safety or solving global problems.
  • AI governance and strategy, bridging economics, law, ethics, and macro strategy.
  • Cyber and information security, critical for protecting biotechnologies and AI systems from theft.
  • Expertise in emerging powers, particularly China-West relations and multipolar coordination.
  • AI hardware expertise (chip design and manufacturing) as the "oil" of the next century.

Strategic Career Decisions and Advice

• Job seekers should avoid specializing exclusively in routine knowledge work (e.g., basic law, accounting, translation) without considering AI's ability to automate these tasks.
• The optimal strategy is to target roles satisfying at least two of the four value categories while being faster to learn relative to their potential value.
• Combining technical skills with social/policy skills creates high-value "bridge" roles (e.g., understanding AI/synthetic biology alongside government policy).
• Career choice should prioritize transferable skills to adapt to shifting market demands rather than seeking a single job that is immune to automation.
• Specific examples of rising value include: radiologists (increased productivity via AI), software engineers (wages up, but junior hiring down), and taxi drivers (wages down, volume up).
• Physical infrastructure constraints (robot production limits, computing power costs) will slow the automation of physical and low-margin digital jobs, extending the window for human labor in these sectors.