Every Generation Thinks This Time Is Different: The AI Job Myth

Summary

The video argues that AI should be understood through a recurring pattern: new technologies eliminate particular tasks and industries while enabling new products, businesses, and kinds of work. During each transition, the losses are visible sooner than the opportunities that emerge later.

Historical Examples and Business Implications

• Automobiles replaced much of the horse economy. Around 1900, New York City had roughly 200,000 working horses. Within about 30 years, many related jobs had disappeared. The auto industry and supporting services—such as mechanics, road construction, gas stations, motels, and auto insurance—created new work that people in 1900 could not readily have anticipated.
• Electricity, PCs, the internet, and smartphones similarly displaced some businesses and roles while enabling new industries, including appliances, software, IT services, e-commerce, digital marketing, apps, gig work, and the creator economy.
• The video cites research attributed in the subtitles to David “Otter” at MIT: more than 60% of the jobs Americans do today did not exist in 1940. The point is that people created new work using technologies that earlier generations could not fully envision.
• ATMs did not lead to fewer bank tellers overall. The number of U.S. tellers reportedly rose from about 300,000 in the mid-1970s to nearly 600,000 by 2010. ATMs lowered the cost of operating branches, banks opened more branches, and teller work shifted from routine cash handling toward customer assistance.
• The video applies Jevons Paradox to AI and software: when a resource or process becomes more efficient and cheaper, its use can increase. If AI reduces the cost and time required for software development, companies may build more software and pursue projects that were previously too expensive or impractical—potentially increasing demand for people who can use the tools effectively.

Frameworks and Operating Principles

• Jobs are bundles of tasks: AI may automate some tasks within a job without eliminating the whole role. Remaining tasks can be reorganized, and new task bundles or roles can emerge.
• Efficiency can expand demand: Lower costs may lead companies to do more, open new lines of work, or serve more customers—not simply reduce headcount.
• Look beyond the immediate disruption: The losses from automation are often apparent first. New markets and occupations may take longer to develop and can be difficult to predict in advance.
• Companies tend to fill capacity with more work: The presenters argue that organizations rarely conclude they have nothing left to do. As tools make work faster, businesses often find additional needs and opportunities.

Practical Recommendations

• Learn to use emerging tools, including AI, rather than treating them only as a threat.
• Build skills around the work the tools do not handle well, such as understanding the underlying problem and helping customers with more complex needs.
• Stay adaptable and open to new roles. Some future job titles may sound implausible today. The presenters cite “prompt engineer” as a recent example, describing it as a six-figure job.
• For businesses, consider whether automation can reduce costs, expand capacity, improve service, or make previously uneconomic projects viable—not only whether it can replace current tasks.

Metrics and Historical Predictions

• New York City working horses, circa 1900: Roughly 200,000; the related economy had mostly disappeared within about 30 years.
• U.S. bank tellers: Around 300,000 in the mid-1970s, rising to nearly 600,000 by 2010, despite ATM adoption.
• Jobs existing today but not in 1940: More than 60%, according to the research cited in the video.
• AI and unemployment: The presenters state that U.S. and global unemployment rates have remained largely unchanged since generative AI arrived; no specific figures or causal analysis are provided.
• The video recalls John Maynard Keynes’s 1930 prediction that technological progress could eventually produce a 15-hour workweek, noting that this outcome has not materialized.

Presenters and sources mentioned: Bri Kopecki and Jeff Crume; economists James Bessen and David “Otter” (as rendered in the subtitles); William Stanley Jevons; and John Maynard Keynes.