The AI jobs debate: alarm, optimism, and real danger
Will the arrival of AI create a job crisis? Most predictions suggest a May Day for entry-level jobs in all sectors.
For the past few days, social media timelines and news headlines have been filled with debates over the future of humanity due to AI.
No aspect of the AI discussion generates more heat, more confusion, or more fear than its impact on employment. This debate involves some of the most prominent creators and most credible critics of the technology. The worrying thing is that they cannot even agree on the basic direction of the destiny of employment.
Alarm: Dario Amodei
In an interview with Axios in May 2025, Amodei warned that AI could eliminate half of all entry-level white-collar jobs and increase unemployment by 10–20% within one to five years. He doubled down on his prediction in an extensive essay in January 2026: “New technologies often bring shocks to the labor market, and in the past, humans have always recovered from them. But I worry that this is because these past shocks have affected only a small fraction of the full potential range of human capabilities, leaving room for humans to expand into new tasks. The effects of AI will be much broader and will occur much faster.” (Dario Amodei, The Adolescence of Technology).
His most profound observation: “Across a wide variety of tasks, AI is moving from the bottom to the top of the capability ladder. Thus we are at risk of a situation where, rather than impacting people with specific skills or specific occupations (which can be adapted by re-training), AI is impacting people with certain intrinsic cognitive qualities, namely low intellectual ability (which is hard to change). It is not clear where these people will go or what they will do, and I worry that they may become unemployed or very The low-wage underclass.”
They also shared an internal Anthropic metric that reveals the trajectory: 60% of users currently use AI for augmentation, 40% for full automation – and the latter figure is rising.
Counterargument: Huang, and Schmidt
Nvidia CEO Jensen Huang has been one of the most vocal optimists. His argument: “You’re not going to lose your job because of AI, but you’re going to lose your job to someone who uses AI”. This has become the default framing for the camp that the abundant era awaits.
“I am confident that AI will boost productivity, revenue growth and therefore more hiring,” Huang told a magazine.
Former Google CEO Eric Schmidt also argued for coexistence. At TED 2025, he said: “If you’re not using this technology, you won’t be relevant compared to your peer groups, your competitors, and people who want to be successful. Adopt it, and adopt it fast.”
This optimism has a long and proven history. To understand the present alarm it is worth carefully reviewing the past and understanding exactly what the cycle of evolution tells us.
Arc: from muscle to brain
The evolution of human labor is the story of shifting the burden from muscles to machines and then to the brain.
For millennia, humans were hunter-gatherers, based primarily in Africa. Survival was linked to physical endurance and local ecology. The Agricultural Revolution (about 10,000 BC) transformed hunters into farmers.
Starting with small farms in the Fertile Crescent of the Levant, labor gradually became repetitive and tied to land ownership. This created the first surplus in history, giving rise to cities, specialized trades, artisans, and a warrior class to defend them.
Europe’s 19th century Industrial Revolution was the first true breakthrough in comparison to AI. The artisan’s hand was replaced by steam and steel. Skilled artisans such as weavers were replaced by less-skilled factory operators.
Although cruel to displaced artisans, it reduced the cost of goods so much that a new middle class emerged. This gave rise to the 40-hour work week and mass education.
In the mid-20th century, automation moved into factories, pushing humans into white-collar roles. Gradually, spreadsheets replaced bookkeeping; Computers replaced the typing pool.
We are now entering the next phase of this evolution: machines equipped with artificial intelligence are powering the automation age.
What history really shows
The economic verdict of the evolutionary arc is clear: it created enormous wealth, raised standards of living, and generated employment which it destroyed. New industries, new cities, new lines of work eventually absorb those displaced by machines.
But there are two caveats. The changes generated new roles but required different skills, and they appeared on different time scales in different cities, requiring different educational backgrounds than the jobs they replaced.
The second factor is time. Each revolution created more opportunities, but ultimately the dominant word is. For many categories of skilled workers, such as the handloom weavers of Yorkshire, the recovery took no more than a few years. They did not live to see it.
Each historical recovery from technological dislocation took decades, involving large-scale disruption and slow growth of new industries. The question in the current debate is whether institutions, education systems, and political frameworks can compress that transition time or whether the pace of AI deployment simply outpaces them.
entry level problem
Most predictions suggest a May Day for entry-level jobs in all sectors. And this is the biggest danger.
Historically, people learned trades by starting with simple, repetitive tasks and moving up. AI is closing that cycle at the entry point.
Consider what a first-year attorney does: research, document review, drafting standard clauses, preparing summaries of case law. What a junior financial analyst does: Building models from templates, formatting decks, running sensitivity analyses. What a junior journalist does: Writing press releases, covering regular council meetings, transcribing interviews. This is not a glamorous job. But these are tasks through which people learn to think within a discipline. AI is replacing them faster than anyone has prepared a replacement for them.
The Class of 2026 is entering a market where more than half of employers view job prospects as poor or uncertain. The question is not whether the economy will eventually generate new categories of work; History shows that this will happen. The question is what happens to people who live in the years between jobs that no longer exist and jobs that don’t yet have a name.
