Key Takeaways
- Anthropic’s latest analysis finds no systematic rise in unemployment for workers most exposed to AI since late 2022, contrary to early alarmist forecasts.
- AI deployment remains far below its theoretical potential; Claude, for example, handles only ~33 % of computer‑and‑math tasks despite being capable of near‑100 % automation.
- Labor‑productivity growth in the AI era has been slower than during the 1990s IT boom, suggesting the technology’s economic impact is still nascent.
- Even AI‑industry leaders such as Sam Altman and Dario Amodei now express doubts about a imminent “jobs apocalypse,” while acknowledging AI’s continuing rapid improvement.
- Broader skepticism is emerging: public opposition to AI data‑centers is high, energy demands are projected to soar, and economists warn that AI may not deliver promised gains at an acceptable societal cost.
Anthropic’s Tempers Early AI‑Job‑Loss Predictions
In March, Anthropic released a study assessing whether intelligent machines were poised to eradicate human labor. The report bluntly stated, “We find no systematic increase in unemployment for highly exposed workers since late 2022,” challenging the narrative that AI would soon cause mass joblessness. This finding. The authors noted that AI deployment “remains a fraction of what’s feasible,” with the Claude chatbot covering only about a third of the tasks it could theoretically automate in the computer‑and‑math category.
Founders’ Shifting Forecasts
Dario Amodei, Anthropic’s co‑founder, had previously warned in May 2023 that AI could “wipe out half of all entry‑level jobs in one to five years.” By January 2024 he suggested AI might become a “general labor substitute for humans,” and in June he warned of a scenario where the economic trade‑off dial sticks on “hypergrowth, hyper‑inequality.” Yet the firm’s own data now show a more modest impact, prompting a recalibration of those dire projections.
Productivity Gains Lag Behind Expectations
Despite massive spending on data‑centers, labor productivity has not experienced the “galloping gains” that technologists once predicted. The article observes that “Labor productivity was, in fact, slower in the first three years of our AI era than during the information technology boom that began in the mid‑1990s.” This slowdown mirrors the historic pattern where transformative technologies take years to register in macro‑economic statistics—a point echoed by Nobel laureate Robert Solow’s quip that the computer age was visible everywhere except in productivity numbers.
Industry Voices Temper the Apocalypse Narrative
Even AI’s most public proponents are sounding more cautious. Sam Altman told reporters in May, “I don’t think we’re going to have the kind of jobs apocalypse that some of the companies in our space advocate or talk about.” MIT economist David Autor added that “A lot of people have noticed that the world is not changing as fast as they predicted,” reflecting a growing consensus that AI’s labor‑market disruption may be gradual rather than catastrophic.
Market Reaction Signals Growing Skepticism
The tech‑heavy Nasdaq index, which had risen largely on AI‑related stocks, has fallen about 8 % since its peak in early June, indicating investor unease. This market dip coincides with a broader shift in public discourse toward a “less cataclysmic narrative” that emphasizes the historical complexity of automation and questions the feasibility of a sweeping AI‑driven transformation of the economy.
The O‑Ring Analogy Highlights Limits of Substitution
One line of critique draws on the O‑ring argument, referencing the 1986 Challenger disaster where a faulty rubber O‑ring caused catastrophic failure despite the spacecraft’s otherwise advanced systems. Applied to AI, the analogy suggests that as long as AI cannot perform every task perfectly, it will increase the value of the remaining tasks. Depending on which functions AI assumes, it may either elevate high‑skill workers by offloading routine duties or create new opportunities for lower‑skill workers by handling expert‑level components.
Research Is Still Early, But Adoption Is Accelerating
Jed Kolko cautions that “research on the labor market impact of artificial intelligence is still in its infancy,” noting that Amodei’s one‑to‑five‑year window still has roughly four years left. Nonetheless, the Federal Reserve reports that AI adoption is expanding fast across businesses, and Autor observes that “AI is getting better… its progress shows no sign that it will soon hit a ceiling.” This tension—between nascent evidence of limited impact and rapid technological advancement—fuels ongoing debate.
Insiders Remain Bullish, Yet Economic Risks Loom
Despite skeptical outward statements, many industry insiders retain strong faith in AI’s future. Daron Acemoglu, Nobel‑prize‑winning economist, said “Insiders are as gung ho as ever… They still believe artificial general intelligence is around the corner.” Elon Musk continues to champion the vision that “AI+Robots will be able to do everything, resulting in universal high income. Work will be optional.” However, Acemoglu also warned that “Companies developing AI models are never going to make money… they are losing hundreds of billions of dollars every year,” highlighting the fragile economics underpinning the AI boom.
Societal and Energetic Pushback Complicates the Outlook
Public sentiment has turned against AI infrastructure: seven in 10 Americans oppose building AI data‑centers in their area, driven not only by fears of job loss but also by the insatiable demand for energy that pushes up local electricity costs. The International Energy Agency projects that power demand from data‑centers will more than double by 2030 to about 945 TWh, surpassing Japan’s total consumption. Such energy intensity, coupled with rapid model obsolescence, raises questions about whether society is willing to bear the environmental and financial costs of AI’s promised benefits.
Conclusion: A Promising Yet Uncertain Future
While AI continues to improve and attract massive investment, the evidence to date shows limited immediate labor‑market disruption, slower‑than‑expected productivity gains, and growing socioeconomic and ecological concerns. As Autor summarized, “Not everything is a computational problem,” reminding us that AI’s strengths in language replication do not equate to mastery of the messy, contextual realities of many jobs. Whether AI will ultimately fulfill its epochal promises—or fall short of delivering benefits at a price humanity accepts—remains an open, critically important question.
https://www.theguardian.com/technology/2026/jul/25/ai-jobs-apocalypse-human-labor