Imagine you are competing for the same client as someone with an always-on assistant that drafts, summarizes, codes, researches, and revises in seconds. That image captures why the economics of artificial intelligence increasingly look less like an old contest over owning factories and more like a contest over using leverage well. Public sources support the idea that generative AI has lowered barriers to entry: the OECD says generative AI stands out for lower deployment costs and reduced technical prerequisites, while OpenAI offers a free tier and a Plus plan priced at $20 a month. That matters because access to capable tools is no longer confined to large firms. The practical question is shifting from who can buy technology to who can turn it into better output.
The people most likely to benefit are workers and firms that can integrate AI into daily tasks rather than treat it as a novelty. Evidence so far points to real, but uneven, gains. A large field study summarized by the National Bureau of Economic Research found that customer-support agents using generative AI became about 14 percent more productive, and the later Quarterly Journal of Economics version reported a roughly 15 percent gain. The OECD, reviewing experimental studies, says productivity gains in areas such as customer support, software development, and consulting have ranged from 5 percent to more than 25 percent. Those figures are meaningful, but they do not clearly confirm a universal “10x worker” effect across occupations. Public evidence supports acceleration, not a single magic multiplier for everyone.
In practical terms, this shift is most visible in knowledge work, service jobs, and digital occupations where language, analysis, and pattern recognition matter. It is most useful when work involves drafting text, handling routine queries, editing, summarizing, coding, or assisting with decisions. Like a bicycle for the mind, a tool is most powerful when the rider learns how to steer it. Public sources do not clearly confirm that AI has already erased the location premium of major cities. In fact, the OECD’s 2024 report on the geography of generative AI says urban labour markets are more exposed to generative AI than non-urban areas, and that cities are, and will be, significantly more exposed than rural areas. That suggests disruption will spread widely, but not necessarily in the simple direction of rural areas displacing cities.
The mechanism is straightforward. Generative AI reduces the time and effort needed for many tasks, letting workers complete more work or redirect attention toward higher-value judgment. That is where leverage enters the picture. A worker who learns prompting, verification, workflow design, and iterative improvement can increase output without hiring a larger team. Public evidence also suggests that this changes hiring signals. A 2025 study of about eleven million UK job postings found a 21 percent increase in demand for AI roles as a share of postings from 2018 to 2023, while mentions of university education requirements for AI roles fell by 15 percent. The same study found AI skills carried a 23 percent wage premium. That is not the same as credentials disappearing, but it does show output-linked skills gaining ground.
The clearest implication is that AI is sharpening differences in adaptability. The World Economic Forum reports that employers expect 39 percent of key job skills to change by 2030, with AI and big data, technological literacy, creative thinking, resilience, flexibility, agility, curiosity, and lifelong learning all rising in importance. That evidence supports a more cautious version of the leverage thesis: the divide is growing between people who can absorb new tools into their work and those who cannot. Public sources do not clearly confirm an inevitable winner-take-all future or a broad collapse of degrees across the whole economy. They do confirm a faster premium on usable skills. One sensible step today is to choose one recurring task at work and redesign it with AI, measuring time saved and quality gained rather than relying on slogans.