A study found that an aging workforce hinders productivity not by a lack of wisdom, but because older workers, often in leadership, slow the adoption of new technologies for the entire organization. This "albatross theory" challenges conventional narratives about experience.
History shows businesses often invest in new technology during downturns. A future recession could trigger a wave of AI implementation as firms restructure to cut costs, potentially accelerating automation and prolonging the negative employment shock more than in past cycles.
An effective method for refining AI output is to instruct the model to adopt an expert persona, such as a "PhD economist," and critically evaluate its own work. This often leads the model to self-identify and correct its own flaws without further prompting.
Anthropic's research shows that experienced AI users get more value because they learn to interact with the model as a collaborator. Proficiency is not just prompt engineering, but a learned skill of engaging the AI in a more sophisticated, iterative partnership to explore ideas.
Anthropic's economics team uses AI not just to accelerate existing work, but to expand capabilities into new areas like building interactive data visualizations. This "broadening of scope" is a key productivity driver, allowing experts to perform tasks they weren't trained for.
Even if AI saves time on tasks like curriculum planning, a teacher's overall productivity is constrained by the need to be in a classroom. This illustrates how job-level productivity gains can be limited by non-automatable "bottlenecks," potentially reducing AI's aggregate economic impact.
Jerome Powell's decision to stay on the Board of Governors indefinitely after his chairmanship ends is a direct response to perceived threats to the Fed's independence. This move is considered a "massive tell" of his deep concern about political pressure on the central bank.
Beyond productivity gains, AI's most transformative impact may be automating R&D to accelerate scientific discovery. This could lead to breakthroughs in health and wellness, solving problems that might otherwise take decades and fundamentally improving quality of life, not just GDP.
While not yet visible in aggregate unemployment, Anthropic's research found a suggestive signal: hiring for younger workers in jobs with high AI exposure seems to have slowed over the past year. This may be an early indicator of AI-driven shifts in the labor market.
Peter McCrory's journey highlights a modern career path for economists: starting with humanities, gaining quantitative skills at the Fed, pursuing a PhD, then moving through corporate finance (J.P. Morgan) and tech (LinkedIn) before a leadership role at a top AI company.
Anthropic's usage data shows that late-adopting regions in the U.S. are catching up to early adopters at a rate 5 to 10 times faster than for technologies like the internet. This accelerated diffusion implies that AI's economic effects could materialize much more quickly than historical precedents suggest.
By analyzing time savings across tasks on its platform, Anthropic calculates a potential 1.8 percentage point annual lift to labor productivity. This bottom-up, data-driven estimate is more than double the typical economist's forecast of ~0.8%, which often relies on historical analogs.