The Anthropic Economic Index report: Economic Primitives – what HR leaders need to know

“AI isn’t a substitute for specific human jobs but rather a general labor substitute for humans,” Anthropic CEO Dario Amodei recently wrote in a 15,000-word essay entitled ‘The Adolescence of Technology.’

For HR leaders reading this statement, a clear message is painted: AI isn’t replacing tasks or professions, but it is rapidly reshaping skills and roles.

AI safety and research company Anthropic recently released The Anthropic Economic Index report: Economic Primitives, providing an exclusive insight into how productivity gains from AI are unevenly distributed.

UNLEASH took a deep dive into the report to find the key takeaways for HR leaders.

What type of work is AI actually being used for?

Gen AI is typically being used for “more complex tasks,” according to Anthropic’s data, therefore allowing it to “yield a greater time savings” from AI.

HR leaders must therefore be aware that AI will not impact all roles equally. 

“We derive this by having Claude estimate both how long a task would take a human working alone and the duration when human and AI work together, which we validated in previous work,” the report stated. 

This “speedup” is calculated by “human-alone” time being divided by “human-with-AI time” – for example, reducing a one hour task to 10 minutes creates a 6x speedup time. Therefore, the highest speed-ups come when there’s the highest human capital.

For HR leaders, this highlights the type of work that AI is and isn’t being used for, exposing any areas which should be focused on for upskilling.

Training, talent pipelines and performance expectations can then be adapted by HR leaders to support organizational strategy.

HR leaders must therefore understand that workforce strategy must shift from broad AI awareness to targeted capability building. 

For example, the report suggests that roles reliant on judgment, analysis, and domain expertise will gain the most from AI augmentation. 

Likewise, lower‑complexity tasks may require different interventions, such as workflow redesign or clearer guardrails, to ensure AI is used reliably and safely.

Further supporting this, conversations with Claude.ai that required a high school education demonstrated a speedup of 9x, while those requiring a college degree attain a 12x speedup. 

This implies that productivity gains are more pronounced for use cases requiring higher human capital, consistent with evidence that white collar workers are far more likely to adopt AI,” the report added.

Yet more complex tasks such as those requiring less than a high school education – asking basic questions about products, for example – generated a 70% success rate, compared to college level conversations – asking for analysis plans, for example – which generated a 4% drop in success rate.

Overall, Anthropic’s findings highlight that AI isn’t simply accelerating work, but rather redefining which capabilities matter most. 

For HR leaders, the pattern that emerges is one where AI amplifies human expertise rather than replaces it, making the underlying skill level of the worker a critical variable in how much value AI can unlock.

Assessing readiness with the five Economic Primitives

Additionally, the report introduces a framework to help HR leaders better understand how AI is used in the workforce, through the five economic primitives.  

These span task complexity; human and AI skills; work, coursework or personal use case; the AI’s level of autonomy; and task success, providing leaders with a clearer insight into where AI is delivering value and where human capability remains essential. 

These primitives are therefore used to further investigate implications for adoption and productivity in the workforce, with the report stating “Claude use diversifies with higher adoption and income” and that the AI tool “succeeds on most tasks, but less so on the most complex ones.”

This highlights that “early adopters in less developed countries” are usually more “technical” users that have “specific, high-value applications use Claude for education,” in comparison to mature markets, which sees usage “diversify toward casual and personal purposes”.

What’s more, the report highlights that the education level of Claude’s responses “tends to match the user’s input,” meaning as the time it takes for a human to complete a task increases, Claude’s success rate falls. 

If HR leaders view AI adoption through these primitives, they will be better equipped to assess where AI can improve productivity, compared to where human capability will be seen as the main limiting factor. 

In turn, this will help organizations avoid overestimating AI’s readiness for complex work. 

Anthropic’s report also found that “widespread adoption of AI” has the potential to “increase US labor productivity growth by 1.8 percentage points annually over the next decade.”

This is based on task speedups alone, but once leaders “account for task reliability,” the estimate falls to approximately 1.2 percentage points per year for Claude.ai usage, and to 1.0 percentage points for API traffic.

For HR leaders planning the continued integration of AI strategies, this is particularly important as it shows that the realities of the Economic Primitives are much more incremental. 

What’s more, HR leaders must ensure they align their expectations by investing in AI integration and workforce training. In turn, this should create benchmarks which are realistic and tied to measurable outcomes.

The message for HR leaders is clear, as Anthropic’s report reinforces that AI’s impact on the workforce will be significant but uneven – with the organizations benefiting the most emerging as those that pair AI investment with deliberate workforce development.