Anthropic: AI Widens the Skills Gap

The Unequal Impact of AI on the Labor Market

A recent study conducted by Anthropic highlights that artificial intelligence, while rapidly transforming the way work is done, has not yet led to significant job losses. According to Peter McCrory, head of the economy at Anthropic, the labor market remains "still healthy," but signs of unequal impacts are beginning to emerge, particularly for young workers entering the job market.

In an interview at the Axios AI Summit in Washington, D.C., McCrory explained that the company's latest report on economic impact does not show clear evidence of large-scale job displacement so far. There is no notable difference in unemployment rates between workers using Claude to automate core tasks of their jobs, such as technical writers, data entry clerks, and software engineers, and those whose jobs require physical interaction and dexterity.

Alarming Forecasts for the Future

However, the growing adoption of AI across industries could quickly change this situation. Dario Amodei, CEO of Anthropic, anticipates that AI could eliminate half of entry-level office jobs and drive unemployment up to 20% within the next five years. McCrory emphasizes the importance of establishing a monitoring framework to understand these changes before they occur, in order to respond appropriately.

A Growing Skills Gap

Anthropic's fifth report on economic impact, published on Tuesday, also reveals a growing skills gap between early users of Claude and newcomers. Early adopters are more likely to derive greater value from the model, using it for professional tasks rather than casual or one-off uses. They leverage Claude in a more sophisticated manner, acting as a "thinking partner" for iteration and feedback.

McCrory noted that AI is becoming a technology that rewards those who already know how to use it. Workers who can effectively integrate AI into their work will increasingly have an advantage. Roles involving tasks that AI excels at, which are already being automated and are linked to concrete use cases in the workplace, are the areas most likely to signal where displacement could emerge.

Geographically Uneven Adoption

The report also revealed that the use of Claude is more intense in high-income countries, particularly in the United States, in regions where more skilled workers are located. This usage is concentrated on a relatively narrow set of tasks and specialized professions. In other words, despite the promises of AI as an equalizer, its adoption may already be skewed in favor of the wealthier, amplifying these advantages as advanced users get ahead.