This post was originally published on this site.
Anthropic has released data on how workers use AI in different occupations. The group collected information from millions of prompts typed into its Claude chatbot. The goal of this was to start understanding exactly how individuals apply AI in daily tasks.
Anthropic introduced what it calls the Anthropic Economic Index to track the share of tasks in each role that use AI assistance. The project draws from real user conversations, though personal data is concealed. This method yields insights on tasks that depend on language-based support, such as coding, writing, and planning.
People can track the Anthropic Economic Index to watch how these trends evolve over time. It ranks tasks and shows where AI-based input might gain traction. Teams that gather these insights hope to guide discussions on how AI tools fit into daily routines.
Which Fields See The Highest Adoption?
Software development comes out on top, with 37% of all prompts drawn from occupations that handle programming duties. Roles in media, design, and creative work stand second at around 10 percent, while education and library roles follow close behind. This pattern signals an emphasis on text-heavy tasks, such as content drafting and research.
Software engineers often delegate code reviews, debugging, or system checks to AI chatbots. Media and design specialists turn to the tool for script drafting or concept brainstorming. Meanwhile, educators might request lesson plans or additional explanations for academic topics.
These high-usage groups reveal an appetite for text-based guidance. Higher-wage positions in software and analysis tasks see major interest in automation or collaboration with AI. Yet some industries stay less involved, as we see next.
Engineers are not alone in turning to AI. Writers turn to chatbots for drafting press releases or narratives, and marketing specialists might request sample campaigns or brand messaging. This method saves time and allows skilled professionals to concentrate on higher-level decisions.
More from News
Are Some Professions Unconvinced?
Transportation, healthcare support, and farming are at the lower end of AI adoption. Their tasks often involve hands-on duties that don’t mesh with chat-based tools. Even at the high end of the salary range, specialists such as surgeons avoid digital assistants for tasks that require direct clinical judgment.
Roles that demand physical interaction or extensive licensing see few uses for text-based help. This difference lines up with Anthropic’s finding that only four percent of occupations apply AI in at least three quarters of their tasks. The majority adopt partial assistance rather than a total takeover.
At the same time, only about 36 percent of occupations apply AI for at least a quarter of tasks. That leaves many workers who rarely involve automated support. Researchers interpret these figures as a sign that the transition toward advanced technology is uneven.
Deep medical expertise and complex machinery supervision do not blend well with a chatbot’s text output. Farming requires fieldwork, and fishers seldom benefit from automated responses. Employers in such sectors often place more weight on tangible tasks that do not involve text creation.
What Are The Main Observations?
Anthropic reports that 57% of AI usage boosts human work, such as writing drafts or clarifying concepts. The other 43% replaces manual effort, for instance code generation that happens without much human revision.
Also, data shows that mid to high wage roles employ AI the most, while both ends of the pay scale show less interest. Researchers cite software-focused tasks as the top driver of frequent queries. Meanwhile, teachers and marketing workers also account for a noticeable chunk of requests.
A separate analysis in Europe explores benefits for women, especially in positions where technology complements their skill sets. Higher education and strong labour participation help cushion any negative effects when AI is adopted. That study points to greater involvement of women in areas that use these tools.
Anthropic says that these early patterns could change as language models evolve further. Many workers who have not embraced AI may do so once they see value in partial task assistance.
Some voices have asked how wages and labour conditions might adapt, but the data so far points to gradual incorporation in certain fields. This is a growing area of study, with plenty of debate on best practices for the workplace. Adoption patterns may change even more as new uses arise.