[3/3] Future of Work with AI Agents: Auditing Automation and Augmentation Potential across the U.S. Workforce (AI column)
Finishing the story (1 and 2) About the study, I will share my final thoughts.
The study also revealed an important trend: as AI agents are introduced, the set of key skills demanded by humans changes. The man of the future will need less narrow analytical skills, but more social and managerial competencies. After all, the analytical work will be taken over by AI.
Separately, the authors discussed the limitations of their study: The authors emphasize that their results are a “slice” of the beginning. 2025 It is impossible to project them directly into the distant future without taking into account the development of technologies. In particular, the WORKBank database covers existing O*NET tasks; but in the future there will be new types of tasks and professions as AI is integrated, which are simply not taken into account now. Respondents may not fully represent the capabilities of the latest AI systems (The authors note that many people do not yet know much about the limits and potential of AI.)which may have influenced their responses
- The survey covered 104 Computerized professions are a wide range, but not all economics professions; for example, working specialties where AI agents are not yet used are excluded. The authors predict that the landscape of preferences and opportunities will shift: according to them, “AI Agent WORKBank reflects the current state of generative AI at the beginning of the project.” 2025 As technology evolves, the picture of the tasks that agents want will change. Future iterations of this audit are necessary to monitor long-term trends.”
The researchers also make several predictions and recommendations.
- They expect a decrease in demand for purely analytical, monotonous information skills. (where AI has already proven effective) Increase the role of skills that require interaction between people. This prediction is backed up by their finding of a shift in high-paying skills and actually paints an image of the labor market, where the value will be social intelligence, the ability to manage and train, not just the ability to work with information.
- The authors call for active measures to retrain workers. Since AI agents are able to change the structure of demand for skills, it is important to develop programs of reskilling and upskilling - retraining employees under new competencies.
- The study points to the need for a more human-centered AI adoption strategy. Organizations and developers are encouraged to consider employee desires when deciding what to automate – that is, the AI systems of the future should be developed in tandem with users so that they are trusted, accepted by employees and improved efficiency without negative social consequences.
- Brynjolfsson emphasizes that special attention should be paid to the areas of “R&D Opportunity” – those tasks that people want to automate, but the technology does not yet know how. Increasing the intensity of research in these areas will create AI tools that bring maximum benefit and demand in the workplace, thereby better preparing the economy for the future.
In general, the authors see the future of work not as complete robotization, but as the evolution of roles where humans and intelligent agents work together, and call for managing this evolution responsibly based on human needs.
P.S. The authors’ conclusions about collaboration resemble ideas from the whitepaper.Rethinking Software Engineering in the Foundation Model Era: From Task-Driven AI Copilots to Goal-Driven AI Pair ProgrammersThe one I mean. told earlier.
#AI #Economics #ML #Work