You may not take interest in AI, but AI will take interest in you. This statement has never been more relevant than today, when 70 million American workers are preparing for the biggest workplace transition in history thanks to AI agents. However, their voices are often overlooked in this discussion.
The Largest Study on AI Agents in the Workplace
Researchers conducted a nationwide study encompassing 1,500 workers from 104 different occupations to understand what workers actually want AI to automate or enhance. The results reveal three key insights that completely transform our understanding of the future of work.
Key Findings: When Reality Meets Expectations
1. Mismatch between worker desires and AI investments
Analysis showed that as many as 41% of startup companies (through the Y Combinator program) focus their AI solutions on tasks that fall into “Low Priority Zones” or “Automation Red Light Zones” – areas where workers don’t want automation or it’s not technically feasible.
2. Workers want partnership, not replacement
The biggest surprise: workers in 47 out of 104 analyzed occupations prefer “equal partnership” with AI agents, not complete automation. They want AI to be their colleague, not their replacement.
3. Skills shift: From information to interpersonal abilities
As AI takes over information processing, it becomes clear that the future of work will depend on interpersonal and organizational skills, not data analysis capabilities.
Where Do Workers Resist AI Automation?
Analysis of worker transcripts reveals three major concerns:
- Lack of trust (45% of cases)
- Fear of job replacement (23%)
- Absence of human touch (16.3%)
Particularly interesting is that in creative sectors (arts, design, media) only 17.1% of tasks receive positive ratings for automation.
What Do Workers Want to Automate?
Despite fears, 46.1% of workers express positive attitudes toward AI automation of certain tasks. Their motivations are clear:
- Freeing up time for higher-value work (69.4%)
- Reducing repetitiveness (46.6%)
- Improving quality (46.6%)
- Reducing stress (25.5%)
Human Agency Scale: A New Approach
Researchers introduced the revolutionary “Human Agency Scale” (HAS) with five levels:
- H1: No human involvement
- H2: Minimal human involvement
- H3: Equal partnership
- H4: Dominant human involvement
- H5: Human involvement is essential
This scale shifts focus from an “AI-first” approach to a “human-centered” approach, which is a crucial difference in AI system design philosophy.
Four Automation Zones
Based on worker desires and current AI capabilities, tasks can be divided into four zones:
1. Automation Green Light Zone: High desire + high capabilities Example: Automatic invoice processing in accounting
2. Automation Red Light Zone: Low desire + high capabilities Example: AI diagnostics in medicine without human oversight
3. R&D Opportunity Zone: High desire + low capabilities Example: AI assistants for creative writing
4. Low Priority Zone: Low desire + low capabilities Example: Complete automation of artistic processes
Future of Skills: The Great Shift
Analysis of 844 tasks reveals a dramatic change in valuable skills:
Skills in Decline
- Data analysis
- Knowledge updating
- Information processing
Skills on the Rise
- Interpersonal abilities
- Organizational skills
- Decision making
- Quality control
Practical Implications for Organizations
For Leaders: Instead of focusing on complete automation, companies should develop AI solutions that enable equal partnership with workers.
For Workers: Investing in interpersonal and organizational skills becomes crucial for future career success.
For Investors: Current AI investment trends don’t match actual worker needs – there’s a huge opportunity in the “green zone” and “R&D opportunity zone.”
Humans and AI as Partners
This study challenges the common narrative of AI as a replacement for human work. Instead, it reveals a more complex picture where workers want AI as a smart partner that will enable them to focus on what they do best – the human aspects of work.
The future of work won’t be about whether AI will replace humans, but about how AI and humans will work together to create more productive, fulfilling, and humane workplaces.
“It’s not a question of whether AI will change our jobs – the question is how we can shape that change to serve both people and technology.”
This blog post is based on research conducted in collaboration with economists from Stanford Digital Economy Lab, representing the largest study on AI agents in the workplace to date.
