Knowledge Workers Boost Performance by Making AI Disagree With Them, Study Finds

Knowledge Workers Face Cognitive Offloading Crisis with Generative AI

Knowledge workers including scientists, teachers, lawyers and business leaders increasingly rely on generative AI for tasks demanding imagination, creativity and problem-solving. However, recent research exposes a troubling trend: professionals who interact with generative AI for work tasks often stop thinking critically1

1

. They surrender to AI's quick and confident answers, leading to speedy but low-quality solutions. Scholars identify this as cognitive offloading, a problem that erodes the very skills making knowledge workers valuable2

2

.

UMass Amherst Study Reveals Power of AI Friction

Research from Monideepa Tarafdar at UMass Amherst presents a different perspective on human-AI collaboration. In one study, Tarafdar and colleagues interviewed 45 knowledge workers at a U.S. public university, all using generative AI applications in core tasks: generating new ideas, solving complex problems and designing novel solutions1

1

. The team analyzed prompts and outputs from iterative AI chats. The findings demonstrated that when knowledge workers intentionally use AI to challenge their ideas—to generate friction—they significantly improve their performance. This approach of asking AI to disagree helps professionals avoid the echo chamber effect where AI models tailor output based on previous prompts and responses3

3

.

Real-World Applications Show Friction-Generating Queries Work

Practical examples from the study illustrate how arguing with AI transforms work quality. A marketing professional asked AI to create virtual customers with personas that would lead them to dislike a product under development—an accounting certificate program. The contrary feedback revealed customer preferences the accounting professional designing the program had not considered, such as using multiple cases from different industries to illustrate principles rather than cases from the same industry1

1

. An attorney used AI to find obscure legal loopholes and consider how companies might exploit them, exposing surprising but realistic scenarios of hard-to-detect unethical behavior in employees2

2

. A supply chain scientist adapted unfamiliar computing programs from signal processing and wireless communication suggested by AI, creating a stronger program despite the programs being new to him1

1

.

Critical Engagement with AI Enhances Professional Expertise

Using AI to find information that contradicts your own reveals unlikely perspectives disrupting familiar logic and sparking new insights. This matters because AI models quickly become familiar with how users prompt the model and interpret output, presenting information according to established patterns2

2

. Knowledge workers should lean on their expertise to question surprising information an AI system provides. A research scholar probed an AI system for journal papers using a concept in contrasting settings, helping him understand the concept from different angles. A lawyer followed up on legal citations by asking for different types of illustrative cases, helping her fully understand the original case and write a strong legal brief1

1

. Research shows creative writers produce better copy when they use AI as a sounding board rather than to ghostwrite, reinforcing that AI for human creativity and problem-solving works best through critical engagement with AI2

2

.

Organizations Must Support Friction-Based AI Strategies

Leaders should encourage knowledge workers to develop friction-generating queries of AI models. A straightforward approach involves using prompts that challenge rather than agree with a person's thinking3

3

. People can draw on their knowledge, work experiences and professional ethics to scrutinize AI outputs, sometimes realizing a response lacks originality even if factually correct. Organizations should create programs for knowledge workers to share experiences and insights through regular sessions led by cross-functional AI task forces or AI champions1

1

. Providing resources remains critical—internal AI boot camps and workshops led by AI experts help workers understand what AI tools can and cannot do. A professor leading an initiative about AI use in teaching developed internal competitions with prizes for professors who successfully completed AI boot camps. Organizations can offer AI sandboxes—safe, private environments where employees experiment and learn how to use AI tools2

2

. Watch for more research on opposing viewpoints and innovation in professional workflows as this approach to improve work quality with AI gains traction.