AI Education Study Reveals Students Increasingly Value Human Intelligence Over AI
A longitudinal study examining classroom poll results from 2020 to 2026 in undergraduate and MSc programs centered on AI within computer science indicates a notable change in students' perceptions of human intelligence (HI) compared to artificial intelligence (AI). The analysis, which involved 471 students enrolled in both technical courses (like Machine Learning and Deep Graph Learning) and design-focused courses (such as Design Thinking for AI), outlines four distinct phases: hype, distrust, trust, and dependency. Initially, in 2020, responses leaned slightly towards AI, but from 2024 onwards, a clear trend favoring HI became evident across all MSc groups. By 2026, HI preference soared to 65 percent in a technical course (up 12 percentage points from 2025) and 90 percent in a design course, indicating that ongoing interaction with AI tools may prompt students to reevaluate the distinct advantages of human intelligence.
Key facts
- Longitudinal analysis of classroom poll responses from 2020 to 2026
- Data from 471 students in AI-focused undergraduate and MSc courses in computer science
- Courses include Machine Learning, Deep Graph Learning, and Design Thinking for AI
- Four recurring phases identified: hype, distrust, trust, and dependency
- Early responses in 2020 slightly favored AI
- Consistent shift toward human intelligence (HI) emerged from 2024 onward
- By 2026, preference for HI reached 65% in a technical course
- By 2026, preference for HI reached 90% in a design-oriented course
Entities
—