 |
| Photo by Solen Feyissa on Unsplash |
The way many of us think is shifting—and not in a dramatic, sudden way. It’s subtle. When AI tools handle tasks that once required effort, our brains don’t always follow the same deep processing they used to. Early research suggests this shift could change how we learn, remember, and generate ideas. The effects aren’t inherently negative. But they are meaningful.
The Shift From Thinking to Simply Getting Answers
Recent work from researchers at MIT’s Media Lab explored what happens in the brain when people write using different tools. In a small, preliminary study of 54 adults, participants wrote timed essays under three conditions: using only their brain, using a search engine, or using ChatGPT. Brain activity was measured with EEG.
The pattern was clear—even with the study’s early and non–peer-reviewed status. The group that used ChatGPT showed the lowest brain engagement. As TIME’s reporting on the study detailed, their essays became increasingly uniform and less original over time. Some participants eventually relied on the chatbot to generate nearly the entire essay.
Meanwhile, the group who wrote without any tools showed the highest neural connectivity, especially in frequency ranges associated with memory and meaning-making. They also reported a stronger sense of ownership and satisfaction over their work.
These results are still emerging and require broader confirmation, but they point to an important dynamic: when AI replaces the thinking process, the brain steps back.
When Convenience Becomes “Cognitive Offloading”
Healthline highlighted a related concept called cognitive offloading—a tendency to hand over mental work such as recalling, analyzing, or problem-solving to technology. The issue isn’t that AI performs tasks poorly; it’s that we stop practicing the skills behind those tasks.
A research review from 2024 summarized potential outcomes when complex thinking becomes automated:
- Reduced mental engagement
- Declines in memory retention
- Shortened attention spans
- Weaker problem-solving flexibility
- Lower confidence in one’s own ideas
This isn't about intelligence. It’s about mental habit formation. What we do repeatedly becomes how our brain defaults to thinking—or not thinking.
A Reveal Hidden in the Study’s Final Phase
A particularly revealing moment came during a final reassignment phase described in both the CNN interview and TIME’s coverage. Participants switched conditions:
- Those who had relied on AI now had to write without it.
- Those who had written unaided were allowed to use AI.
The outcomes diverged:
The previously AI-reliant group recalled less and showed weaker neural connectivity.
The previously unaided group actually showed an increase in neural engagement once AI was introduced.
This suggests the issue is not AI itself but the timing of when we use it.
When the brain engages first, AI can enhance.
When AI engages first, the brain tends to follow passively.
What This Means for Real-Life Learning and Work
Other research from MIT, referenced in reporting by TIME, has found that people who spend significant time talking with AI can feel more isolated over time. Meanwhile, a 2025 analysis cited by Healthline described a shift from active critical thinking to passive verification when AI becomes the primary source of ideas.
The concern isn’t about dependence in a dramatic sense. It’s about losing the early struggle phase of thinking—the part where memory builds, meaning forms, and creativity sparks.
What I’ve Learned Along the Way
The pattern emerging across these studies is not that AI harms the brain—but that the brain requires active practice to grow and maintain complex thinking skills. The MIT team emphasized that neural connectivity strengthened when participants engaged directly with a task before turning to AI. When participants handed the thinking over immediately, that internal processing weakened.
A practical principle from the research is simple:
Think first. Then ask AI.
Try outlining the idea yourself before requesting elaboration.
Try forming your own interpretation before asking for clarification.
Try solving the problem before checking it.
This keeps AI in the role of a partner rather than a substitute.
AI doesn’t erase our ability to think.
It shapes the habits that determine whether we do.