Is AI Making Us Stupid? Exploring Cognitive Offloading and Brainpower

In this episode of Science Weekly, Ian Sample and Madeleine Finlay explore whether AI is changing how our brains work. They discuss an MIT study where students wrote essays with a chatbot, a search engine, or their own effort, finding reduced brain connectivity and memory recall in technology-assisted groups. The discussion expands to cognitive offloading, memory, and the idea that external tools can free brainpower for other tasks. The takeaway is nuanced: offload when it helps, but preserve opportunities for deep learning and desirable difficulty to build lasting knowledge.

Overview: AI and the Brain

Is AI making us stupid? The episode centers on a pressing question about cognitive skills in the age of AI, exploring whether we rely too much on technology for thinking tasks. The discussion anchors on an MIT study in which 54 students were split into three groups to write essays. One group used a large language model, another used a search engine, and the third relied on their own thinking. The researchers employed EEG to gauge brain activity and conducted linguistic analyses and interviews after writing. The findings suggested that technology-assisted writing correlated with lower brain connectivity and, in the chatbot group, weaker memory for the details of their essays. The researchers term this a cognitive debt, though some scholars remain cautious about the conclusion. "The human brain is opportunistic. It will use whatever opportunities it can find to simplify tasks, and that's actually potentially a very good thing because it means that we can find the most efficient way to solve the problem rather than using these very limited capacity systems in the brain when they're not necessary to be used" — Sam Gilbert, Professor of Cognitive Neuroscience, University College London.

“The human brain is opportunistic. It will use whatever opportunities it can find to simplify tasks, and that's actually potentially a very good thing because it means that we can find the most efficient way to solve the problem rather than using these very limited capacity systems in the brain when they're not necessary to be used.” — Sam Gilbert, Professor of Cognitive Neuroscience, University College London.