The Cognitive Cost of AI
This video discusses an MIT study investigating how using Large Language Models (LLMs) impacts human brain activity and cognitive performance. The research employed Electroencephalography (EEG) on subjects completing essay assignments, comparing those using AI, those using a search engine, and those using no technological aids to assess cognitive load and neural connectivity.
The study found that groups utilizing more technological support exhibited less brain activity and a lower degree of "ownership," or recall, of their work, with the AI users showing the least neural engagement and producing less varied essays.
Importantly, the research also suggested that initial unaided mental effort followed by AI use may lead to higher cognitive integration, but habitual LLM use from the start leads to diminished long-term cognitive skills.
This video discusses an MIT study investigating how using Large Language Models (LLMs) impacts human brain activity and cognitive performance. The research employed Electroencephalography (EEG) on subjects completing essay assignments, comparing those using AI, those using a search engine, and those using no technological aids to assess cognitive load and neural connectivity.
The study found that groups utilizing more technological support exhibited less brain activity and a lower degree of “ownership,” or recall, of their work, with the AI users showing the least neural engagement and producing less varied essays.
Importantly, the research also suggested that initial unaided mental effort followed by AI use may lead to higher cognitive integration, but habitual LLM use from the start leads to diminished long-term cognitive skills.
