0 Comments
Fresh research from the Massachusetts Institute of Technology Media Lab suggests that frequent use of ChatGPT to write essays might be linked to cognitive decline.
It turns out that using ChatGPT every day to write essays and perform literary tasks may be having a negative impact on our brains.
New research from the Massachusetts Institute of Technology Media Lab involved measuring the electrical activity of three different groups of people.
All three were asked to write multiple SAT essays. The first group heavily used ChatGPT, the second were allowed access to traditional search engines, and the final group had no external resources available. There were 54 participants in total and all ranged from 18-34.
Each of their brains were monitored using electroencephalography over a four-month period – try saying that three times fast. Activity was measured across 32 different regions of the brain.
According to the study’s findings, those who had used ChatGPT were discovered to have the ‘weakest’ brain connectivity and struggled the most to remember details about the work they had written.
The experts believe this effect could be heightened by frequent use of AI over time.
The study says that ChatGPT users ‘consistently underperformed at neural, linguistic, and behavioural levels,’ adding that, ‘these results raise concerns about the long-term educational implications and underscore the need for deeper inquiry into AI’s role in learning.’
The group who wrote essays without any other resources were found to have the ‘strongest, most distributed networks,’ which is hardly surprising given the implications of AI use throughout the study.
Lead author Nataliya Kosmyna spoke to Time Magazine, noting that individuals who rely on ChatGPT are not ‘integrat[ing] any of it into their memory networks.’ There is also concern around how these impacts could affect children, as their brains are still developing and may be more at risk.
‘I am afraid that in 6-8 months, there will be some policymaker who decides to [start a] GPT kindergarten,’ says Kosmyna. ‘I think that would be absolutely bad and detrimental.’
It’s worth noting that as of yet, the paper has not been peer reviewed and its sample size is small. The findings were published early in order to spread awareness about the potential damage that ChatGPT might be having on educational programmes and courses.
