Neuroscientists Raise Alarm Over the Risk of a Future Dementia Epidemic

Could the habit of asking ChatGPT about everything under the sun one day backfire on our brains? Scientists are asking this question more and more often amid the rapid spread of artificial intelligence. The trigger for the new wave of discussion was an interview with neuroscientist and cognitive science specialist Vivienne Ming, who warned that excessive delegation of mental work to neural networks could weaken one of the brain’s main defense mechanisms against dementia.

What exactly is worrying scientists

According to Ming, the problem is not that AI directly causes Alzheimer’s disease or other forms of dementia. The danger lies in the so-called “chronic substitution of thinking” — a situation in which a person solves tasks independently less and less often and increasingly shifts them onto algorithms.

The brain works on the principle of “use it or lose it.” If certain cognitive skills are not engaged for a long time, the corresponding neural networks gradually weaken. This especially applies to memory, concentration, critical thinking and spatial navigation.

Scientists call the protective mechanism that helps the brain resist age-related changes for longer cognitive reserve. Research shows that people with a high cognitive reserve have about a 35% lower risk of developing dementia compared with those whose reserve is lower. This reserve is built over the years through learning, intellectual work, reading, problem-solving and constant mental activity.

What the MIT study showed

One of the most widely discussed studies in recent years was the work of specialists from the MIT Media Lab titled “Your Brain on ChatGPT.” The experiment involved 54 people divided into three groups: some wrote essays on their own, others used search engines, and the third group used ChatGPT. During the work, brain activity was measured using electroencephalography.

The results were unexpected:

— ChatGPT users showed the weakest neural activity;

— their brain connections were less pronounced compared with the other participants;

— about 83% of participants were unable to reproduce a single phrase from the text they had just written;

— many noted a weaker sense of authorship over their own work.

The study’s authors called this effect “cognitive debt” — a person gets a fast result but pays for it with a reduced involvement of the brain in the learning and memorization process.

But there is an important caveat

Despite the dramatic headlines, scientists stress that there is still no evidence that using ChatGPT causes dementia. The MIT study was a preliminary scientific paper (a preprint), and a number of specialists later pointed out methodological limitations and the small sample size.

Therefore, claims that “AI causes Alzheimer’s disease” are still an exaggeration. However, experts agree on another point: if a person systematically stops training memory, analyzing information and independently solving intellectual tasks, this could potentially weaken cognitive abilities in the long term.

The biggest problem is the younger generation

Particular concern is caused by the fact that today’s schoolchildren and students begin using AI from an early age. If previous generations first learned to write, count and analyze independently and only then gained access to digital assistants, many teenagers today already see neural networks as their primary tool for solving tasks.

If such a habit becomes entrenched for decades, the consequences may only appear in 30–40 years. That is why neuroscientists are urging that the impact of AI on the brain be studied now, while the technology is only beginning to change people’s everyday habits.

Conclusion

For now, artificial intelligence remains more of a tool than a threat. But researchers warn: if you use it as an assistant, it enhances human capabilities. If, however, you completely delegate thinking, memory and decision-making to it, the brain may gradually lose skills that took millions of years of evolution to form. And that is precisely what is causing growing concern among neuroscientists today.

So the main question is no longer how intelligent AI will become, but how intelligent we ourselves will remain.