Signs of Human Brains Becoming Dull, Scientists Sound Alarm
The debate surrounding Artificial Intelligence (AI) has frequently centred on one primary fear: will AI steal human jobs?
That fear is valid. OpenAI CEO Sam Altman even warned world leaders at the United Nations Security Council that humanity could ‘lose control over the future’ if the pace of AI development accelerates without government regulation.
However, there is another threat that is far more silent, occurring more rapidly, and is currently eroding offices worldwide: humans becoming so dependent on technology that their brains become ‘dull’.
‘Agency Decay’: When Machines Think for Humans
Cornelia Walther, a researcher from the Wharton School, has coined a specific term for this phenomenon: ‘agency decay’, or the weakening of the ability to act independently.
She defines it as the gradual erosion of an individual’s ability and desire to observe closely, think independently, make conscious choices, and act responsibly.
Ironically, this occurs not because AI is inherently dangerous, but because the convenience it offers places human cognition into a ‘default’ lazy mode. ‘Tools that were meant to help us think are slowly starting to think for us,’ Walther wrote, as cited by TechTimes.
Several empirical studies prove this diagnosis is not mere speculation:
An MIT Media Lab study found that participants writing essays using ChatGPT showed significantly weaker neural connectivity compared to those writing without assistance. Surprisingly, 83% of them were unable to quote even a single sentence from the text they had just produced themselves.
A cross-campus trial involving over 1,200 participants from Carnegie Mellon, Oxford, MIT, and UCLA found that even brief AI assistance—as little as 10 minutes—was sufficient to reduce independent problem-solving abilities once the tool was removed.
A global BCG survey of 70 C-suite executives noted that half had already observed skill erosion within their respective companies, and over 60% believe this will become a real threat within the next 3 to 5 years.
The Trap of ‘Cognitive Offloading’
Many companies are racing to adopt AI. However, the greatest problem is often not access to technology, but how humans utilise it.
There is a widening gap between using AI to expand thinking capacity versus using it to replace the thinking process itself.
Imagine two executives. The first executive delegates analysis, argument construction, and decision-making entirely to AI, until the machine acts as the primary source of thought. Meanwhile, the second executive develops a framework independently first, then uses AI to challenge assumptions, seek alternatives, and test reasoning.
Only the second executive’s brain becomes sharper. The first, in the field of cognitive psychology, is experiencing ‘cognitive offloading’—an extreme delegation of cognitive load that causes the neural pathways for independent reasoning to undergo atrophy.
The Irony of Automation: Are Doctors Becoming Dull Too?
The impact of this ‘collective stupidity’ is beginning to encroach upon crucial sectors. An International AI Security report noted a 6% decrease in doctors’ ability to detect tumours without AI assistance—occurring in just three months since the AI system was introduced in hospitals.
The current industry-standard implementation model—where AI handles routine tasks while humans only handle exceptions—creates a deadly trap. As Mohammad Jarraver, an information scientist from the University of North Carolina, stated, companies may ultimately hand the most complex tasks to employees whose analytical skills are untrained, because the heavy lifting of thinking has been outsourced to AI.
Rising Wages, but Mental Quality at Risk
Ironically, the current labour market demands dual expertise. Data from PwC’s 2026 Global AI Jobs Barometer shows a wage premium of up to 62% for workers with AI skills. Entry-level positions involving AI now seven times more frequently require senior-level skills such as leadership and judgement.
This means that while the system forces humans to engage in high-level thinking, the ease of AI actually dulls the brain’s capacity to achieve it.
If companies do not immediately balance technology adoption with investment in honing human capabilities, the result will not be a super-intelligent workforce, but organisations rendered paralysed without the ‘scaffolding’ of machines. Ultimately, the greatest promise of AI is not a world where humans no longer need to think, but a world where humans should be capable of achieving far greater things—rather than becoming dull spectators in their own homes.