Is AI Making Us Stupid? The Uncomfortable Truth from MIT
While institutions like MIT may prefer diplomatic language, the reality presented by their research is stark: Large Language Models (LLMs) like ChatGPT appear to be making a significant portion of users, particularly younger individuals, less intellectually capable. Words like “stupid” or “brain rot,” while inflammatory, arguably capture the concerning trend of intellectual surrender observed with widespread AI adoption.
The Grok Phenomenon: Outsourcing Your Thoughts
A quick look at platforms like Twitter reveals a disturbing pattern: thousands unironically ask AI models like Grok to explain complex topics – from politics and economics to current events. This isn’t seeking information; it’s outsourcing critical thinking and opinion formation. Instead of engaging in rational evaluation or personal research, users are actively surrendering their cognitive processes to an AI program. This level of intellectual passivity is a worrying sign of potential intellectual decline.
Measuring the Impact: Insights from MIT Research
Given the recent proliferation of accessible AI, compelling research on its neurological impact is crucial. A recent pre-print paper from MIT, released urgently due to the pressing nature of its findings, utilized EEG scans on students to study the effects of LLM use during essay writing compared to using search engines or no external tools.
The results were concerning. The group using only their “brain” showed significantly stronger semantic processing networks and enhanced connectivity between key brain regions associated with understanding and creativity. In contrast, the LLM group exhibited substantially fewer neural connections during the writing process.
Deficiencies in Neural Activity, Content, and Learning
Digging deeper, the study highlighted deficiencies across multiple levels for the LLM-assisted group:
- Neural: Significantly weaker alpha and theta brainwaves associated with memory and cognitive effort.
- Linguistic: Essays were statistically homogeneous, lacking original thought, and described by educators as “soulless.”
- Behavioral: Students became progressively lazier over months, often resorting to direct copy-pasting with minimal effort.
Furthermore, when the LLM group was asked to write without AI assistance, they struggled significantly, demonstrating that reliance on AI bypassed their own memory and hindered their creative process. This indicates a measurable, negative impact on learning and critical thinking.
The Educational Crisis: Rising Illiteracy and AI Dependence
This research aligns with broader trends. Pew Research data shows a doubling in US teenagers using ChatGPT for schoolwork. Educators across various fields are observing what one professor termed a generation becoming “essentially illiterate,” lacking both basic literacy and cultural/historical knowledge due to AI dependence.
Adding to this concern is potential government action. Proposed legislation like the “Big Beautiful Bill” includes provisions that could impose a 10-year moratorium on state-level regulation of AI models in interstate commerce. This push for deregulation comes at a time when we are only beginning to understand the negative impacts of these tools on developing minds.
The Alarming Prospect of “GPT Kindergarten”
The lead researcher of the MIT study, Nataliya Kosmyna, explicitly stated her motivation for releasing the pre-print early was the fear of seeing “GPT kindergarten” within months – a prospect she views as “absolutely bad and detrimental” to developing brains, which are “at the highest risk.”
While MIT’s desire to avoid provocative language is understandable, the urgency of the research combined with moves towards deregulation paints a picture of an existential threat to human intelligence. Sometimes, blunt language is necessary to provoke awareness and action.
Conclusion: The Real Cost of AI in Education
The unchecked integration of LLMs, particularly in education, poses a severe risk. Programs like ChatGPT and Grok are enabling a culture of intellectual dependency, leading to declines in memorization, motivation, creativity, and original thought. The next generation is being described as “essentially illiterate” by experts.
While calling someone “stupid” is harsh, it highlights the potential outcome of over-reliance on these tools. Just as we restrict substances that impede brain function for young people, the unrestricted use of LLMs in education warrants serious reconsideration. The potential for negative consequences far outweighs the current understanding of long-term effects. When faced with potential billion-dollar contracts by deploying AI in education systems, the ethical concerns about stunting mental growth may unfortunately be ignored by corporations.
The MIT findings underscore a critical warning: while AI is a powerful tool, its unchecked and unexamined use, especially in formative years, could have profoundly negative and lasting impacts on human intelligence. We must question everything and be acutely aware of the unseen costs.
