Does ChatGPT Make Us Smarter — or Just Faster? A Professional Look at the Evidence

In recent weeks, a series of viral posts by Alex Vacca raised urgent questions about the impact of ChatGPT and other generative AI tools on our brains, our thinking, and our ability to write. His claims were provocative — but are they true? And if so, what does that mean for educators, students, and professionals using AI every day?

At DTO, where we help dance studio owners and educators integrate tools like ChatGPT into their businesses and classrooms, we believe in using tech responsibly, not blindly. So, we dug into the original data and peer-reviewed research to separate facts from fear.

Vacca cites a statistic from a recent experiment:

“83.3% of ChatGPT users couldn’t quote from essays they wrote minutes earlier.”
–@itsalexvacca on X (June 2025)

He describes this as a form of “cognitive offloading” — where the brain doesn’t retain information because the AI did the thinking. A bar graph (visible in his post) showed ChatGPT users underperforming compared to those who used Google or wrote without any tools.

Backed by Research?
Yes — partially. A recent Microsoft Research study titled The Impact of Generative AI on Critical Thinking (Lee et al., 2025) surveyed 139 knowledge workers and found:

• Using GenAI reduced self-reported cognitive effort
• Users became less confident in their critical thinking
• The more people relied on AI, the less mental effort they exerted

One of the more eye-catching claims:

“Brain scans revealed neural connections collapsed from 79 to 42 — a 47% reduction in connectivity.”

According to EEG results referenced in the same study, participants who wrote without AI showed:

• Stronger neural connectivity

• Especially in the theta and high-alpha bands, associated with deep concentration and memory formation

By contrast, those using AI demonstrated less brain activity across these regions.

Vacca also shared quotes from two English teachers who reviewed essays — some AI-generated, some not:

“Soulless.”
“Empty with regard to content.”
“Perfect language without personal insight.”

While the teachers didn’t know which essays were AI-assisted, they could feel something was “off.” This suggests that while ChatGPT can produce grammatically correct, high-scoring content, it may lack original voice and emotional nuance — unless the user actively guides it.

Perhaps the most striking insight:

“When researchers forced ChatGPT users to write without AI, they performed worse than people who had never used AI at all.”

This mirrors concerns from the Microsoft paper, which called out a “confidence-effect paradox”:

• The more confident people felt using AI, the less actual critical thinking they applied

• Over time, skills eroded, like a muscle that hasn’t been exercised

And here’s the twist. AI does help:

• According to research cited in the posts, ChatGPT makes you 60% faster at completing tasks.

• But it also reduces “germane cognitive load” — the mental effort needed for learning — by 32%.

In other words:
You’re trading long-term brain development for short-term speed.

Absolutely not — but we must use it better.

AI tools are like calculators, spell checkers, or search engines. They can help or harm, depending on how they’re used. The danger isn’t ChatGPT itself, but:

• Using it uncritically

• Submitting responses without editing

• Letting it become a substitute for learning, rather than a partner in it