Why Big Tech’s Push to Automate Thinking Could Harm Human Cognition

Leading artificial intelligence figures have long framed the human mind in mechanical terms. Google DeepMind’s Demis Hassabis has described the brain as "a biological approximation to a Turing machine," while Elon Musk has asserted that people should view it simply as a "biological computer." However, emerging perspectives from neuroscience suggest this computational metaphor is deeply flawed—and that aggressively inserting AI into our cognitive lives could carry serious consequences for human learning.

Rather than an algorithmic machine that simply processes inputs into outputs, human nervous systems evolved as dynamic feedback-control systems, argues Paul Cisek, a neuroscientist at the University of Montreal. According to Cisek's biological model, our brains developed over millions of years to interact with, explore, and exercise agency within our physical and social environments. By treating the mind merely as an information processor, tech developers risk bypassing the active engagement that builds lasting knowledge.

The Threat of "Cognitive Automation"

Critics are comparing generative AI to a form of mental junk food: convenient and tempting in the short term, but detrimental to long-term health if it becomes a regular mental diet. Former Google software engineer François Chollet has labeled this shift "cognitive automation," defining it as encoding human abstractions into software to automate tasks typically performed by people.

This dynamic is especially visible in education, where technology firms are pushing their tools at an unprecedented scale:

• In July 2025, OpenAI VP of Education Leah Belsky stated that "learners" made up more than half of ChatGPT’s 900 million average monthly users, calling the tool "the world's largest learning platform."
• Anthropic recently introduced its "Claude for Teachers" initiative, offering classroom educators free access to premium models.

Yet evidence indicates that offloading mental tasks harms genuine learning. Researchers point out that when students use AI to avoid effortful thinking—a phenomenon dubbed "cognitive delegation"—they weaken their autonomous reasoning capabilities. A study conducted in China showed that thousands of students substantially harmed their learning outcomes after they effectively stopped doing their homework upon turning to AI tools. Other research indicates that habitual reliance on large language models leads students to relax critical judgment in wider contexts.

Pushback and "Cognitive Immunization"

In response to concerns that cognitive delegation undermines the human thinking process, calls are growing for what researchers term "cognitive immunization"—preserving structured practices like unaided problem-solving, verification, and deliberate disengagement from AI tools.

Policymakers and educators are increasingly turning to bans to protect young learners:

• Norway recently prohibited nearly all uses of AI in schools for students under the age of 13.
• In the United States, a prominent teachers' union has advocated against chatbots in elementary schools.
• Major US metropolitan school districts, including New York City and Los Angeles, have introduced restrictions and bans.

Resistance is also emerging directly from students. At Oberlin College, members of the Oberlin Luddite Club issued an open letter challenging the administration's institutional adoption of AI, advocating instead for "embracing human imperfection and raw inquiry."

As schools and institutions navigate the influx of AI tools, the debate is shifting from how fast AI can be adopted to how societies can protect the essential, effortful processes required for human cognitive growth.