Gone are the hours spent reading; now, a chatbot can process hundreds of pages and summarize or explain content in mere seconds, replacing human effort. Nearly four years after ChatGPT's initial public demo in 11/2022, artificial intelligence has significantly altered human reading skills and societal perspectives.

A February report by the Pew Research Center on how US teenagers perceive AI revealed that half of 1,458 adolescents and parents use chatbots for learning and work. Specifically, 57% of teenagers use AI to find information, and 42% use it to summarize books, articles, or videos. This data indicates that AI usage has become widespread in the US market, consequently changing users' reading skills, habits, and critical thinking since ChatGPT's introduction.

ChatGPT's interface demonstrates its capabilities. Photo: Unsplash

People are increasingly less patient when it comes to deep reading. This isn't due to a lack of information; rather, they are overwhelmed by it. According to the Pew Research Center, one teenager, when asked about AI's negative impact, stated, "People will fear becoming creative or will find creativity unnecessary. AI makes people lazy and takes our jobs."

Forbes reports that reading has become more technical: podcast listeners often speed up playback to 2x, people skim articles for bullet points, or use ChatGPT to summarize books to feel like they've read them. This convenience saves time but transforms the reading process into a mere collection of conclusions.

While artificial intelligence can "read" and provide results or answers that satisfy user needs, human reading offers context, experience, and deep insights into a text. AI merely reproduces language without truly understanding the meaning behind the words. This distinction leads to separate outcomes for the human brain's process of understanding, learning, and evaluating issues.

AI cannot replace critical thinking

Critical thinking begins with the ability to grasp multiple perspectives, tolerate ambiguity, and ask better questions. These skills are not formed through a few bullet points. They are slowly honed through the reading process, as readers grapple with complex ideas, connect facts independently, and question their initial understanding.

In 12/2025, the MIT Media Lab published results from a study on human brainwave activity during essay writing, clarifying this risk. The research team divided 54 individuals from five universities in Boston, US, into three groups: writing essays using ChatGPT, using Google Search, and using no assistance tools.

While this study has not undergone academic peer review and has a small sample size, its results are noteworthy. The ChatGPT group exhibited the lowest level of brain engagement and "performed worse" across neurological, linguistic, and behavioral metrics. Additionally, two graders observed that many essays from this group were similar, lacked critical thinking, showed little original thought, and were largely "soulless."

Image introducing the latest GPT-5.5 model in April. Photo: OpenAI

Nevertheless, the issue is not about restricting AI from reading. The more important question is whether it is used before or after humans have reflected on the content. AI has the potential to erode critical thinking, but this outcome is not inevitable. When used correctly, artificial intelligence can help readers compare multiple viewpoints, test assumptions, pose counter-arguments, or act as a conversational partner to assess the strength of an argument.

In practice, AI should serve as a reading aid, not replace the act of reading itself. For a difficult text, users can ask AI to explain terminology, list characters, suggest historical context, or present opposing viewpoints. However, readers must then return to the original text, verify sources, and independently answer: What is the author arguing, which evidence is credible, what has been omitted, and do they agree?

Reading, therefore, is not an outdated habit in the AI era. It is a process that hones attention, patience, and independent judgment. Artificial intelligence can summarize, suggest, explain, and debate. However, it cannot replace the human experience of grappling with a difficult idea, the empathy felt when immersing in a story, or the responsibility of drawing one's own conclusions.

In an era where machines can generate instant answers, the capacity for slow reading and critical thinking is a human advantage.

Thao Uyen (according to Forbes, Pew Research Center, MIT Media Lab)