Economic Observer Follow
2026-05-11 13:27

At 8 pm, Li Lin (pseudonym), a second year junior high school student from Guizhou, sat at her desk, opened the AI dialogue window, and took photos of the reading questions on her Chinese exam paper to upload. Three seconds later, a well-organized and beautifully worded answer appeared on the screen. She quickly copied and finished writing, closed her workbook, and the whole process took less than five minutes. I know this is not right, but the answer given by AI is so good that I can't even write it myself. ?
On the other end of the city, Teacher Lei, who works at a training institution, inputs students' homework and classroom performance into AI, and the system quickly generates a well worded and well-organized feedback text. It did save a lot of time, "he admitted," but I also wonder, if one day there is no AI, can I still write such feedback? ?
Sun Xiao (pseudonym), who is engaged in new media copywriting, is struggling with the promotion of Party A's new products. She has become accustomed to using AI generation, and sometimes even a sentence is casually thrown in, slowly forming a dependency. This mechanical state of receiving and outputting seems like a disease.
At the same time, more and more netizens are claiming to have been diagnosed with "frontal lobe damage", describing themselves as having difficulty concentrating, unstable emotions, procrastination, and lack of self-control after heavily using AI. However, this is a self mockery that follows the trend and is fundamentally different from medical diagnosis.
When the professional skills, creative thinking, and judgment abilities that people take pride in are quietly outsourced to algorithms, are those thinking and muscle memory abilities that originally required repeated practice to form, such as students' problem-solving ability, creative people's hand drawing skills, and professionals' analytical judgment, shrinking? Will excessive reliance on AI have an impact on the function of the prefrontal cortex in the brain? Will a person become a hollow person?
Tool Dependency
The 2026 Artificial Intelligence Trends Report shows that by 2026, 87% of creative practitioners will have used AI tools in their daily work, with 66% reaching a weekly high usage level. Looking at the entire workplace, the use of AI is more common, and most workers have tasted the sweetness of AI.
Duan Yang, who is transitioning from manual creation to AI assisted manual work in new media and visual learning and work, has a deep understanding of this role transition. After AI enters her job, she can get rid of some time-consuming and laborious work, and spend more energy on thinking and creativity, but sometimes her initiative in thinking and creativity weakens, and her hand drawing skills also decline.
In the past, people would start from scratch and do a lot of divergent thinking and manual operation work. Now, almost every day, AI is used to provide inspiration, find information, or perform simple one click 3D modeling. Occasionally, when you need to model yourself, many shortcut keys are not commonly used and cannot be remembered. The speed and precision are not as good as before. All creativity has become a mechanized product that can be quickly generated and consumed, which is actually very sad. ?Duan Yang said.
Zheng Dan (pseudonym) is a self media blogger with 20000 followers. In the past, it took several days to write copy and create images and text, and the update frequency could only be three times a week. With AI, he can basically do daily updates.
I have been extensively using AI creative tools recently. The efficiency of NotebookLM, Doubao and ERNIE Bot is a bit unreal. You don't need to modify any information just by searching, and you can generate a seemingly problem free piece of content in just a few minutes. If the content requirements are not particularly strict, AI can basically meet the needs and allow me to devote more energy to other work. ?
In the recruitment industry, the phenomenon of widespread use and dependence on AI is not uncommon. According to the "2025 Workplace AI Tool User Experience Report" conducted by Zhilian Recruitment, approximately 80% of workplace workers use AI tools in their work. 33.4% of professionals use AI every week, with 13.4% using it 1-3 times a day and 10.1% using it more than 3 times a day, making it a valuable tool for the workplace.
Zhang Li (pseudonym) is a human resources practitioner in a private enterprise, mainly responsible for the company's daily attendance and human resources recruitment. Previously, screening resumes and checking matching accuracy relied solely on manual labor. However, during peak season, hundreds of submissions were made in a single day, and it was easy to miss out on some outstanding candidates for job matching due to eye strain. Now she throws her resume into the AI system and completes the first round of screening in just a few seconds. Even the candidate's project experience keywords can be automatically highlighted in red. "The most time-consuming coarse screening process used to be almost brainless now, and the system's results are even more accurate than mine," she said
But there are also hidden concerns when using tools for a long time. During an interview, she asked the other party as usual, 'What is the biggest challenge of this internship?' The other party's answer was stuttering, completely different from the perfect summary extracted by AI. The labels given by AI are too clean, which actually grinds away the real edges and corners of people. ?Zhang Li said, 'I will deliberately use less AI now, afraid that I will forget how to judge a person based on intuition.'. ?
AI has helped many practitioners improve work efficiency, but as people lack thinking, they become more and more like "hollow people" who only know how to click, copy, and paste.
In 2025, research by Microsoft and Carnegie Mellon found that the more AI is used, the weaker its ability to think independently and solve problems. Many programmers admit that if they don't read the documentation or check for errors, their first reaction when encountering bugs is to ask AI. Over time, they even become unfamiliar with the basic skills of debugging (troubleshooting program problems). When AI crashes or cannot provide answers, one is at a loss.
"It took me a few minutes to get an answer, a few minutes to identify the problem, and a few hours to solve the problem," roast an intern of Internet operations.
AI users from different professions expressed the same dilemma: I seem to have gradually lost my creativity and thinking ability, unable to do without AI while feeling a little scared.
Will AI make people stupid?
In fact, the entry of AI into human life is more like a two-way rush between AI and humans. It is not surprising that AI can write essays and quickly correct them in schools; In the workplace, from writing resumes, answering interview questions, to creating work ideas, AI is ubiquitous; In life, AI takes children to liberate parents, AI companions soothe the soul... It cannot be denied that the evolution of AI has become a value-added point for human experience of happiness. It can be an efficient assistant or a cyber friend who understands the language and needs you. But the more perfect the relationship, the more dangerous it is.
Will AI make people stupid? This may be a very practical problem. AI may cause a decline in human thinking ability, but this decline is not simply about becoming stupid, but rather about using it to make progress.
Nowadays, it is difficult for AI to understand things like humans do, with a fixed style and logic. AI only has almost complete and common sense general knowledge, which can replace some low-end thinking and reasoning processes of humans, but cannot replace human innovation ability. ?Scholar Li Mu (pseudonym) in the field of artificial intelligence talked about it.
As AI becomes increasingly integrated into work and life, some studies have revealed a worrying trend. Some new studies suggest that excessive reliance on AI models may lead to "AI fog", causing unclear thinking and decreased attention.
A study conducted by the Massachusetts Institute of Technology in the United States provides direct evidence: researchers conducted EEG scans on 54 college students and found that those who relied on ChatGPT writing had a decrease in brain neural connectivity activity of about 47% compared to those who only used brain thinking. Even more alarming is that when asked to retell content they had just completed a few minutes ago, 83.3% of AI users were unable to recall any valid information, compared to only 11.1% in the pure manual writing group.
On the other side of the coin, AI is amplifying individual potential in unprecedented ways, such as the wave of entrepreneurship sweeping across the country with OPC (One Person Company). Data shows that as of June 2025, the number of one person limited liability companies in China has exceeded 16 million, and the number of new registrations in the first half of 2025 has surged by 47% year-on-year. Chen Mei, an entrepreneur from Xiamen, runs around ten AI employees on her own, and over 90% of the company's online business is completed by AI. For these entrepreneurs, AI is not a tool to replace their brains, but a lever that helps them amplify their personal abilities several times.
AI may be exacerbating the Matthew effect of capabilities while improving efficiency. ?Teacher Lei believes.
Wang Hui (pseudonym), a graduate student majoring in translation, has a deep understanding of this. She admitted that AI has indeed impacted traditional translation positions, but the real danger is not that AI makes people stupid, but that people voluntarily give up thinking. Language proficiency must be continuously learned, with grammar and vocabulary as the foundation. At the same time, it is necessary to master AI tools to improve efficiency and establish its own irreplaceability in areas where AI is not yet proficient, such as legal translation and cultural slang. ?
The answer to whether AI will make people stupid may not lie in the technology itself.
Denoising The Brain
Can I still work without AI? ?This is a question that many professionals ask themselves late at night.
Mr. Du, a senior figure in the enterprise marketing industry, gave an interesting analogy: "AI is like wings, some people put in the wings of a pheasant, while others put in the wings of an eagle. ?He pointed out that currently AI can only solve basic problems, and people who rely entirely on AI will eventually encounter obstacles in real market challenges. The 'hollow person' either degenerates or adapts to the environment and becomes a person who can truly cope with work
Education economist Yang Dong (pseudonym) found in long-term research that 60% -70% of college students in China are already using AI to complete assignments, and almost all essay writing, programming, and open-ended test questions can be done by AI. Without a solid foundation, no matter how big the platform is in the future, it will not be able to withstand it. "He emphasized that college students should start from their freshman year to solidify their basic skills and let AI become an amplifier of their abilities, rather than a substitute.
Wang Ming (pseudonym), a psychological counselor at a university in Beijing, interprets this phenomenon from a psychological perspective. "People who cannot do without AI often have lower self-efficacy, worry that their creations are not as exquisite as AI, or are unwilling to face the pain of thinking. ?He said that 'being anxious and unable to think without AI' may be the result of insufficient training of abilities, or it may be a manifestation of self-identity deviation.
From a physiological perspective, excessive reliance on AI is weakening the frontal lobe function of the human brain. The prefrontal cortex is an advanced center in the brain responsible for decision-making, reasoning, personality expression, and social appropriateness regulation. Long term disuse can directly lead to attention deficit, working memory deficits, abnormal emotional responses, and even weakened judgment and creativity.
Faced with the anxiety of "hollowing out" caused by AI dependence, some users on platforms such as Xiaohongshu, Weibo, and Zhihu have summarized a series of coping measures.
Some users of Xiaohongshu are forced to schedule a "fixed one hour no AI writing time per day" for themselves, writing diaries and essays by hand, and resolutely not using AI editing. The effect is very obvious: at the beginning, the hand feels sluggish, but after two weeks of persistence, the thinking becomes noticeably smoother, and there will be no more stuttering when writing. Another Zhihu user defined this rhythm as "Weekly Digital Fasting Day without AI", which means choosing one day every week without using AI, and relying solely on oneself to search for information, write solutions, and fix bugs. Equivalent to rebooting the brain.
A design blogger insists on hand drawing sketches and writing initial drafts before using AI to optimize details, and firmly refuses to let AI directly produce initial drafts. This can not only preserve the original logic and personal style, but also avoid losing the basic skills of hand drawing and conception.
A workplace blogger has invented the "AI usage three question self check method" - every time he uses AI, he must ask himself three questions: Is the answer correct? Do you have a better idea? Can I explain it clearly independently? His idea is to not deceive himself and gradually get rid of passive dependence.
Professor Luo Xiang from the School of Criminal Justice at China University of Political Science and Law shared his views on technology and thinking on public social media: technology is a tool, not a master. The reason why people are human lies in independent thinking and value judgment. We can use AI to save effort, but we cannot give up thinking with it. True wisdom is knowing when not to use tools.
(The authors of this article are Lv Rong, Sang Meng, Shi Xingyu, Lv Ruijia, Liang Xin, and Feng Sipei, graduate students of the School of Language and Communication, Beijing Technology and Business University, Class of 2025)

Tech Show | The Science and Technology Expo has come to an end, with various districts in Beijing showcasing their 'black technologies'

Technology Show | Visiting the Beijing Science and Technology Expo, how can these technological exhibits rewrite the future?

What is more important than algorithms when AI competitions enter the 'deep water zone'?