Image source: Generated by Unbounded AI

On campus, AI is quietly causing a revolution.

Papers and assignments that once made students rack their brains can now be completed in a few minutes with the help of AI tools. The sharp increase in the rate of AI inclusion has attracted the attention of colleges and universities, and they have introduced regulations to regulate its use.

But the impact of AI goes far beyond that. It is everywhere from daily office work to medical transportation. The recent DeepSeek-V3 launched by Chinese start-up Deepseek has triggered a new round of industry discussions on the inclusiveness of AI with its low cost and high performance. When AI inevitably invades people's lives, how should we pay attention to the hidden risks?

Less than 10% of teachers and students rarely use AI

Recently , "Universities issuing AI bans" has become a hot topic, triggering heated discussions in the education sector and even society.

On the school's side, the situation is a bit tricky. In the past, students had to spend a lot of time completing course papers and defense PPTs, completing experiments, reviewing literature, and repeatedly revising and polishing papers... In the process, students' critical thinking and knowledge integration and application were indeed exercised. ability.

But now, with the help of AI tools, you can produce a decent graduation thesis in a few minutes, and the defense PPT is also a piece of cake. This makes schools and teachers very headache, after all, once academic fraud is rampant , the foundation of education will be shaken. Therefore, many universities, including Fudan University and Communication University of China, have introduced relevant regulations or measures to regulate the use of AI tools by college students.

According to Phoenix Technology, many students are indeed interested in AI. Whether it is daily homework or graduation thesis, I want AI to help. For example, Xiao Jia, a student at a university in Zhejiang Province, told Phoenix.com that in his graduation thesis of 30,000 to 40,000 words, in addition to data models, many text descriptions sought help from AI tools.

Research by Max, a three-party research institution that provides data services to universities, also shows that more than 30% of undergraduates mainly use AI to write papers or assignments. Michaels interviewed more than 3,000 university teachers and students in October 2024. The results of the questionnaire showed that almost all teachers and students interviewed had used AI tools in study and work. “Teachers and students who have never or rarely used AI The proportion is less than 10%.”

The popularity of AI papers among universities is closely related to the accurate introduction of new AI products.

Open social platforms such as Xiaohongshu and Douyin, and you can see posts like this everywhere: "Hey guys, I found a super powerful AI writing artifact!" "Wow! This AI product is simply awesome. It’s my paper savior!” “Family members, you must try this super popular AI translation tool.” These propagandas are overwhelming and targeted directly at college students.

College studentsOften faced with a variety of learning tasks, including the above-mentioned paper writing, coursework, information inquiry, etc. Xiao Jia said that AI can help him polish his paper. "When he encounters no inspiration, AI can also give good ideas," which can greatly improve the completion efficiency. It is precisely because they have recognized the needs of college students that major model manufacturers have heavily advertised here.

Picture | Screenshot of Xiaohongshu

Take the AI ​​tool "Doubao" as an example. A promotional post it posted on Xiaohongshu wrote: "Doubao AI can analyze 2025 graduate students. Change trends in the number of applicants, share efficient postgraduate entrance examination study methods, develop personalized sprint preparation plans, and provide real-time methods and suggestions for psychological adjustment before the examination, etc.

Under this post, a candidate left a message: "It's easy to use, I use this every day (referring to the bean bag)." This shows its penetration into the student population.

Looking at the start-up Kimi smart assistant, the new promotion also targets the student group: "College students are blessed! Let Kimi read the papers you can't understand; practice the English you can't speak with Kimi; If you have no clue about the internship interview, let Kimi prepare it; if you don’t want to do PPT, let Kimi handle it!” Every line seems to express: AI is the best helper for students’ study and life.

Therefore, it is easy to understand that different universities have successively introduced regulations to regulate students' use of AI tools.

Universities explore the boundaries of AI

In the past two years, AI tools have It is becoming more and more common on campuses, and universities have to adjust their management standards for AI accordingly.

As early as early 2023, the University of Hong Kong issued a notice prohibiting students from using AI tools such as ChatGPT to complete coursework, assessments, etc. If a student secretly uses the material without the teacher's written permission, it will be treated as plagiarism.

However, a few months later, the University of Hong Kong’s attitude changed. It has begun to provide teachers and students with a variety of generative AI applications for free, and allows students to issue up to 20 instructions to AI every month. The reasons given by the University of Hong Kong are also very real: Generative AI is rapidly becoming popular and developing, and it is very important for students to master the ability to apply AI tools - in addition to the four important communication skills of speaking, writing, video, and digital, students must have The fifth important ability.

The attitude of Hong Kong universities towards AI tools has changed according to the actual situation, and universities in the mainland are no exception. In the first half of 2024, many universities such as North China Electric Power University, Hubei University, Fuzhou University, Nanjing University of Technology, Tianjin University of Science and Technology, and Communication University of China have issued notices saying that when reviewing undergraduate graduation thesis, they will try to add a new Detection, specifically detecting whether there is a risk of AI cheating in the paper.

Some people in the industry pointed out that colleges and universities have increased the detection of AI in the review of undergraduate thesis in order to supervise students.Students complete the writing of thesis themselves to maintain academic integrity and standards. In addition, "only testing but not blocking" also shows that colleges and universities have clearly seen the general trend. AI is "conquering cities and territories" in the field of education. Schools must actively respond and guide according to the trend.

This is not over yet. Many schools are holding AI competitions to encourage students to explore new applications of AI technology in different fields.

Although, the "Fudan University Regulations on the Use of AI Tools in Undergraduate Thesis (Design) (Trial)" issued by Fudan University on November 28 is more stringent and clearly states "six prohibitions. ", even called "the world's most stringent AI ban" by many media.

But it does not simply and crudely prevent students from using AI. On the contrary, Fudan announced in June last year that it would launch at least 100 "AI major courses" in the 2024-2025 academic year. According to public reports, many teachers at Fudan also take the initiative to integrate AI content into their courses.

It can be seen from these initiatives of universities that they are all trying to grasp the "heat" of the integration of AI and education. On the one hand, graduation thesis is related to academic rigor, and AI is not allowed to become an "accomplice" in students' fraud, so strict review and accurate testing are imperative. On the other hand, colleges and universities have not killed AI with one stroke. They are also actively embracing AI competitions and offering AI courses to open the door to AI for students. Behind the seemingly strict "ban", there is actually a deeper educational intention. To put it simply, we hope that students can not only maintain their academic bottom line, but also keep up with the development of the times by leveraging the power of AI.

How to dance the double-edged sword of AI development?

More than 20 years ago, a science fiction masterpiece "The Matrix" told the story of a human being immersed in a virtual world and ultimately saving himself. By exploring the relationship between humans and AI and the consequences of AI running out of control, it has triggered the thinking of countless moviegoers and opened the door for many people to yearn for the AI ​​world.

More than 20 years later, the rapid development of AI has allowed many unattainable "science fiction" settings to penetrate into our lives, and it has also allowed us to see the intelligent productivity brought by AI. Lots of room for imagination.

Students skillfully use AI to polish their papers, office workers habitually ask AI to "generate PPT according to requirements" and "extract keywords from pictures", and bloggers ask AI to "summarize the top ten Internet memes in 2024" ""Think of a few more attractive titles based on the video content" "Create a Hitchcock-style video"...AI has infiltrated the entire Internet. As long as you open the Internet, the AI ​​content of various platforms and channels is very high.

On Xiaohongshu, there are a large number of accounts that use AI to generate a picture, a sentence, or a video. They publish only a dozen pieces of content, but have hundreds of thousands of fans, allowing a large number of original Xiaohongshu creations The readers are confused and don’t know how to fight against the problems caused by AI.Influence.

Picture | Screenshot of Xiaohongshu

What’s interesting is that even AI itself is being polluted by AI.

Deepseek (Chinese name "Deep Search") has recently made a splash in the global AI industry due to the release of the lower-cost but better-performing V3 model. However, someone discovered during testing that DeepSeek-V3 occasionally claimed to be ChatGPT. Later, industry insiders analyzed that DeepSeek may have used a public data set containing text generated by GPT-4 through ChatGPT. At the same time, DeepSeek-V3 is not the first model to misidentify itself. Google's Gemini and others sometimes claim to be competing models. It can be seen that the concentration of AI on the Internet has reached a record high.

The impact of AI goes far beyond daily office work and content consumption. The autonomous driving we are familiar with is breaking through the bottleneck through large models. New car-making forces have begun to announce that they will become artificial intelligence companies; AI is even used in the field of drug research and development, and AI doctors will also be implemented in the near future. … When the day comes when AI fully enters thousands of industries, how should we get along with AI? While enjoying the great convenience and unlimited possibilities brought by AI, we need to be alert to issues such as technology abuse, data privacy leaks, AI fraud, and even infringement and violations. How to figure out the appropriateness is a question that each of us must think about.