I dare not say whether the 21st century is the century of biology, but 2026 will definitely be the year when AI dominates everyone’s attention. Programmers were laid off, one-man companies exploded, lobsters invaded the world, and the carnival in the AI circle resulted in increasing anxiety among ordinary people.
AI is really the future, but what about my job? So many people regret that they did not choose an AI-related major in the college entrance examination, as if they missed the biggest trend of this era.
After all, in the eyes of laymen like us, those who study these majors now have well-written scripts, spend all day studying how to make large models in the laboratory, and are better at AI than anyone else. If we say that they are natives of the AI circle, we are just despicable outsiders who do nothing but work on AI all day long.
So, what is happening to these people who got the tickets for the New Era?
With this curiosity in mind, I chatted with several bachelor's, master's and doctoral students from different universities who are still studying AI. Unexpectedly, in this ivory tower that has been infinitely elevated by the outside world, there are not so many well-informed trendsetters of the times. In fact, they are just like ordinary people, stumbling and groping within the framework of reality.
First of all, the school’s training plan has not kept up with last time, and the use of AI is never taught in class.
But it’s reasonable. The school’s expectation for them is to create AI, and they must first teach basic knowledge. There’s nothing wrong with that.
But the reality is that the AI majors in many schools are split from the School of Information or the School of Computer Science, and the training programs are still in the transition period. Some are more interested in electronic communications, and some are still using their old background in accounting.
Li Mo, a student from a 985 school in Anhui Province, complained that their class coincided with the first year when the AI major became independent from the College of Letters. The training plan has not yet been changed. "Signals and Systems" in the communication major is required, but there is no need to take "Compilation Principles" and "Operating Systems" that are essential for computers.
The most torturous thing is that I have to take a bunch of physics courses such as "Electromagnetism" and "Quantum Physics". . For computers, this stuff is difficult and useless.
In addition to theoretical courses, programming is actually not taught so perfectly.
Friends who know AI should know that Python is currently the most widely used programming language in the field of artificial intelligence. A student who calls himself "Mo 9 AI" said that their training plan has always been based on C++, which has led to a separation between "learning" and "doing". They use Python for daily projects, while coursework and exams are forced to be written in C++. This is also because they have followed the training plan of the computer major next door.
What you learn is different from what you use. It can be regarded as a test left by the God of AI to the students.
Another thing is that today's AI programming has become so powerful that it is out of the industry. There are reports of layoffs every day on the Internet. Dario, the founder of Claude, also revealed that most of the code is no longer written by employees. .
But the school is definitely not that advanced with the times. During the investigation, I learned that there are indeed no schools offering courses on AI programming, and everyone is sticking to ancient programming.
When the teacher is explaining PPT, he may occasionally mention it in class, "The recently popular GPT, Claude, is used by many people to write code, so what does this tell us? What students need most in the future is actually creativity." Then it is over. If you want to really learn some AI programming skills, you can only understand them by paying attention to the technology circle and watching videos.
More than one classmate mentioned,Some of the knowledge and technology stacks taught in schools are outdated. The role of the teacher is to open the door to this field for you, and you have to explore it on your own at a deeper level.
The negative reviewer thinks it is reasonable. My friends have graduated so many years ago and the school version update is still so slow, so I feel relieved.
Until pony from a 211 in Zhejiang told me, "The teachers don't write the code, we do it." I suddenly understood that students are the best agents.
Teachers’ attitudes towards AI are quite divided.
Zhang Facai, a classmate from a 985 program in Hunan, told the reviewer that he studied network security. Regarding the use of AI, the network security teachers have a clear attitude and do not encourage its use. They believe that many operations must be completed by hand in order to truly internalize the principles. Especially in the security field, handing over operations to AI is a hidden danger in itself.
The teacher in the web development direction, on the other hand, took the initiative to mention in class that using AI correctly can save a lot of trouble in web development.
To put it another way, the teachers on both sides are actually right. Cybersecurity requires practitioners to have an almost muscle memory grasp of the underlying principles, and this step cannot be skipped; and there is a lot of patterned work in web development, and AI can indeed significantly improve efficiency.
The problem is that no one can lead students to form their own judgment system. It is up to the students to figure out in which scenarios AI should be used, in which scenarios they must learn by themselves, and how to use it without losing basic skills.
This may also lead to a problem: people who can use AI use it as a tool and learn faster the more they learn it; those who can’t use it throw everything to AI and accumulate nothing after one semester.
In this way, the students in the audience will no longer pretend. If you don’t teach me, I will learn on my own. When asked about the class situation, Xiaoxue, a student from a certain 985 program in Shanghai, said, "Everyone doesn't listen to the teacher very much now. They are all immersed in learning. There is indeed a gap between what the industry needs and what the school teaches."
And because AI is becoming more and more convenient to use, in the coursework part, everyone’s AI usage rate is already very high. "Now the basic AI participation rate in writing courses is the whole process, and there is no one who does not use it." Zhou Kai, a doctoral student from Shanghai, told us. Students in other grades also got the same answer.
What's interesting is that when everyone talks about "it's very convenient to use AI", they always bring a "but". No one feels that they have fully mastered AI.
Classmate Zhang Facai said that if you rely too much on AI, you will lose control of the project and the code will turn into a mountain of shit. The time saved by writing code is spent on reviewing the code.
"What is written is different from what you think, and it is very time-consuming to repair. You still have to have development experience and type the code by hand before you can truly control AI."
Zhou Kai, a doctoral student, said that he spent 200 million tokens during the winter vacation, and his final feeling is: it is no problem for AI to write simple scripts directly, but high-level design still has to be done by yourself. "Because if you talk to AI too much, you will find that you become stupid. The summary that AI gives you looks good, but you can't see where it is wrong, and it is slower than doing it yourself. Many experts dare to use AI with confidence because they already know what is right and what is wrong."
Xiaoxue, who was preparing to submit her resume and answer the questions, immediately started complaining: After using the AI for a long time, she could no longer write code and could not even solve a single question.
Everyone is using AI crazily, but no one feels that they have the answer.
But the reality for them is completely different.
Nowadays, the requirements of major Internet companies for fresh graduates are all kinds of strange. Whether or not to "rip the code by hand" (typing out a piece of code without using any tools) during the interview has become a metaphysical question. Some people still insist on ancient programming, and some have evolved to the stage of AI programming.
Teachers usually allow the use of AI, but they may not be allowed to use it during interviews. It is equivalent to suddenly asking you to test the Pythagorean Theorem in the college entrance examination. Only you know the tears in it.
But the people I talked about earlier, whether they were people who figured it out or stepped on the trap, are at least people who have the energy to think.
For students in related majors, the real situation is actually, who cares?
Those taking postgraduate entrance exams are reviewing, those taking postgraduate entrance exams are busy with competitions and scientific research, and those looking for jobs are scrambling. What are the new AI tools, the latest progress in cutting-edge AI? The more hype outside, the more noisy people inside are.
Classmate Zhang Facai said bluntly that the reality is too heavy. “It was only when I was in my third year of college that I realized that scientific research experience would not grow on its own.”
Take OpenClaw, the most representative product of 2026, for example. It has exploded on the Internet. The media has reported overwhelmingly, and major manufacturers have not hesitated to follow up.
However, most of the AI-related students responded indifferently. “I feel like I can’t earn back the token money,” said classmate Zhang Qiu.
And classmate pony told us, "There are basically no splashes around me, maybe some teachers want to play."
Zhou Kai considered it more from a safety perspective. It was too dangerous and he was afraid of being injected into the attack.
A product that is regarded by the outside world as "changing the world" is not so popular in the hands of people who really learn AI. The bad reviewer summarized the following two voices, one is seeing the essence, "the essence of lobster is just on the roadside", and the other is pressure. Everyone has more pressing things.
Classmate pony also told us that people in the same laboratory react half a beat slower than the outside world.
"Most of the traditionalists around me are still traditionalists. When everyone used LLM before, we liked to use our hands. Now that we are all coding in AI, we still like to use LLM. We are too lazy to try new tools, and more often think about whether we can publish papers. For my own workflow, I use vim, hand button on the terminal, and Copilot of VSCode - these old things are enough, why should I change to new ones?"
From undergraduates to Ph.D.s, there is a heated discussion outside about how AI will change the world, but people inside are thinking about how to survive the current situation. Because the one-third of an acre of land in front of them is their world.
From this, we can summarize the current situation of the AI circle:The old ones are busy instructing others to embrace AI, the newbies are too busy looking forward to the future and have no time to play with AI, and the middle-aged ones are playing with AI while shouting "the future is here".
The rest are ordinary people who are forced to accept the influence of AI.
Finally, when we talked about whether we should continue to work in this industry in the future and what specific direction we should take, most people said they were indifferent.
However, Zhou Kai still put forward his own opinions. "AI is bringing everyone's engineering capabilities to the same starting line - if Code Agent allows everyone to write code, what is your advantage over others?" He has recently begun to consciously learn expression skills because he feels that in this era, describing one thing clearly is the most valuable.
But more people are still confused, confused about the coursework, interviews and eight-part essays in front of them, and confused about the rapidly sweeping AI wave.
Everyone is in the same water, choking on the water, and learning to swim.