A deepSeek core engineer's lengthy post goes viral: "I had to bury my talent in yesterday."
On September 14, Liu Shengyu, a young engineer who had just delivered the core operator of DeepSeek V4.1, published a long article titled "I Had to Bury My Talent in Yesterday," which recorded technological changes and his personal journey. The article went viral in technical communities at home and abroad, sparking widespread discussion about the fate of developers and the evolution of technology in the AI era.
The hashtag #I had to bury my talent in yesterday# also topped the Zhihu hot list.
As one of the core builders of large-scale models, Liu Shengyu, a machine learning systems engineer at DeepSeek, boasts an outstanding technical resume. He was a member of the 2021 Turing Class of Computer Science at Peking University, served as the captain of the Peking University Weiming Supercomputing Team, and represented the university in the international student supercomputing competition SC23. After joining DeepSeek in April 2025, he undertook key research and development work on underlying operators, and just a few days ago completed the code delivery of the main Attention operator (MQA attention with head dim = 512) in DeepSeek V4.1.
In the article, Liu Shengyu frankly admitted that although he was proud of the success of the DeepSeek V4.1 core operator, he was also very clear about the irreversible evolution of technology.
“In another six months or a year, the operators written by AI will most likely be as good as mine, or even surpass mine,” Liu Shengyu said frankly. “AI can think of 300 tokens in one second, type out a line of commands in half a second, and write a piece of code in twenty seconds, while I can’t; AI can continuously improve in terms of model depth, thinking intensity, tool usage, and even parallelism, while I can’t.”
This rapid evolution has plunged developers into a profound technological paradox: the better they optimize the operators, the faster the inference and training of new models become; the faster the model's capabilities improve, the sooner AI will replace human operators.
Faced with this irresistible trend, Liu Shengyu wrote: "Human beings have always been ruthless in their self-destruction... Of course, I hope I will not be revolutionized, but if I am to be revolutionized, I hope that I will be the one to revolutionize myself."
"I won't lose my job, but I'll have to change careers."
In response to concerns about programmers' livelihoods, Liu Shengyu offered a sober assessment: engineers will not face true "unemployment," but they will inevitably need to undergo a profound "career transition."
"Changing careers means I have to give up the field of operator design, writing, and optimization that I have been deeply involved in and passionate about, and instead become an 'armor pilot' for agents," he wrote. "Before, my interests, my strengths, and what the industry needs were basically aligned; but now, the industry's needs have shifted from 'people who can write high-performance operators' to 'people who can use AI to produce high-performance operators faster'."
He likened the loss of this craft to the impact of automated sock knitting machines on veteran artisans. Even if they could still use the machine to their best advantage based on their past experience, the pleasure of "listening to the rain by the window, threading the needle, and slowly passing the time" was ultimately crushed by the roar of the machine.
“I have more gears in my hands, but less rhythm in my heart,” he lamented.
In addition to his personal career change, Liu Sheng also expressed concern about the decline in engineering skills among the younger generation in the article. If students generally rely on AI to complete complex coding assignments in minutes, there may be a gap in system architecture and abstract design capabilities. "A person with poor engineering skills, when paired with AI, can produce mountains of garbage at several times the previous efficiency, thereby planting all sorts of problems in the system and making the world more chaotic."
Clarifying the original intention: A dignified farewell to the "era of handicrafts"
Because the article sparked excessive interpretations on various platforms regarding geopolitics, institutional competition, and even unemployment anxiety after its publication, Liu Shengyu subsequently issued a supplementary statement clarifying his original intention in writing the article.
He stated that the article was not intended to express anxiety about his job, nor was it meant to provoke a confrontation between DeepSeek and closed-source organizations like Anthropic. Its core motivation was simply to bid a serious farewell to the pure days of "purely handwritten operators."
"I imagine that one day in the future, 'handwritten operators' or even 'programming' may become a form of entertainment rather than a productive activity, just as almost no one uses javelins for hunting now, but rather as a competitive activity," Liu Shengyu wrote. "This is equivalent to forcing me to give up what I once loved and forcibly turn in another direction. Even if the new direction is equally fascinating, this feeling is not pleasant. As the title says: I have to bury my talent in yesterday."
When asked why he insisted on staying at DeepSeek, Liu Shengyu reiterated his belief that cutting-edge intelligence should be supplied to the whole society in an open and inclusive manner to avoid core capabilities being completely monopolized by a very small number of commercial oligarchs.
"Even after having to bury my talent in yesterday, I still have the light of tomorrow to pursue," he said at the end of his supplementary statement. He added that he will continue to devote himself to the new process of introducing AI Agents to write operators, embracing new technological tools while seeking a foothold in the new era.
The original text is as follows:
I had to bury my talent in yesterday.Original post by intlsy's Doghouse, September 14, 2026, 00:26 Zhejiang
A few days ago, DeepSeek v4.1 was released, taking the capabilities of small models to a whole new level.
The speed of AI development has far exceeded everyone's expectations. From the initial version of ChatGPT, which could only chat in a babbling manner with a context length of only a few thousand tokens, to OpenAI o1, DeepSeek R1, and Kimi K1.5 Thinking with reasoning capabilities, only two years have passed; from reasoning models to intelligent agents that can now smoothly execute commands and complete complex tasks in various harness tools, only a year and a half has gone by. It's hard to imagine what AI will be like in another year, two years, or three years, how powerful it will be, whether it will have already possessed the ability to self-evolve, and whether it will have deeply penetrated fields such as embodied intelligence.AI is getting better and better at writing operators.
AI has also made rapid progress in my field of operator design and programming. In just one year, it has transformed from a simple assistant that could only help me look up documentation, read code, and find bugs into an operator master capable of independently reading CUDA, PTX, and SASS code, analyzing the pause time of each instruction using professional tools, and then independently optimizing operators. I believe that in the near future, it will also have the ability to independently design operator schedulers, evaluate the performance of different scheduling schemes, implement them, and optimize them.
Of course I am proud of the success of DeepSeek v4.1—after all, I wrote its main Attention operator[1], and its excellence is an affirmation of my operator. However, the wheels of time keep rolling forward, and no one can stop the development of technology. I am well aware that in another six months or a year, the operators written by AI will most likely be as good as mine, or even surpass mine. AI can think of 300 tokens in one second, type out a line of commands in half a second, and write a piece of code in twenty seconds, while I cannot; AI can continuously improve in terms of model depth, thinking intensity, tool call volume (frequency of interaction with the environment), and even parallelism, while I cannot.
Humans have always been ruthless in their self-destruction. Why, knowing that "the better my operators are, the faster our new models will train and infer, the faster their capabilities will improve, and the sooner I'll be replaced," do I still choose to optimize my operators? On one hand, writing operators is indeed like playing a video game, providing immense pleasure. The excitement I feel when inventing a new technology or seeing my operator's performance improve is comparable to a speedrunner breaking their own record. At the same time, seeing my operator's performance far surpass the manufacturer's official operators fills me with immense pride. But beyond that, a more important reason is that even if I were to "give up" or deliberately sabotage model training, other models would continue to develop and ultimately destroy me. "Of course, I hope I won't be revolutionized, but if I have to be, I hope I'm the one who revolutionizes myself." With everyone so obsessed with self-destruction, I'm forced to join this brutal arms race.What about me?
What will happen to me when AI's ability to write operators truly surpasses mine?
My assessment is that I won't necessarily lose my job, but I'll have to change careers. I can still keep my current job, but this might mean I'll never have the chance to do the work I once loved again.
I once made a judgment about the changes of the times and my personal future situation: because the times are changing so rapidly (the development of AI mentioned above is a good example), I have absolutely no way of predicting what will happen in five or ten years. But in any case, I believe that with my vision, judgment, initiative, and intelligence, I can stay at the table of the times and stand at the forefront of the times again. However, this judgment can only guarantee that I will not be "unemployed," not that I will not need to "change careers." Rather, this judgment encourages me to avoid unemployment by changing careers.
What does this career change mean? It means I need to abandon my long-cultivated and passionate field of operator design, writing, and optimization, and instead become an "Agent pilot." Previously, my interests, my strengths, and the industry's needs were largely aligned; now, AI has made what I'm good at even better, and the industry's demands have shifted from "people who can write high-performance operators" to "people who can use AI to produce high-performance operators faster." To adapt to the industry's needs, I must abandon my previous passion and turn to an unknown new direction. I believe I can continue to produce high-quality, high-efficiency operators based on my understanding of engineering, upper-level model requirements, and underlying hardware. I also know I might love this new direction (or I might not), but the feeling of having my passion taken away is truly unpleasant. The quiet joy of sitting at my desk, calmly writing operators for an afternoon, may become a thing of the past this summer. I have to bury my talent in yesterday and become a mecha pilot. My hands have more gears, but my heart has lost some of its rhythm.
To illustrate, imagine you're a master knitter, especially skilled at creating various patterns and matching colors. Your sweaters are of excellent quality and beautifully patterned, attracting wealthy people from far and wide who commission you to knit for them, earning you a considerable sum. You thoroughly enjoy the feeling of sitting by the window, brewing a pot of tea, gazing at the green mountains, clear waters, cattle, sheep, and rising smoke outside, quietly knitting for an afternoon. But one day, someone invents a magical machine that, with just yarn and a pattern, can automatically knit sweaters of comparable quality and texture to your hand-knitted ones, and at a much faster speed. You know your competitors could easily reach your former level with this machine, so you feel compelled to use it too. You also know that, with the knitting skills you've accumulated over the past twenty years, even if everyone has a machine, your knitting speed and quality will still surpass your competitors. But the pleasure of listening to the rain by the window, threading a needle, and slowly passing the time was ultimately crushed by the roar of the machines.
I know this is frustrating, but there's no other way. I can keep my job, but I'll most likely have to give up my old passions. I'm someone who keeps my rationality and emotion quite separate; I can be very rational when I need to handle problems, but sometimes I also show my emotional side. I remember crying my eyes out when I moved out of the rented apartment I'd lived in for a year, reluctant to part with the memories of the past. Saying goodbye to the era of handwritten arithmetic and human brain optimization today is undoubtedly even more cruel than this.
I don't know if any readers have similar feelings, but I think this is the only way things can be.What about people?
While AI continues to advance, I also have some concerns:Are students today more likely to use AI to complete assignments, especially practical labs? Imagine having two choices: one is to spend eight hours painstakingly completing a lab, perhaps without even getting a perfect score; the other is to launch an AI model, costing only a few cents and a few minutes, and have the AI write perfect code. Which would most students choose?The above point leads to a severe deficiency in the engineering skills of many students, including the ability to organize code, build systems, anticipate future potential needs and address them in advance through design, and abstract thinking. Given the ever-increasing power of AI, are these "engineering skills" still necessary? Will these engineering skills gradually become obsolete like the old ability to "proficiently write x86 assembly language," or will they remain valuable forever, like the ability to "understand the entire computer system from software to system to hardware"? If the latter, then it's dangerous—a person with poor engineering skills, when paired with AI, can produce mountains of garbage several times more efficiently than before, thus sowing various seeds of disaster for the system and making the world even more chaotic.In the future society, will power be more important than technology or intelligence?
These questions may require answers from the times themselves.Conclusion
With the development of AI, future societies may tend towards two extremes: communism and cyberpunk 2077. In the former, productivity is greatly liberated, and people's living standards are significantly improved; while in the latter, a few technology companies control most resources, and only a very small number of people can use the most advanced AI and various technologies to achieve near-mechanical ascension, while most people can only use very weak AI. Social mobility will become increasingly difficult: you must first have the most powerful AI to transcend social classes, creating a vicious cycle.
Do you think that if An****pic Corporation forever controls the world's most advanced AI, will future society become communist or like 2077? What do you think?
Therefore, I still believe that cutting-edge intelligence should be made available to everyone in an open and affordable way. I don't trust Anthropic or OpenAI to do this, and I especially don't want Anthropic to master the most advanced artificial intelligence or AGI; to exaggerate, the seriousness of this is no less than allowing Hitler to acquire atomic bomb technology before the Allies. This is why I chose to stay at DeepSeek: by researching powerful, fast, and inclusive artificial intelligence and open-sourcing it, we might be able to pull the world back from that 2077 timeline.
May the world be at peace in the future. May all beauty be blessed.
[1] The “main attention” only includes the MQA attention with head dim = 512, excluding the indexer used to select the top-k important tokens, which was written by other (very skilled) colleagues (and their AI Agent).
Supplementary notes to "I Had to Bury My Talent in Yesterday"
Original post by intlsy's Doghouse, September 15, 2026, 19:50 Zhejiang
This article went viral, garnering over 100,000 views on WeChat, topping the Zhihu hot list, and sparking considerable discussion on X.
This sudden surge in popularity was quite unexpected, but unfortunately, the focus of everyone's attention is not entirely aligned with what I wanted to express.
My original purpose in writing this article was not to express anxiety about unemployment, nor to emphasize the openness and accessibility of DeepSeek and the unpopularity of Anthropic, but rather to say goodbye to the days when I hand-wrote operators. Before the advent of Agents, most of my code was typed out word by word. This seemingly tedious process was actually a pleasure for me: I could calm down and carefully consider every detail, from the organization of modules and the arrangement of functions to the logic of the code and the naming of variables. I also enjoyed racking my brains to study the scheduling and optimization schemes of operators, improving operator performance, and finally beating the generally accepted implementations (such as Flash Attention, the operator written by NVIDIA, or even some tasks that were previously considered "impossible to optimize well," such as token-level sparse attention)—"The excitement I felt when I invented a new technology or saw the performance of my operator improve was no less than that of a speedrunner breaking their own record in a game." However, this joy is now being taken away by AI: for the sake of productivity and operator performance (current AI works faster than me, and future AI will be both fast and good), I must embrace the new things of this new era, researching how to use AI to write operators better. During work hours, I will no longer have the opportunity to quietly and slowly write, tune, and optimize operators. Presumably, one day in the future, "handwritten operators" or even "programming" may become a recreational activity rather than a productive one, just as almost no one uses javelins for hunting now, treating it as a competitive activity. This is tantamount to forcing me to abandon what I once loved and forcibly switch to another direction. Even if the new direction is equally fascinating, this feeling is not pleasant. As the title says: I have to bury my talent in yesterday.
I'm someone who highly values memories and emotions, but unfortunately, memories fade, gradually buried by the fine snow of time. I even wrote another article about this before: 2025 Year-End Summary Part Two: Memories . Therefore, this article is actually a summary and reminiscence of my past. I hope to use this article to freeze those handwritten arithmetic images from the past, to preserve this genuine memory and sincere emotion, for me to cherish ten or twenty years from now. Incidentally, I also hope to use this article to give myself the courage to let go of the past, pick up new tools, and bravely step into a new era.
↑ A quote I really like, excerpted from the comments section of a public WeChat account.
However, it seems that everyone's understanding has deviated significantly from my original intention... The last two sections of the article were originally just some random thoughts: the penultimate section briefly mentioned some social problems brought about by AI, but didn't really analyze or answer them in detail; the last section was some of my personal fragmented thoughts from some time ago, believing that cutting-edge intelligence should still be "provided to everyone in an open and inexpensive way." However, since I am not an expert in history, sociology, or other humanities disciplines, I only briefly explored this part and touched on it (by the way, I'll take a jab at A/, which I've never really liked), and the debate about communism and 2077 may not be accurate. Although the statements "One of the reasons I chose DeepSeek is that I don't want the world to become like 2077" and "May all the beauty be blessed" are true, and these are indeed my ideals, they are not the focus of this article. This paragraph only represents my personal opinion and does not represent the position of any company I work for. Perhaps one day in the future, after I have thought this issue more clearly, I will write another article about it.
As a result, after this article was published, although many people did resonate with my nostalgia for the past, most people focused on the last paragraph. The top-ranked answers on Zhihu were decent (indicating that Zhihu users have good reading comprehension skills), but the WeChat public account comment section was flooded with discussions about Anthropic and DeepSeek's attitudes towards open intelligence, and there were even answers on Xiaohongshu titled "DeepSeek fights back against Anthropic"... I'm quite impressed by the hype-generating abilities of some journalism students, and at the same time, I feel helpless after seeing all this... Everyone should focus more on the article's main theme.
Returning to the main point, it's certain that AI will continue to grow stronger in the future. Even if limited to the current paradigm of Agent + Harness + Long Context + CoT, AI's capabilities will steadily advance as data quality, model depth, and context length continue to expand (scale). Not to mention, embodied intelligence, RSI, and other technologies yet to be discovered are waiting in the future. I'm optimistic about the future capabilities of AI, and relatively optimistic about my place in future society. However, I'm pessimistic about whether people will still be able to dedicate themselves to doing one thing seriously, and I'm also pessimistic about whether I can sustainably pursue one thing as both work and hobby for a long time.
Even so, I will continue to write operators and introduce AI Agents to write operators. First, because "I certainly hope that I will not be revolutionized, but if I have to be revolutionized, I hope that the person who revolutionizes me is myself" (it is better to evolve myself than to be eliminated by others); second, because I still hope that we (or other AI companies that also uphold the concept of openness and sharing) can create the most powerful intelligence before some companies and benefit everyone. I still believe in my original argument about communism and 2077, and I hope that the world can be as far away from 2077 as possible. Hobbies and interests are one thing, but ideals and beliefs are another, so I will continue to explore this field. In the future, I will try to realign my hobbies, strengths and the needs of the times, and try to find my own paradise in the era of AI. After having to bury my talents in yesterday, I still have the light of tomorrow to pursue. After being drenched by this cold rain of the end of the world, I will still use my warmed heart to dispel the gloom and find the blue that gradually appears after the night [1].
[1] The lyrics of songs such as COP’s “The End of the World Singer” were adapted.
Risk warning and disclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.