Hallucinations of a Prodigy

Someone who has founded a startup. 

Someone who once worked at Huawei and later left. 

Someone who served as a chief scientist. Yet he stayed up all night over a single job interview.

This is the most striking, unexpected detail from the recent viral DeepSeek interview controversy. 

Dual screens, live coding, suspected cheating... To be honest, none of these are the real point. 

What truly deserves discussion is: why would a "Huawei Genius Youngster" get so shaken by a seemingly ordinary interview?

1. The Incident Itself Is Very Simple

On July 6, Li Bojie posted a long thread on social media. 

During his remote second-round interview, his eyes habitually glanced at a second screen while he was coding. The interviewer suspected he was cheating, and asked him to prove his innocence. 

He left the meeting immediately. 

The post sparked massive debate. One side criticized DeepSeek for being overly arrogant, while the other side questioned why a so-called genius should be exempted from basic skill assessments. 

Arguing over who was right or wrong here is pointless. 

What matters more is that a person who has seen many ups and downs in his career could have his entire emotional state shattered by a remote interview — losing sleep, writing long posts, giving interviews, and repeatedly explaining himself. 

The answer lies in another line from his later interviews. 

2. What He Was Waiting For Was Not Just An Offer

"DeepSeek, in my mind, is widely recognized across China's tech industry as the absolute peak, the most prestigious hall of the field."

This sentence matters far more than the dual screens or the coding test. 

He submitted his resume, finished the written assessment, waited for three weeks, and followed up five times. Even after securing offers from other companies, he was still holding out for DeepSeek's response. 

This was no ordinary job hunt. He was waiting for a form of identity validation. 

Li Bojie wanted to prove that he belonged at the center of this era, while DeepSeek only needed to judge whether he was qualified for the specific role.

This misalignment of expectations between the two sides forms the core of the conflict. 

The coding task was just the trigger. What truly stung him was that the institution he had looked up to for so long did not give him the reception he had anticipated when they finally met. 

In an interview with 36Kr, the reporter asked Li Bojie: Did you lose sleep because the interviewer rejected you, or because of the gap between expectation and reality? 

Li Bojie answered: It was mostly the sense of disparity.

It was never the offer itself that kept him awake. It was the shattering of his long-held expectations. 

3. Treating Screening As Peer Exchange, And Standard Procedures As Arrogance

During the interview, Li Bojie said: I actually really dislike the media constantly labeling me as a "genius youngster". 

But here is how he evaluated interviews at other companies: 

At MiniMax, Yan Junjie gave him targeted feedback on his work with speech model training. The chief scientist at StepFun had read his papers and pointed out issues with his parameter evaluation methods. At Xiaomi, Luo Fuli discussed team management topics with him. 

Then Li Bojie added that he could learn something valuable from every single interview at those companies — but not at DeepSeek. 

For Li Bojie, the standard of a good interview was not the rigor of the process, nor the role fit, but whether the other party treated him as an equal peer.

An interview is first and foremost a screening process, a two-way selection mechanism, not an academic peer review. But he carried a sense of privilege into the interview, and treated it as a casual discussion between industry peers. 

A person who repeatedly claimed "don't call me a genius youngster" subconsciously expected every interviewer to treat him exactly like one. 

The aura was never on the label. It existed entirely inside his own expectations.

This kind of misalignment is rooted in decades of accumulated inertia across the entire internet industry. 

Over the past 20 years, China's internet industry has developed a set of unspoken rules. The company you work for defines your status, and your job title acts as instant credibility.

Being a "Huawei Genius Youngster" inherently means your competence is pre-recognized, and your credibility is pre-approved. You are treated like a star wherever you go. 

MiniMax treated him that way, StepFun treated him that way, Xiaomi treated him that way. 

It was not that Li Bojie explicitly made special demands. It was that the industry had grown so used to these unspoken rules that anyone at his level would be received with the corresponding level of respect. 

He simply never expected that DeepSeek would break this long-standing unwritten rule. 

4. Why DeepSeek Refuses To Recognize Pre-Existing Auras

Why is DeepSeek so strict about following standardized procedures today? 

The biggest risk in the AI industry is not missing a genius — it is misplacing the wrong person in a critical role.

For a single AI model role today, companies might receive thousands of resumes in a single day. As an organization scales up, its screening criteria have to become more rigid and objective. It simply cannot afford to trust anyone based on reputation alone. 

Huawei Genius Youngster? You still need to do the coding test. Former chief scientist at a startup? You still need to do the coding test. MSRA background? You still need to do the coding test. 

You can call it rigid, mechanical, and lacking in human warmth. But at the very least, the standards are applied equally to everyone. 

Many truly high-performing organizations share one common trait: The corporate culture comes before individual prestige, and rules are never bent for so-called outstanding talents.

Other companies treated him with that peer-to-peer attitude, which was the pre-existing aura working for him. 

DeepSeek was the first company that saw his impressive resume, yet still insisted that he prove his ability by writing the code in real time. 

If the story ended there, it would have remained nothing more than a single ordinary interview. 

What really resonated widely with the public was the broader era shift unfolding beneath the surface. 

5. "Outstandingness" Is Turning Into Real-Time Calculation

Over the past two decades, the internet industry had a very simple way to evaluate people: which company you worked for, what your title was, and what projects you had delivered. Tencent, Alibaba, ByteDance, Huawei — a strong resume was all you needed to get your foot in the door. 

The underlying assumption of that system was that excellence could be accumulated. Your degree, past experience, and job title could carry you for many years. 

In the AI era, this assumption has completely collapsed. The speed of knowledge obsolescence now outpaces the speed of experience accumulation.

The most important methodology from last year is already obsolete this year. The architectural design skills you were proud of yesterday can be fully replicated by a new graduate using AI agents in half a day. 

AI has turned personal excellence into a real-time calculation.

This leads to a strange paradox: the more outstanding a person is, the more anxious they tend to feel. The more glorious your past was, the greater the sense of disparity you feel when your identity starts to depreciate. 

More and more top talents today are constantly switching jobs, launching startups, and pushing their way into the most cutting-edge AI companies. This cannot be explained by money alone. What they are truly afraid of is falling behind the times. 

Li Bojie waited for three weeks, followed up five times, and insisted on finishing the entire interview process with DeepSeek. 

His real underlying anxiety was: if DeepSeek rejected him, would that mean he had already been left behind by the era? 

A final note beyond the article:

In the past, your aura could define who you were. Today, your aura only proves what you have already achieved in the past. 

Li Bojie is just the first person who spoke out loud about this widespread anxiety. 

Very soon, every programmer, every product manager, every researcher, and every entrepreneur will have to go through this exact same kind of interview.

The interviewer will be the AI era itself. 

It will ask every single person the same question, every single day: 

You were outstanding yesterday. What about today?

This article is from the WeChat public account "Beyond the Layout", written by Huahua, and published by 36Kr with authorization.