2026 Workplace Insights: Some made a fortune tracking the market with 'Crayfish', while others were overwhelmed by AI-generated content

With rapid technological iteration recently, professionals are feeling a bit anxious about the constant updates in AI.

Anxiety, like the air, permeates the workplace of 2026. It is no longer an abstract term but rather the constant notifications popping up on every worker's phone, and the lurking concern in the dead of night as they stare at their screens, thinking, 'Am I about to be replaced?'

This feeling is akin to standing next to a never-stopping high-speed train. In the break room during lunch, the conversation has shifted from 'What’s for lunch?' to 'Which AI tool do you use to write weekly reports?' and 'Have you tried that new code assistant?'Various learning materials and videos are circulating across everyone’s computers, while news pop-ups constantly announce another AI company releasing its latest version.

This anxiety is not without reason.

The latest data from Zhaopin shows that 78.2% of employees use AI tools at work at least once a week, with penetration rates of 'digital employees' in roles such as finance, customer service, and administration exceeding 60%.From intelligent investment advisory in the financial sector to AI teaching assistants in education, AI is deeply embedded in workflows as a 'digital colleague,' capable of handling standardized tasks around the clock while capturing details in data flows that humans might overlook.

Perhaps this is the most authentic footnote to today's workplace. AI is not an opponent here to replace us but a mirror forcing us to rethink the meaning of work. When it takes over standardized tasks like data organization, document writing, and basic screening, humans can finally focus their energy on areas requiring empathy, creativity, and decision-making.

So, what does the situation look like on the frontlines of this transformation?

Recently, we interviewed several professionals who heavily rely on AI: among them were a meticulous owner of a medical aesthetics clinic, a university instructor embracing change, and a financial practitioner navigating the frontlines of risk control.

In this 'human-machine collaboration' revolution, there are no absolute winners or losers—only working individuals seeking their own balance between dependence and transcendence.

Xiaomei: My AI colleague is the 'two-faced assistant' in the medical aesthetics clinic.

At eight in the morning, I pushed open the door of the medical aesthetics clinic. On the front desk's computer screen, the AI assistant had already automatically generated yesterday's financial report, with income details, material costs, and customer prepayments clearly categorized. This is the first 'colleague' I work with every day; it has no physical form but quietly helps me manage the most tedious accounts like an indefatigable accountant.

Its greatest strength lies in organizing customer information. In the past, I always had to spend half an hour extracting customer preferences from consultation records. Now, by simply inputting keywords such as 'anti-aging treatments' or 'sensitive skin,' it consolidates scattered information from WeChat and appointment systems into a clear profile, even marking customers' birthdays and their last treatment dates. Marketing content is another area where it excels: poster slogans for holiday promotions, promotional messages for social media posts, and short video scripts—it always provides three to five versions tailored to current trends. Last week, its slogan for the 'Spring Skin Renewal' event, 'Plant spring into your skin,' unexpectedly garnered many likes from clients.

However, when it comes to core business operations, this 'all-around assistant' reveals its limitations.

Previously, I reached out to every company I could find related to 'AI + healthcare' and discovered that most domestic firms rely on large underlying models developed domestically. At present, these technical personnel lack understanding of the specialized knowledge (know-how) in our industry and need to collaborate extensively with experts and businesses within the sector. It will still take considerable time before they can truly be integrated into the industry as tools capable of replacing employees.

I believe there are currently no companies that can genuinely provide AI-based solutions specifically tailored for the medical aesthetics industry.

A few days ago, a client interested in rhinoplasty brought me materials she found online and asked, 'Given my nasal bridge foundation, should I opt for expanded polytetrafluoroethylene or silicone? Will the surgery affect my breathing?' When I entered the question into the AI, its response, though filled with medical terminology, was as rigid as a textbook. It didn't consider her nasal skin thickness or mention the minimally invasive techniques we commonly use at our clinic. Ultimately, the customized plan was devised by me and the director based on her specific conditions, which finally put her mind at ease.

Another time, a long-term client felt anxious during her postoperative recovery period and repeatedly asked me, 'Why is the swelling reduction slower than expected?' The AI-generated reply stated, 'Individual differences exist in postoperative recovery; patience is advised.' However, looking at the photos she sent, I sensed her unease. What she truly needed wasn't a cold, medical explanation but reassurance: 'I understand your concerns. Last week, a client with a similar situation started looking much more natural. Let me send you her recovery pictures for reference.'

Some say AI will make the medical aesthetics industry feel cold, but for me, it acts more like a mirror reflecting my own value.Judgments that require consideration of individual differences, empathetic communication, and decisions based on clinical experience will always remain my irreplaceable domain.

Investment manager Dodo: When 'Crayfish' became my U.S. stock trader

In the investment community of 2026, if someone hasn't heard of 'raising shrimp,' they might really be on the verge of being phased out by the market.

Dudu, the investment manager, is a pioneer in this wave. As a professional, he faces the same dilemma as all fund managers: information overload. The global market operates 24/7, with macro policies, individual stock earnings reports, sudden geopolitical events... the human brain simply cannot process such a massive amount of unstructured data in real time.

I used to spend four hours a day reading research reports and news, but now, all that work has been handed over to my 'little shrimp.' What Dudu refers to as 'little shrimp' is OpenClaw, the recently viral open-source AI framework in financial circles.

Unlike most peers who only use AI to write weekly reports, Dudu's application of OpenClaw has entered the deep end—assisting in decision-making.

He deployed this system on his cloud server and connected it to FinnHub's real-time market data API. For Dudu, OpenClaw is not just a chatbot; it's a 'super analyst' capable of autonomously breaking down tasks, searching the internet, and conducting logical reasoning.

I set its role as an aggressive tech stock hunter. Dudu showed us his workflow: every morning, OpenClaw automatically captures overnight movements in the US stock market, analyzes the trend of Nasdaq index futures, and generates a briefing based on the latest tech news.

But this is just the foundation. Dudu’s core strategy is to leverage OpenClaw’s hierarchical scheduling capability.He lets the 'fast thinking' model handle data retrieval and organization, such as looking up NVIDIA's Q1 earnings data for the past five years, while delegating in-depth analysis and strategy formulation to the 'deep thinking' model.

The most memorable live application occurred in February this year. At that time, a company released seemingly impressive earnings, and its stock price surged during after-hours trading.

Based on past experience, I would have tended to chase the uptrend. Dudu recalled, but my 'little shrimp,' after analyzing the transcript of the earnings call, keenly picked up on management's vague wording about next quarter's capital expenditures, combined it with negative rumors from the supply chain, and issued a 'high risk, recommend observation' red alert.

Dudu chose to trust the AI's judgment, not only refraining from adding more positions but also reducing some of his profitable holdings when the market surged at the opening. In the following two weeks, the stock plummeted by 20%.

“At that moment, I realized it had helped me avoid a massive emotional trap.” Dudu admitted that AI lacks human greed and fear, enabling it to execute trading discipline more objectively.

Of course, Dudu didn’t completely “sit back and do nothing.” He emphasized that OpenClaw is more like an indefatigable junior trader, responsible for monitoring the market, reviewing trades, and preliminary screening, while the final decision-making authority still rests in his hands.

University Lecturer Chen Mo: My AI Jury Roundtable and “Counter-Surveillance” Intuition

Chen Mo’s computer always runs several different AI agents simultaneously. Some are academic types proficient in literature review, one is a powerful search type, others may be responsible for creativity, and there’s even one dedicated to writing code.

To put it simply, Chen Mo’s research method no longer involves him sitting alone in front of a blank document racking his brain as he used to. Instead, he initiates an “AI roundtable meeting” — using the same prompts to seek the optimal solution.

“Now when I conduct research, it feels like I’m running a horse race.” Chen Mo pointed to the different dialogue windows lined up on his screen. When he needs to outline a new research framework, he inputs the same detailed prompts — including background, core hypotheses, and expected goals — into entirely different intelligent agents.

Chen Mo’s workflow has become about drawing on the strengths of many. He combines the theoretical framework generated by Agent A with the innovative methodology proposed by Agent B, then uses Agent C’s coding capabilities to validate data, and finally lets Agent D conduct a round of “simulated peer review.”

“It’s like forming a virtual team composed of different experts,” Chen Mo described. “What used to take several seminars to finalize can now be completed through this ‘human-AI competition’ and ‘human-AI collaboration’ in just one afternoon. The end result is often a ‘well-rounded expert’ synthesizing the strengths of all parties involved.”

However, this deep reliance on technology has brought about a side effect: Chen Mo’s sensitivity to student assignments has reached unprecedented levels.

"In the past, when I reviewed students' theses, I focused on whether their arguments were novel and their reasoning sufficient. Now, I can immediately detect the 'AI flavor' at first glance." Chen Mo said with a wry smile. This intuition doesn't come from any detection software but stems from his deep understanding of human-machine interaction characteristics as a heavy user.

He can easily tell whether students are in control of AI or being fed by it.

"Many students simply throw their topics to AI and then copy-paste the generated answers. Those bland and soulless platitudes, that typical 'first, second, finally' three-part structure, and even certain overly polite transitional phrases unique to AI—I can spot them right away." Chen Mo explained, "Because I interact with these models daily, I know their comfort zones and where their logical pitfalls lie."

In Chen Mo's view, AI hasn't made academics easier; instead, it has raised the bar."In the past, effort alone could produce a good thesis, but now you need 'discernment' and 'integrative ability.' The outstanding students now know how to use different AI tools to spark ideas like I do, and then connect the fragments with their own thoughts. The lazy ones, however, are just producing digital junk."

Conclusion

AI is reshaping workplace productivity, but the 'heart' of the workplace still belongs to humans.

Whether it’s cyber risk-control investing in financial circles or human-machine debates in education, these cases reveal a harsh truth: technology will淘汰 the lazy but reward those who make good use of tools.

In this new era of human-machine collaboration, the real winners in the workplace aren’t those who run the fastest but those who know best how to simplify. Leave the tedious tasks to AI and keep the core work for yourself. When you stop trying to compete with machines on computational power and start leveraging your decision-making and empathy, the journey toward self-consistency in the workplace truly begins.

12K Views