In the age of artificial intelligence, "AI Pioneers" are gradually emerging as a highly sought-after new talent category, where portfolios are replacing traditional resumes and real-world problems have become the new measure of capability.
After a short intermission, the host returned to the stage and announced encouraging news: Zhang Qingqing from Sichuan University's Ammonia Scheduling Team had just received an on-the-spot offer from Yuntu Holdings.
Applause erupted. Yang Li, General Manager of Yuntu Holdings' Human Resources Center, who took the stage next, said candidly: "We came here today specifically to recruit talent capable of implementing AI in real business scenarios. After watching the project presentations, we were worried other companies might poach this candidate, so we extended the offer immediately."
This sense of urgency was no exaggeration. Just ten minutes earlier, during the tea break, a quiet battle for talent had already begun. One person whispered, "I've already claimed this person." Another quickly responded, "No problem—I can still steal them."
This wasn't a scene from a TV drama, but a true account from the semifinals of the "AI Pioneer Future Talent Competition" held in Chengdu on September 20, 2024.
Within just two days, four regional finals unfolded consecutively in Guangzhou, Beijing, Chengdu, and Shanghai, with 40 university student teams taking turns presenting their work. Before the competition even concluded, companies including XPeng, Huatu, Hisense, Yuntu, Raytrix Micro-Nano, and Feishu had already issued eight internship offers, campus recruitment invitations, and final interview opportunities on site.
Why are companies rushing so urgently? The answer likely goes beyond simply having "exceptionally talented students this year." As AI permeates corporate operations, hiring standards themselves are undergoing a fundamental transformation.
A New Type of Talent Is Being Recruited
In the past, when recruiting young graduates, companies relied on clear criteria: evaluating academic background, major, and internship experience, followed by written tests and interviews to assess actual skills.
But at this competition, companies have completely reinvented how they identify talent.
There are no written exams or classic interview questions. Each team gets only seven minutes to present tangible results they’ve actually built: What problem did they identify? Why was it worth solving? How did they apply AI? And what value did they ultimately create?
The work itself becomes the strongest testimony. In the Guangzhou round, the "Autumn Sunshine Pipeline" team integrated AI into the entire short-video production chain; their two works topped Douyin’s trending charts, accumulating over 970 million total views while cutting production time from 15 days down to just 7.
In Beijing, the Hisense team introduced AI into their after-sales system, automating a large volume of repetitive service tickets, allowing engineers to focus on complex issues requiring human expertise—potentially saving hundreds of millions in costs. In Chengdu, students from Sichuan University developed a digital employee for ammonia plant scheduling, reducing a task that previously took 2–4 hours down to just 12 minutes.
In Shanghai, a PhD team from Beihang University presented a low-altitude radar solution, while a research group from Zhejiang University used AI to optimize 3D printing decisions. These outcomes, once confined to labs or academic papers, are now directly addressing real corporate challenges.
HR face-to-face booths at the competition venue
As a result, company judges are increasingly transforming into recruiters.
In Guangzhou, XPeng Motors issued internship offers on the spot and invited contestants to “take a ride in a Robotaxi tomorrow.” In Beijing, Huatu identified two students who developed a product helping recent graduates find career direction and immediately granted them access to the presidential final interview—a channel reserved for fewer than ten candidates each year. In Chengdu, Yuntu Holdings issued an offer mid-event before the competition ended. In Shanghai, after hearing the Beihang PhD team’s low-altitude radar proposal, Raytrix Micro-Nano immediately went on stage to extend an internship offer.
Sun Peng, PhD from Haisen Pharmaceuticals, commented: "Once the work is shown, success naturally follows."
The talent being competed for is no longer merely traditional AI specialists, but a different kind of individual: practitioners who excel at identifying real problems, understanding business logic, knowing when and how to deploy AI, and turning ideas into tangible outcomes.
Redefining the Talent Profile in the AI Era
This shift is particularly evident across the four regional rounds.
When people think of AI talent, images of algorithms, models, and computer science backgrounds typically come to mind. But the students stepping onto the stage this time represent far more diverse backgrounds than expected.
Among them are law graduates, medical students, researchers—and even individuals who previously knew almost nothing about programming.
In one team from the western region, a law master’s student from Sichuan University and a medical student from Shanghai Jiao Tong University—both distant from programming—taught themselves to build an AI-powered product selection system for cross-border e-commerce.
AI has lowered the technical barrier to "building things," blurring professional boundaries. Law students can now develop products, medical students can write code, and business staff can build their own agents.
However, lower barriers do not mean talent is no longer scarce. On the contrary, as more and more people gain access to AI tools, the real dividing line shifts from "knowing how to use the tool" to "being able to solve real problems."
Can you identify key pain points within operational complexity? Can you determine which tasks should be delegated to AI and which require human judgment? Can you integrate models, data, tools, and workflows so that AI doesn’t just generate answers but completes actual tasks? And when AI participates in business decisions, do you clearly understand its limitations?
One outbound team from the western region established a principle: "Evidence precedes generation; humans retain final approval rights." The Shanghai 3D printing team required every AI conclusion to include sources, enforcing a rule: "If uncertain, refuse to answer—no guessing allowed."
These abilities cannot be fully demonstrated through resumes or standard interview questions. Therefore, as the profile of future talent evolves, screening methods must evolve as well.
Over 100 Companies Brought Real Business Challenges Into the Arena
This is where Feishu’s approach to organizing the competition stands out.
Rather than setting standardized AI test questions to compare problem-solving skills, the event flipped the model—inviting companies themselves to define the challenges.
In June this year, over 4,300 teams registered to participate, while more than 100 enterprises brought their real, ongoing business problems into the competition.
Competition venue
The challenges spanned multiple sectors including manufacturing scheduling, automotive, retail, consumer goods, education, scientific research, and content creation.
Students were no longer solving textbook problems with predefined answers, but tackling genuine, often unresolved challenges faced daily by real businesses.
This shift created a profound impact: companies are no longer merely selecting talent after the competition ends, but actively participating in talent development and identification from the very beginning—from problem definition onward.
Companies pose real problems, students use AI to propose real solutions, and companies evaluate them against actual business standards: Is the solution feasible? Can this person perform the job?
Thus, an interesting scene unfolded at the venue: right after a contestant’s seven-minute pitch, their resume would instantly appear on the big screen; if a company showed interest, they could scan a QR code to contact the candidate immediately.
From problem-setting and problem-solving to talent acquisition, the entire process is tightly connected. This integration may be even more significant than the competition itself. If the AI era demands people who solve real problems, then the best exam hall is reality itself.
From AI Pioneers to Future Talent
This isn’t the first time Feishu has turned to real-world business scenarios to explore AI applications.
Over the past few years, Feishu has consistently searched for "AI Pioneers"—those who were among the first to introduce AI into enterprises and genuinely transform workflows.
They might be technical staff, or they could come from sales, operations, manufacturing, R&D, or frontline roles. Their common trait isn’t mastery of model parameters, but rather the ability to pinpoint specific problems and drive AI adoption into actual work processes.
More and more cases show that when AI enters a company, it changes not only tools but also job design, collaboration methods, and the very definition of "excellent talent."
This year, the AI Pioneer initiative extends into universities, shifting focus from "who is currently using AI to transform work" to exploring "who will become the next generation of AI practitioners." At its core lies the same question: When AI becomes a fundamental workplace skill, what kind of people do we need?
The competition is delivering increasingly concrete answers: Not everyone needs to become an algorithm engineer, and knowing how to write prompts doesn’t automatically make someone an AI talent. What will be increasingly scarce in the future are individuals who understand business, exercise sound judgment, create value, and can orchestrate AI to accomplish real tasks.
For this reason, Feishu hasn’t designed a set of standardized AI answers for young people. Instead, together with over 100 partner companies, it has placed real-world problems directly in front of them.
On-site at the four regional finals
Careers Begin Even Before the Competition Ends
The four regional rounds ultimately selected 12 finalist teams: Autumn Sunshine Pipeline, Fan Dage and Fang Wuji, It’s My Turn Now, TenderTrace, Love Vibe Coding Most, Walnut World, Ammonia Scheduling, Andon Evergreen Tree, Guoran Smart Chain, Jianwei Intelligence Explorer, XwX, and DOUBLE AUNT (listed in no particular order). They will advance to the national finals on October 20.
One week after the competition, a unified talent pool comprising candidates from the 108 participating companies will open for access.
Looking back at that tense intermission in Chengdu, Yuntu Holdings’ urgency in extending an offer now makes perfect sense.
What companies see is not just a student who performed well in a competition, but their problem-solving strategy when facing unfamiliar challenges, their grasp of business context, their ability to leverage AI effectively, their execution in turning ideas into working prototypes, and their sense of responsibility toward outcomes.
In the past, these qualities would only emerge gradually after hiring; today, a single project can prove them all upfront. As AI reshapes the way we work, corporate hiring practices are evolving accordingly.
In the AI era, AI Pioneers are becoming the new premium talent, portfolios are becoming the new resumes, and real problems are becoming the new exam questions.