When Gen Z Meets AI Interviewers: Finding Long-Term Careers in an Era of Rapid Change | 5Y Capital's Tavern x Niuke Network's Xiangyu Ye [Podcast]
Only in change do new opportunities arise.
For this episode of 5Y Pub, we invited Xiangyu Ye, founder and CEO of Nowcoder. Nowcoder is a campus recruitment platform that has been operating in this space for ten years. Data shows that in 2024, the number of university graduates in China reached 11.79 million, setting yet another record high. Under the combined pressures of economic, technological, and social factors, young people's career choices and corporate hiring practices are both undergoing significant shifts.
In this rapidly changing environment, Nowcoder has continually adjusted its business to adapt to new trends. Xiangyu Ye shared his observations on how young people's attitudes toward work have changed, as well as what his team has learned from developing AI-powered interviews.
Finding What You Want to Do Long-Term
5Y Pub: Let's start with an introduction — tell us about yourself and how Nowcoder came to be.
Xiangyu Ye: Hi everyone, I'm Xiangyu Ye, founder of Nowcoder. We started in September 2014, so we're coming up on ten years. Before Nowcoder, I was an engineer. After graduating from Zhejiang University in 2007, I joined Google as a software engineer. After Google, I had a brief stint in game startup entrepreneurship. When that first venture failed, I joined a gaming company, then later moved to Renren, and from there jumped to Alipay. I left in 2014 to start Nowcoder, and I've been at it ever since.
The journey with Nowcoder has changed quite a bit. We started as a consumer-facing community for coding practice and training, then moved into the online written exam market, pivoted to B2B, and focused on campus recruitment services. After exams, we added video interviews, and now we're doing AI interviews. We've also built out employer branding, competitions, and other services. Ten years, gone in a flash.
5Y Pub: We've seen the recruitment reports Nowcoder puts out. You've been in this space for nearly a decade now — what direct changes have you noticed in graduates compared to earlier years?
Xiangyu Ye: The changes are quite noticeable, especially in mindset. When we graduated, the internet was still booming. Domestic companies were growing rapidly, particularly in information technology and the internet sector. There weren't as many graduates back then, and market demand was strong. When we started Nowcoder in 2014, we clearly felt robust hiring demand in these industries — that's why we built services around written exams, video interviews, and take-home assignments.
By 2021, the internet was reaching saturation. The mobile wave had matured, and companies started cutting costs and improving efficiency. Graduate mindsets shifted accordingly. Today, our user growth has slowed significantly compared to before. What are these users doing instead? Our research found two main things: first, a huge number are pursuing graduate school; second, some are preparing for civil service exams earlier than before. And this isn't something they decide at graduation — from day one of university, they're already in a mindset of preparing for the future. I wouldn't say this change is good or bad; it's probably tied to broader societal development. Back then, we all wanted to earn more in emerging industries, work overtime, buy cars and apartments — what people might call being a "corporate beast" today. Now students seem more focused on their quality of life; they'll quit if they're unhappy. Looking back, they might think we were all "socially conditioned," but at the time that just felt like how things were supposed to be. These are all subtle societal influences.
5Y Pub: Young people today do face many challenges. Having worked in campus recruitment for so many years, what advice would you give them about making choices?
Xiangyu Ye: Though my personal advice may not be universally correct. Looking back at my own path — I worked at a foreign company, tried a small startup, then joined a domestic gaming company and later a major internet firm. At first I didn't really know what I wanted; I just knew I wanted to do IT, to be a software engineer. I remember some classmates worrying: if you take a wrong first step, won't the deviation become enormous? Personally, I don't think so. The choice you make today might look different from what you expected in three to five years, and you might still want to try something else. So I'd suggest keeping an open mind early on. We suffer because we fear making wrong choices. But if you understand that there's no right or wrong, only fit or misfit, choosing becomes easier. You might hit obstacles later — well, then pivot.
I only truly started my entrepreneurial journey after seven years of working. Those first seven years, I didn't have complete clarity either, but I did gain valuable workplace experience through serious effort. For graduating students, I think the most important thing is not to let the anxiety-inducing noise about employment affect your motivation and mindset. Because of recommendation algorithms, short-video apps will feed your anxious emotions, constantly pushing content that amplifies anxiety — you can put that down.
5Y Pub: In your early working years, were you set on eventually starting what you're doing now, or did you discover it through exploration?
Xiangyu Ye: I had a vague idea. When I first graduated, my roommates and I discussed our futures, and I said I'd probably start a company someday. That concept had formed during college, but I had no idea how to become that kind of person. So what to do? I joined a company first, and realized I wasn't mature enough. After my gaming startup failed, I wanted to see how others succeeded at games, so I joined a gaming company. Later I moved to the internet industry — by then I was five years behind my classmates who'd gone straight into internet companies in 2007. In hindsight, my choices were quite blind: wherever society's winds blew, I went. Only later did I realize this wouldn't work, and eventually found something I wanted to commit to long-term, so I started my own company.
5Y Pub: How did you discover that this was what you wanted to do long-term?
Xiangyu Ye: It was fuzzy at first. We simply wanted to do IT training, to replicate our own job-hunting experiences, so we built Nowcoder as a coding practice platform. Our monetization path was training — teaching people to code, design systems, and so on. That commercialization path ultimately didn't succeed. Then we discovered demand for campus recruitment exams, so we poured energy into that business. In short, the path you end up taking always deviates somewhat from your initial vision. No one maps out every step with perfect clarity from the start; there's just a vague destination in the distance.
5Y Pub: Mapping things out too clearly can sometimes become a limitation.
Xiangyu Ye: Exactly. Many people plan everything out, then after ten years of effort realize this isn't the life they wanted. You feel like you've lived those ten years for others, which also becomes a negation of your past decade. So I believe the core is to keep an open mindset and do things you genuinely like and can believe in.
5Y Pub: AI has accelerated change across industries. It's hard for young people to see where industries are headed when making plans. In this situation, how can they be more accurate?
Xiangyu Ye: Accuracy is relative — there's no most accurate, only more suitable. The only way is to try things quickly, don't work in isolation. I did something interesting back in college: I sold computers, helped people assemble them. Because I didn't understand the field yet, I thought if I wanted to do IT, I needed to know about computers, hardware and software. Later I realized it wasn't for me — it was fundamentally a sales business, just selling computers rather than something truly IT-related.
I still encourage people to try things that seem difficult today. Many problems you avoid now will still confront you in three to five years. Looking back at myself — why did I job-hop so much in those first seven years? Because I encountered many problems in the workplace that made me uncomfortable, unhappy. But why have I persisted with entrepreneurship for so long? Because I found those same reasons for job-hopping reappearing in entrepreneurship, and finally realized I just had to push through them myself — I had to face these difficulties eventually. I always encourage young people to keep an open mindset, but stay proactive in doing things, because problems you encounter today will definitely reappear. There are no truly new problems; life has just a few core ones.
AI-Driven Changes and New Opportunities
5Y Pub: What's been the hardest thing since you started Nowcoder?
Xiangyu Ye: Two phases. The first was around 2016. In the beginning, everyone was excited and idealistic — it wasn't that hard. But in the two years after starting, our education and training business failed to commercialize successfully, and we couldn't raise funding. That was painful, with considerable mental stress — I'd often wake up to find hair on my pillow. But we got lucky; the capital market recovered, and we raised funding in 2016 that kept the company going. After that funding, we urgently pushed commercialization — that was the first phase.
The second phase is the past year or two. Around 2021 was the peak of internet growth. Because things had gone relatively smoothly before, the company expanded hiring. But then 2022 and 2023 came, and the market underperformed expectations. We had to lay people off, adjust the business, and didn't know what the future held or if opportunities still existed. We tried many things that failed. Fortunately, we got lucky again — AI arrived. Large model technology emerged in late 2022, and by 2023 there was broad consensus. We moved quickly to follow. I think university students job-hunting may encounter similar phases — confusion early on, then needing to pause and find direction. When a new opportunity appears, if you've prepared more than others, you might move faster.
5Y Pub: What sustained you before that opportunity appeared?
Xiangyu Ye: Always believing opportunity would come — an optimistic, positive attitude is crucial. You have to believe that what you're working hard on today will definitely be useful later.
5Y Pub: You've also launched AI interview services. When did you start, and what was the catalyst?
Xiangyu Ye: Around 2023. We didn't start with AI interviews — we started with AI-assisted interviews. The plan was to put an assistant alongside video interviews, prompting interviewers on what questions to ask, what not to say, and so on. But that product failed, because we found that for an AI assistant to guide different, professional interviewers with advice — that's still difficult even today, and AI capabilities were weaker then. We co-developed with one client and ran over 10,000 interviews, but it still failed.
Today we're combining our AI-assisted interview experience with AI interview services, truly enabling AI to interview candidates. The capabilities we built during AI-assisted interviews — understanding interview context in real-time, understanding what interviewers want to assess, how to ask questions — all proved useful. The technologies and scenarios we validated could largely transfer to AI interview products. It goes back to what I said: when you're lost, believe that what you're doing has value. With AI interviews too — AI needs to understand what questions a position requires, and give evaluations after interviews. Before, AI might help interviewers draft evaluations for them to edit; now AI gives evaluations directly.
We started with foreign-language AI interviews first, because that's a hard need — say, interviewing for a position requiring Vietnamese. After building that and seeing decent results, clients asked for more, so we moved to general competency interviews, then professional competency interviews, and gradually did well. This product has no interviewer present; in the recruitment process, AI replaces the first half — HR interviews and earlier assessments, the high-volume initial screening. This improves overall hiring efficiency for companies, and when efficiency improves, candidates also have a better experience.
If the product matures, my vision is that it could truly replace at least first-round interviewers. Achieving that would create significant social value. Of course, AI interviews can only replace part of the process — job searching is ultimately about mutual assessment of fit, and people still need to meet face-to-face. AI interviews are primarily an efficiency tool.
5Y Pub: To reach that more ideal outcome, what are the main bottlenecks now?
Xiangyu Ye: First, the underlying AI capabilities still need improvement. When people first use large models, they seem impressive. But in real commercial applications, you'll find there are still 10% of cases the model can't handle well. For serious business scenarios, companies can't accept AI making mistakes. If I ask you a question today and want you to respond on behalf of the company, even if you answer nine correctly but mess up the last one, the company can't accept that. So the foundation needs to strengthen, and we're hoping AI takes another leap forward — more instruction-following, more controllable.
Additionally, while companies know AI is the inevitable trend and are trying various AI tools, there's still a phase before real adoption. After using AI interviews or AI training products and seeing efficiency gains, they'll believe in AI. At first it's intellectual belief; after using it, they become true advocates. When society's acceptance level rises quickly, that's when AI interviews truly succeed. Compared to this vision, I think Nowcoder and our peers are still in early stages — we all need to keep working, always believing tomorrow will be better.
5Y Pub: For your company, developing AI products is also something relatively new. What adjustments have you made internally to seize the AI opportunity?
Xiangyu Ye: Once there's a clear goal, people are actually very happy, and their motivation to work becomes much stronger than before. We started using AI technology, felt this was probably viable, and so invested in building it.
5Y Pub: There was a survey of many engineers where 90% of respondents said they worried about being replaced by AI, or that finding jobs would become harder. AI may make hiring more efficient for companies, but for technical workers, what challenges or changes might AI bring?
Xiangyu Ye: In absolute numbers, I think there is impact. Because the previous work model was a strong lead managing several junior people, forming a team to deliver a project. After AI emerged, especially for junior roles, the impact is quite significant. A lead who previously managed three to five people might now manage two people plus AI, with similar productivity. Of course, for some people, junior level is just a brief phase — they use AI as a tool to strengthen themselves, and their career continues advancing. Those affected more are people doing repetitive work.
AI also brings many benefits. Some work previously couldn't be done due to insufficient people or capabilities — with AI, new positions emerge. It depends on whether you can seize them. When we were in school, there was no e-commerce major, no product management courses. Today's AI wave is the same — new things will grow. If you can catch the opportunity and work hard enough, there is future opportunity. For us, we welcome change. As they say, running water doesn't rot — only in change do new opportunities arise. For a company or a society, if everything becomes fixed, that's actually not good.
Interactive Giveaway
What are your takeaways after listening to the podcast? Share your thoughts in the Xiaoyuzhou Podcast comments section, and we'll select 2 featured comments to receive a gift prepared by 5Y Capital :)
5Y Capital seeks out, supports, and inspires entrepreneurs who are alone in their journey, providing support from the spiritual to all operational aspects. We believe that if the "crazy you" in others' eyes begins to be believed in, the world will become refreshingly different.
BEIJING · SHANGHAI · SHENZHEN · HONG KONG
