Market Data 
This as-told-to essay is based on a conversation with Vivi Mengjie Xiao, an AI product manager and content creator on RedNote, China’s social media platform. The conversation has been edited for length and clarity.
I’m an AI product manager in China, and earlier this year, my CEO asked me to explore how AI could go beyond cost-cutting, to drive innovation.
Outside of work, I’m a content creator on RedNote, where I share AI tools, workflows, and insights with over 45,000 followers.
I used to spend about four hours a day gathering AI industry news: reading posts on X, newsletters, blog posts, and translating English sources into Chinese. I thought: “Can I automate this?” If AI can handle information gathering, what else can it do? If I’m doing something repetitive, I should automate it.
Each agent was born from a real problem I was experiencing. I created six AI employees, and they’re split between work and personal life.
My foray into OpenClaw
At first, I set up only one “lobster” — a nickname Chinese netizens use for deploying an OpenClaw agent — and tried to make it do everything.
I wanted it to manage my calendar, schedule, to-do list, and monitor my work. I get distracted easily, so I wanted it to help me focus on what I needed to do in the moment and help connect the main thread of my work.
I kept stuffing other tasks into it as well, such as assigning it to manage my finances.
The result of putting all of that on one lobster was that its context became long and messy. It basically became ADHD like me: jumping from one thing to another without helping me focus. It was running three work streams at once. That wasn’t going to work, so I split tasks up and assigned them to different lobsters.
Over time, the six AI employees naturally organized into personal vs. work, and within each category, into clear roles.
I have three work agents: the administrative assistant, the researcher, and the chief of staff. The chief of staff simulates my boss’s communication style, and I use it to practice and polish presentations. For personal agents, I have a life coach, a content and expression assistant, and a finance assistant.
It felt like building a real team. It makes sense — you don’t hire six people on day one. You start with one, and as the workload grows, you specialize.
The compound effect of having them connected surprised me. The life coach can read conversations from all five other agents. I use the life coach agent to help me journal daily, and now 70% of my journaling is automated. The agent knows everything — what I researched, what I invested in, and what I stressed about in my presentation rehearsal.
I’m more productive, but also more tired
About 60% to 70% of my daily operational work is handled by these AI agents, including information gathering, research, and content distribution.
However, my workday hasn’t gotten shorter. I’ve shifted from doing “grunt work” to doing more creative, strategic, and high-leverage work. The AI employees freed up capacity for significantly more output.
I’m more productive by any conventional metric. I publish podcast episodes daily, monitor financials in real time, run a knowledge management system, and create content for RedNote and X — all while working full-time.
Honestly, I’m also more tired. This is a paradox I’ve been thinking about: When your efficiency goes up, you don’t work less. You just attempt more.
My bedtime has shifted from midnight to 2 a.m. because there’s always one more thing I want to do, or one more agent I could spin up to solve a new problem.
The future of work
We’re witnessing a fundamental shift in what “work” means.
The Industrial Revolution standardized physical labor. The information revolution standardized knowledge work. And now, AI is standardizing execution work — the “how” of getting things done.
This means the premium is shifting from execution ability to three things: taste and judgment, ability to direct AI, and emotional intelligence.
The future of work is “one-person studios,” solo creators and operators who leverage AI to produce at team-level scale. For companies, the question becomes: do you need 10 junior analysts, or one senior thinker with 10 AI agents?
This isn’t about replacing humans. It’s about liberating humans to do more human work. The parts AI takes away were never the parts that made work meaningful. The parts that remain — creativity, judgment, connection, purpose — are what make us human.
Building a team of six AI agents feels like going from being a solo freelancer to being the CEO of a small company, except your team never sleeps, never complains, and works for the cost of API subscriptions.
I’ve become a more structured thinker, a clearer communicator, and a more ambitious creator. I now think in terms of “which agent should handle this?” for almost every task. AI expanded my sense of what’s possible for one person to build.
Do you have a story to share about tech in China? Contact this reporter at cmlee@businessinsider.com.
Read the full article here
