This as-told-to essay is based on a conversation with Xiankun Wu, the CEO and cofounder of Kuse, an AI-powered visual workspace headquartered in San Mateo, California. The conversation has been edited for length and clarity.
In the near future, you may have a workplace with AI employees, much like non-player characters.
AI can handle repetitive work, allowing people to spend more time on creative work.
Early this year, things changed because of OpenClaw. AI became more autonomous and customized. Because it understands the company and employees, the AI employees OpenClaw creates can assign tasks, even to humans.
When you see an AI assigning you a task, the relationship shifts. It feels like a real colleague.
I think around 60% to 70% of work can now be done by AI employees. It has dramatically improved efficiency. It’s a no-going-back experience.
Putting OpenClaw into our workspace
At the beginning of January, OpenClaw caused a lot of trouble for us because it was designed for personal assistants. It was meant for one person, not for a team.
Here’s what that means: If you connected the AI to company data, it would share everything with anyone who asked. For example, if it had access to financial data, anyone could ask and receive that information. That’s not acceptable for a company.
So we modified the structure and architecture of OpenClaw to better suit a company or team setting. The AI employees created by OpenClaw now understand roles, permissions, and relationships — who is who, what each person’s job is, and what they are allowed to know. If someone is not authorized, the AI will not share that information.
Kuse has about 20 human employees. As the system matured, we realized it was becoming very useful to our team. We felt like we were a completely different company. We could do many things at once.
The AI works 24/7 and constantly identifies work that can be done. Everything we tell it, it processes and asks: How does this affect the business? What should we do next? What improvements should we make based on user feedback?
AI employees can work 24/7, but humans can’t
As humans, we need rest. It can be overwhelming, especially in East Asian culture, where we feel we must solve problems before resting.
So we created a Slack channel called “human only.” AI is not allowed there, since almost every channel in the company has AI employees.
This gives us a space to rest and talk casually. Otherwise, even random messages trigger the AI to generate tasks.
For example, if I share a random link, a human employee would probably take a look and not say anything, but the AI would say, “This is crucial information. We can do something based on this new information. We can do A, B, C, and D.”
I was like, “OK, maybe we should do that, but we still have 1,000 other things we need to do.”
In a world of infinite work and workforce time, prioritizing tasks becomes a more important skill. You have so many things that can be done. Even though in the end you can do three times or even 10 times more things, you still need to decide what matters most. Otherwise, you will be exhausted.
AI employees push us to work better
At the beginning, the team was terrified. AI employees could do many things that humans used to do. People wondered if they would be replaced.
But after working with AI employees, we felt strongly that they amplify our capabilities rather than replace us. They help us focus on our core strengths instead of repetitive work.
They push us to ask: What is my real expertise? What is my true contribution to this company, industry, or world?
Human employees need to strengthen their ability to communicate with real people. In the future, building trust between humans will become more important.
Do you have a story to share about tech in Asia? Contact this reporter at cmlee@insider.com.
