Skip to main content

The Problem Solvers: Kay Zhu at Genspark

TL;DR

  • Kay Zhu reflects on how AI changed his advice to his college-bound son, who is now studying commercial dance rather than computer science because AI lets people pursue what they love.
  • Genspark is an all-in-one AI workspace (AI spreadsheets, slides, and documents) for everyday white-collar workers, and as a small startup it relies on trusted partners like Anthropic's applied AI team.
  • He argues no one really has a moat anymore in such a fast-moving field; the only durable advantage is team culture, and successful partnership comes down to openness, deep mutual trust, and a tight feedback loop.

Takeaways

  • AI has shifted career advice: three years ago Kay might have pushed his son toward computer science, but now encourages him to follow his heart — the son is studying commercial dance.
  • Genspark positions itself as an all-in-one workspace covering AI spreadsheets, AI slides, and AI documents for normal white-collar workers.
  • Working with the applied AI team feels like "an invitation to a party" — Genspark's engineers are curious about the latest models and enjoy each new request.
  • As a small startup, Genspark can't build everything itself, so working with a trusted partner is really important.
  • Kay believes nobody has a moat today because everything moves so fast; the only moat is the team's culture of exploring and trying the newest technology to execute faster.
  • In collaboration, openness is key — any secret you keep today will be worthless tomorrow — and partnership success comes down to deep, mutual trust and a tight feedback loop.
  • When a candidate asked what Genspark would be in two years, his honest answer was "I don't know," expecting new things that may even seem like magic.

Vocabulary

Genspark — An all-in-one AI workspace offering AI spreadsheets, slides, and documents for white-collar workers. Applied AI team — Anthropic's team that partners with companies like Genspark to deploy and adopt the latest models. Moat — A durable competitive advantage; Kay argues the only real one now is team culture. White-collar workers — Office-based professional workers, Genspark's target users. Commercial dance — The performance-dance field Kay's son chose to study instead of computer science. Tight feedback loop — Fast, continuous exchange between partners that signals a working collaboration. Trusted partner — A reliable external collaborator a small startup leans on to do what it can't build alone.

Transcript

I have a son who is turning 18. Three years ago, if you asked me, I might have forced him to major in computer science. He is actually going to college to study commercial dance, because AI can help people do so much. You basically should pursue your heart — just follow your heart.

Genspark is the all-in-one workspace. We do AI spreadsheets, AI slides, AI documents — all kinds of work for normal white-collar workers.

When you think about how we work together — the applied AI team and Genspark — what's it like on your side when you get something new from us? All our engineers are very curious about the latest advancement, the latest models, so every time we get this request, it's kind of like an invitation to the party, so we enjoyed it a lot. We are a very small startup, so we cannot do everything by ourselves, and working with a trusted partner is really, really important.

What do you think are the things that you do that make you unique and help you succeed? Everybody is talking about moat these days. I don't really think nowadays anybody has any moat, because everything is happening so fast. The only moat is the team's culture. We are always exploring and trying the newest technology to execute faster.

In the collaboration, the openness is the key. The secret you keep today will be worthless tomorrow. The success of the partnership comes down to deep, mutual trust. When it's really working, it's a very tight feedback loop. I just did an interview yesterday and the candidate asked me, "What will Genspark be in two years?" My answer is, I don't know. A lot of new things will happen and it will even seem like magic.

Feedback / ReportSpotted an issue or have an improvement idea?