Missed Billion-Dollar Zhipu Options: A Tale of Choice and Timing
“Fate, luck, and feng shui” — a popular saying online claims each person has only 3–5 life-altering opportunities to defy destiny.
In 2023, at age 23, Hua Yang — a graduate student from a tier-4 city in Henan — received an internship offer from Zhipu AI, the Beijing-based LLM pioneer backed by Tsinghua University and creator of ChatGLM. After six months, he was offered full-time employment — complete with vesting stock options.
He declined.
Driven by stability and familial expectations, he chose academia: a PhD track, aiming for a university faculty position. Today, that decision means he missed out on a potential nine-figure (HKD) wealth event.
📈 Zhipu’s Meteoric Rise
By late June 2026, Zhipu’s market capitalization surged past HK$1.2 trillion, following its successful HKEX IPO. According to its prospectus:
- Two employee shareholding platforms — “Huihui” and “Zhideng” — collectively hold ~14.96% of total shares.
- These platforms cover 451 employees, representing 51% of the workforce.
- At HK$1.2T valuation:
- Huihui (426 employees): ~HK$210M per person (average)
- Zhidend (25 core staff): ~HK$2.3B per person (average)
💡 For context: Zhipu’s early-stage valuation in 2024 was just US$2B — a 60× increase in under three years.
🔍 Inside the Internship: “Accidentally Entered the Palace”
Hua Yang joined Zhipu’s ChatGLM core algorithm team, mentored directly by Chief Scientist Prof. Jie Tang (Tsinghua). What struck him most wasn’t prestige — it was unprecedented resource access:
- ✅ Immediate permissions to 8x NVIDIA A100 GPUs — rare even among top-tier Chinese AI labs.
- ✅ Full access to thousands-node compute clusters, dedicated annotation teams, and cross-functional R&D autonomy.
- ✅ Interns could propose projects — and get greenlit with new business units formed to support them.
Contrast this with his prior big-tech internship, where he operated as a “screw in a rigid machine.”
Figure | Hua Yang’s official Zhipu internship offer
⚖️ The Crossroads: Stability vs. Leverage
In early 2024, Hua Yang faced a pivotal choice:
| Option | Pros | Cons |
|---|---|---|
| Accept Zhipu’s full-time offer | HK$30K+/month + equity; frontline LLM R&D; global impact | High intensity; uncertainty pre-IPO; long vesting schedule |
| Pursue PhD | Tenure-track security; academic freedom; family approval | Minimal stipend (~HK$2K/month); delayed real-world leverage; opportunity cost |
Every advisor — parents, professors, peers, even industry contacts — echoed one message: “PhD first. Stability is priceless.”
He accepted the narrative. Sent the rejection email. And watched — from afar — as Zhipu’s valuation soared.
Figure | Zhipu’s headquarters — lit late into the night during peak model training cycles
🚀 From Regret to Reinvention: Betting on IP Robotics
By January 2026 — after Zhipu’s IPO — Hua Yang re-evaluated his path. He realized linear accumulation (grades → papers → tenure) no longer guarantees transformational outcomes.
“I needed to learn how to leverage my life — not just accumulate.”
He pivoted:
- Left his PhD program (with advisor’s reluctant blessing)
- Co-founded an IP-driven robotics startup, targeting character-based humanoid agents
- Secured seed funding and signed licensing deals with 3 major intellectual properties
- Positioned at the intersection of LLM reasoning + embodied AI + branded interaction
Figure | Hua Yang presenting at a 2026 AI robotics symposium
🧭 Reflection: Judgment ≠ Failure
Hua Yang doesn’t vilify his younger self:
“That 23-year-old saw only the safe road — because that’s all his environment showed him. His ‘lack of vision’ wasn’t ignorance. It was context.”
His takeaway isn’t regret — it’s recalibration:
- Opportunity recognition requires exposure, mentorship, and pattern literacy.
- Equity isn’t just compensation — it’s co-ownership of inflection points.
- The next chance won’t look like the last — but action creates visibility.
“A book says we get seven life-changing chances. Missing one isn’t failure — refusing to reach for the next one is.”
— Hua Yang, Founder & CEO, PersonaBot Labs
Article by AI Story Project | Author: Zhong Chudi