From StarCraft to ChatGPT’s Engine Room
Our Guest-Jerene Yang
Jerene Yang is an engineer at OpenAI, where she works on the behind-the-scenes systems that help AI models like ChatGPT run reliably at global scale. Her work sits at the intersection of infrastructure, research, and real-world deployment, focusing on how advanced AI can be built to be fast, efficient, and useful for millions of people.
Originally from Singapore, Jerene’s path into AI didn’t follow a straight line. It grew out of a love for math, problem-solving, and curiosity about how complex systems work. Along the way, she moved between regions and teams, gaining a firsthand understanding of how culture, context, and lived experience shape the way technology is built and used.
Today, her work at OpenAI reflects that perspective. She cares deeply about responsible AI, practical impact, and using technology to make everyday tasks easier rather than more complicated. Jerene brings a thoughtful, human lens to some of the most powerful AI systems in the world, always asking not just what we can build, but who it’s really for.
“Being in AI is really about being brutally efficient with your time, and using technology to automate everything that isn’t the best use of your energy.”
- Jerene Yang
Most people meet AI at the surface. You type a question. It answers. Maybe it helps plan a trip, summarize a document, or explain black holes like you’re five.
What you don’t see is the invisible machinery behind it or the humans quietly keeping it all running.
That’s where Jerene Yang comes in.
Born and raised in Singapore, Jerene didn’t grow up dreaming of Silicon Valley. She grew up loving math, games like StarCraft, and tinkering, the kind of curiosity that doesn’t look flashy at first. Today, she works at OpenAI, helping build the systems that make ChatGPT fast, affordable, and reliable for hundreds of millions of people around the world.
This isn’t a story about genius lightning bolts. It’s about saying yes, crossing oceans, and learning how to think differently.
Saying Yes by Default
When Jerene moved from Singapore to the US for university, the biggest challenge wasn’t the coursework. It was the mindset shift.
In Singapore, she explains, the default is often “no” you move only when permission is clearly given. In Silicon Valley, it’s the opposite. You assume “yes” unless someone stops you. That subtle difference changes everything: what you attempt, how boldly you design systems, and how willing you are to explore ideas that might fail.
That shift, from careful optimization to open-ended possibility, shaped her entire career. It helped her move from computer science at a time when it wasn’t trendy, through startups and big tech, and eventually into AI infrastructure, one of the least visible but most critical layers of modern technology.
The Job You Can’t See
So what does Jerene actually do?
When you use ChatGPT and it responds quickly instead of freezing or crashing, that’s infrastructure at work. Jerene explains it like this: AI has two hidden engines. One trains the models so they’re accurate and knowledgeable. The other delivers answers fast and cheaply when millions of people ask questions at once.
If infrastructure fails, AI becomes slow, expensive, or unreliable. Subscriptions skyrocket. Errors multiply. Trust erodes.
That’s why infrastructure matters, even if most users never think about it.
Jerene now leads research teams focused on the data that trains AI models, asking hard questions: Is this information accurate? Is it culturally aware? Does it work for users outside the US and Europe?
When AI Grows Faster Than Expected
ChatGPT feels permanent now, like search or email. But it’s barely three years old.
Jerene describes working behind the scenes during sudden viral moments, like when users flooded the system with requests to generate Studio Ghibli-style images. What looks like a fun trend on the surface can cause massive strain underneath. Teams scramble to rebalance capacity so users don’t hit error messages or long delays.
Growth at that speed forces humility. Inside OpenAI, she says, no one assumes the product is “done.” The focus is always on what still doesn’t work well enough, and where AI needs to improve to actually be useful, safe, and affordable.
AI Isn’t an App. It’s an Assistant.
Jerene doesn’t think the future of AI is people constantly “using” AI.
The real shift, she says, is when AI disappears into everyday life. You’re booking flights. Shopping online. Planning meals. Talking to customer support. AI quietly helps in the background, saving time without demanding attention.
Her favorite examples are practical, not futuristic: an airline chatbot that actually understands your question, or snapping a photo of your meal so AI tracks calories without tedious logging.
The goal isn’t novelty. It’s less friction.
Fear, Reality, and What’s Overblown
People worry about AI for good reason, privacy, job loss, hallucinations. Jerene doesn’t dismiss those concerns. But she puts them in context.
She compares today’s anxiety to early fears about online banking. Trust grows gradually. You don’t have to share everything. AI can be useful without touching sensitive data.
As for doomsday scenarios? She’s blunt. Leading AI companies are deeply cautious. Safety and alignment aren’t afterthoughts, they’re central to every release.
The real risk isn’t robots taking over. It’s people assuming AI is always right and switching off their own judgment. Jerene encourages users to challenge AI, ask for sources, and push back, just like you would with a human.
Why Asia Matters So Much
Singapore has one of the highest per-capita ChatGPT usage rates in the world. Jerene isn’t surprised.
Curiosity runs deep. Education is prized. And governments across Asia have invested heavily in digital literacy. AI also solves real problems here: language barriers, access to knowledge, and time-consuming bureaucracy.
For OpenAI, Singapore isn’t just an office. It’s a gateway to understanding culture, language, and nuance across Asia Pacific, things models can’t learn from English data alone.
The Real Advice
If Jerene could go back and talk to her younger self, she wouldn’t say “be safer” or “follow the plan.”
She’d say: experiment earlier.
Today, the tools are everywhere. You don’t need permission to learn, build, or explore. Whether you want to research AI deeply or just use it to improve a small business, the barrier to entry has never been lower.
Curiosity still matters. So does courage. And sometimes, the biggest career shift starts with something as simple as saying yes.
