Nick Turley
The fastest-growing product in history, covering AI product work, product design, and go-to-market execution.
Episode
Inside ChatGPT: The fastest-growing product in history | Nick Turley (Head of ChatGPT at OpenAI)
Summary
Nick Turley, Head of Product at ChatGPT, gives an inside account of building the fastest-growing product in history — from a hackathon project called "Chat with GPT-3.5" to 700 million weekly active users. He discusses OpenAI's product philosophy (ship fast to discover what to polish), the decision to keep health use cases enabled, and ChatGPT's emerging role as a traffic and discovery engine.
Key Takeaways
In AI product development, you won't know what to polish until after you ship — shipping raw is strictly less bad than waiting to polish the wrong things.
Evals are just the modern version of articulating success before you do anything else — specifying ideal behavior for use cases now communicates to ML researchers who can optimize against them.
Hire for curiosity over AI credentials: filtering for people who've worked in AI gives a narrow, lucky sample — curiosity about how the technology works is a stronger predictor of success.
Run recruiting like executive search: diagnose the specific skill gap per team from first principles — sometimes a team doesn't need a PM at all; it needs a data scientist or front-end specialist.
ChatGPT driving traffic is best optimized the same way you'd appeal to a real user — make genuinely high-quality content — because AEO tricks will self-correct as models improve.
Notable Quotes
“It's quite interesting to see that stuff play out because I think all these people entered school for genuine reasons. They were excited about the space, they were researching it, they were pursuing knowledge, and I'm happy that that's being rewarded. And I don't know what the rewards will look like in the future, especially in a post-AGI world. But I just a feeling that if you follow that advice, you'll end up okay.”
“So we want to run towards these use cases by making the model behavior really, really great. That can mean connecting you with external resources when you're struggling. That can mean not directly answering your question, but instead of giving you a helpful framework in the case of like, "Should I break up with my boyfriend?" ChatGPT should probably not answer that question for you, but it should help you think through that question in the way that a thoughtful companion would. So I think it's really important to do the work because I think the upside is immense.”
“So it's situational. Again, you kind of have to be first principles about it. But I do think using velocity, especially early on, as a tool... Actually this has been said about consumer social for example. It is not the first space where people have said, "Hey, you just got to try 10 things because you're probably going to be wrong." So I don't think this has never existed before as a dynamic either, but I do think with AI, it's important to internalize.”