Nick Turley

AI product development requires an empirical approach because capabilities and risks are emergent rather than pre-defined.

"I've never ever worked on a product that is so empirical in its nature where, if you don't stop, and watch, and listen to what people are doing, you're going to miss so much, both on the utility and o..."

In AI applications, the model and the interface are inseparable, requiring the model itself to be managed with a product mindset.

"One thing we've learned with ChatGPT is that there really is no distinction between the model and the product. The model is the product and therefore you need to iterate on it like a product."

Lowering friction through free access and a clean UI is critical for the mass adoption of complex AI technology.

"I think that in the original release, making it free was a big deal... making it free and putting a nice UI on it, very consequential in the way that you take for granted now. And this is why I think..."

A key litmus test for AI features is whether they scale proportionally with the underlying model's intelligence.

"If we're shipping a feature and it doesn't get 2X better as the model gets 2X smarter, it's probably not a feature we should be shipping."

Writing evaluations (evals) is the primary way for product managers to communicate desired behavior to AI researchers.

"I started writing evals before I knew what an eval was because I was just outlining very clearly specified ideal behavior for various use cases... it might be the lingua franca of how to communicate w..."

The ultimate platform goal for AI is to enable users to build and monetize entirely new businesses within the ecosystem.

"I also would love it if you could start a business on ChatGPT. I think there really is a world where, as this thing hits a billion user scale, it can get you distribution, it can get you started on ma..."

The Van Westendorp survey is an effective, rapid tool for determining price points even for novel technologies.

"What I did do is ship a Google Form to Discord with, I think the four questions you're supposed to ask on how to price something... the Van Westendorp survey. ... I remember distinctly A, you [got] a..."

For utility-based AI, retention is driven by problem-solving efficiency rather than time-spent metrics.

"We don't care at all how much time you spend in the product. In fact, our incentive is just to solve your problem and if you really like the product, you'll subscribe, but there's no incentive to keep..."

Prioritize hiring 'barrels'—individuals who can independently drive projects to completion—to maintain a high-output, lean team.

"We try to maximize the number of empowered people who can ship... people who can make stuff happen. And then you can add ammunition around them, which is people helping those people. ... we try to max..."

A leader's role is to establish the 'resting heart rate' or default pace of execution for the organization.

"I always felt like part of my role here, obviously, was to think about the direction of the product, but also to just set the pace and the resting heartbeat for our teams."

The long-term vision for AI is a proactive entity with deep personal context that reduces the need for detailed prompting.

"What we envision is this entity that can help you with any task, whether or not that's at home, or at work, or at school, really any context, and it's an entity that knows what you're trying to achiev..."

The current chat interface is a primitive starting point; the future involves more sophisticated, non-linear ways of interacting with intelligence.

"ChatGPT feels a little bit like MS-DOS. We haven't built Windows yet, and it will be obvious once we do. ... I think natural language is here to stay, but this idea that it has to be a turn-by-turn ch..."

Social media platforms like TikTok are invaluable for discovering emergent, unpredicted use cases for general-purpose AI.

"I go through those in detail because it's not like I knew about those use cases either. They're very, very emergent and I just go through the comments and process because there's so much to learn. ......"

The 'maximally accelerated' principle acts as a forcing function to strip away non-essential blockers and identify the critical path.

"I just really want to jump to the punchline of like, 'Okay, why can't we do this now?' or, 'Why can't we do it tomorrow?' ... if this was the most important thing and you wanted to truly maximally acc..."

In AI, shipping early is necessary to discover which specific areas actually require polish based on real-world usage.

"You won't know what to polish until after you ship. And I think that is uniquely true in an environment where the properties of your product are emergent and not knowable in advance. ... shipping is j..."