Follow Curiosity and Smart People, Not Money

Nick Turley's approach to building ChatGPT reveals a leadership philosophy centered on rapid iteration, empirical learning, and first-principles thinking in the face of unprecedented technology.

At the core of Turley's philosophy is the belief that with AI products, you cannot reason your way to understanding what will work—you must ship to learn. "This is a pattern with AI where you won't know what to polish until after you ship," he explains. This led to the principle he calls "maximally accelerated," where teams constantly question if they're moving as quickly as possible. This isn't about cutting corners, but about recognizing that theoretical planning has diminishing returns when building something fundamentally new.

Turley approaches team building by focusing on skills needed rather than roles. Instead of automatically hiring for traditional product positions, he assesses what each team specifically lacks: "I really like to go through every single team and figure out what are the skill sets that team needs and how do you put it together from principles rather than just assuming we're going to do a bunch of pipeline recruiting." This first-principles approach extends to how he structures teams—small, empowered groups with high autonomy.

For decision-making, Turley emphasizes the balance between speed and deliberation. While most product decisions favor velocity, safety decisions demand rigorous process: "We're very deliberate on that where process is a tool, and one of the areas where we have an immense amount of process is safety." This dual-speed approach allows the team to move quickly on product features while maintaining appropriate caution on consequential safety decisions.

When facing uncertainty, Turley believes in running toward difficult use cases rather than away from them. While many companies avoid risky territories like health advice or relationship guidance, he argues: "If you had a model that was state of the art on health benchmarks and you didn't use that to help people... I think the duty is to make it awesome and to do the work."

For those working with AI products, Turley's approach suggests several practical implications:

• Ship early versions to discover what users actually want rather than trying to perfect features in isolation
• Focus on building small, high-autonomy teams with complementary skills rather than filling standard roles
• Use whiteboarding sessions to break down silos and get teams thinking generatively together
• Apply different decision processes for different types of decisions—move quickly on product features but methodically on safety issues
• Measure success by retention and problem-solving effectiveness rather than engagement metrics

Perhaps most fundamentally, Turley advises following curiosity rather than trends: "If you surround yourself with people that give you energy and if you follow the things you're actually curious about, you're going to be successful in this era."