Z.AI's culture mandates that technical leaders, including the founder, remain hands-on practitioners. The AI field evolves too quickly for a delegated, hands-off management style to be effective. Leaders must personally run experiments and engage with research to make sound, timely decisions.

For Instagram's "Whiteout" redesign, co-founder and CTO Mike Krieger's initial directive was to "build and ship it" without A/B testing. This reflects a philosophy that for major, vision-driven product changes, data-driven incrementalism can be a trap, preventing the big leaps necessary for innovation.

In today's fast-paced tech landscape, especially in AI, there is no room for leaders who only manage people. Every manager, up to the CPO, must be a "builder" capable of diving into the details—whether adjusting copy or pushing pixels—to effectively guide their teams.

The now-ubiquitous "hold to pause" feature in Stories was created because engineer Ryan Peterman felt it should exist while dogfooding the product. He instinctively tried to pause a video with his thumb, and when it didn't work, he simply built it. This shows how engineers can drive product innovation by implementing their own user instincts.

A great tech lead provides a safety net without micromanaging. The analogy is a driving instructor who starts with their hands near a second steering wheel, ready to intervene, but gradually backs off as trust builds with the student. This approach gives engineers freedom to grow while ensuring the project stays on track.

The best leaders don't just stay high-level. They retain the ability to dive deep into technical details to solve critical problems. As shown by Apple's SVP of Software, this hands-on capability builds respect and leads to better outcomes, challenging the 'empower and get out of the way' mantra.

As a junior IC at Instagram, Adrian was told leadership had "bigger fish to fry" than his A/B testing idea. He built a scrappy, functional prototype anyway, recruiting a PM for air cover. This bottoms-up initiative proved its value and ultimately led to his first senior promotion.

When the Instagram Stories project was churning, leadership made a counterintuitive move: they significantly cut the team size. This resulted in clearer ownership, less communication overhead, and faster decision-making, allowing a tiny core team to build and ship the massive feature in just a few months.

A technical CEO shouldn't ship production code. Their most effective use of coding skills is to build quick demos. This proves a feature's feasibility and can effectively challenge engineering estimates, demonstrating that a project can be completed faster than originally projected.