In the rapidly evolving world of technology, few voices carried as much insight as Gokul Rajaram. Having shaped products at Google, Facebook, Square, and DoorDash and backed hundreds of startups as an investor, Gokul reflected on how AI had rewritten the rules of product development, leadership, and career growth.

At the heart of his perspective was a provocative idea: in an era where AI could generate endless code, the one skill that remained truly future-proof was judgment. “With thousands of AI engineers producing output, the challenge wasn’t the code itself, it was deciding what actually mattered,” Gokul explained. “Every product leader I spoke to worried about the AI slop, the noise versus the signal. Human judgment was the differentiator.”

YOU CAN ALSO READ: From Setback to CEO: How Jamie Dimon Turned Adversity into Triumph

AI was not just a tool; it had become a force reshaping how products were built. Roles that were once clearly defined had begun to merge. Product managers, designers, and engineers increasingly collaborated hands-on, while autonomous AI agents handled much of the heavy lifting. “Six months ago, I tried building a video transcription tool with early AI. It failed repeatedly,” Gokul recalled. “But more recently, I could create a working version in under an hour. Tools like Claude had made design and engineering converge, and even non-technical people could ship products.”

This shift had transformed expectations for teams. Traditional top-down specifications were giving way to iterative, hands-on creation. Product managers focused on articulating the “why,” while cross-functional teams worked closely in code. What was cutting-edge six months prior was often obsolete, so product managers and engineers had to prototype actively and check AI outputs. AI reduced reliance on designers, increasing engineer-to-designer ratios, sometimes as high as twenty to one. Small changes in input could produce vastly different results, making the evaluation of outputs a core responsibility for PMs and researchers. Gokul compared this pace to an industrial revolution for services, where the frontier of capability moved every two months, making product development both more exciting and more challenging than ever.

For Gokul, a product manager’s role remained timeless: balancing customer needs with business priorities while serving as the keeper of the “why.” Every product had to drive measurable changes in customer behavior, turning non-customers into loyal, paying users. “Thousands of engineers might produce code,” he said, “but humans had to decide what mattered, evaluate outputs, and ensure alignment. Judgment was the skill that could never be automated.”

When building AI products, he recommended focusing on deep, compelling problems in high-value workflows where AI could replace repetitive tasks. Ownership of scarce assets such as data, hardware, and network effects created defensibility. Companies often had to migrate legacy systems to deliver complete solutions, and stickiness came from unique assets, network effects, or combined software and hardware systems. Great founders shaped their companies around their strengths. At Google, Larry Page and Sergey Brin focused on technology while Eric Schmidt facilitated strategy. Zuckerberg optimized engagement, and Jack Dorsey perfected intuitive design.

Judgment extended beyond product to communication. Gokul cited producer Rick Rubin, who described himself as a reducer, someone who simplified and refined. In leadership, reduction meant clarifying priorities and communicating effectively. Startups often began with two or three people in a room, but as companies scaled, structured communication became essential. Weekly all-hands meetings aligned teams on product, business, and progress. Weekly CEO emails shared two or three key updates across product, business, and team dimensions. Repetition was crucial to ensure ideas sank in. Transparency fostered trust and invited team input, particularly in small companies.

Success in advertising came from owning a coveted user base and surface, delivering measurable results even without owning inventory, and maintaining exclusive access to advertisers. Trust and exclusivity were critical, and middleman models rarely survived. Gokul warned that consumer behavior could shift quickly, especially as AI interfaces replaced traditional apps, making close monitoring and experimentation essential.

North Star metrics had to reflect both growth and customer value and were coupled with check metrics to prevent unintended consequences. Self-serve products encouraged organic adoption, better onboarding, and broader reach without direct sales. The AI era demanded builders, not middle managers, and managing AI agents became a core skill. Candidates were assessed through real projects, demonstrating initiative and problem-solving. Career advice was straightforward: focus on long-term impact, avoid rapid job hopping, and build a superpower.

YOU CAN ALSO READ: Can Bosun Tijani Deliver Nigeria’s Digital Transformation?

Authenticity was key in assessing founders. Understanding their story and exploring the choices and lessons that led to their solution, what Gokul called the “idea maze,” was essential. Board dynamics evolved, with management increasingly attending meetings, and assigning board buddies as advisors helped maintain alignment.

The conversation landed on a singular point: in a world awash with AI-generated output, judgment and editorial skills differentiated success from noise. Tools could scale, automate, and generate, but only humans could decide what truly mattered. “AI could do a million things,” Gokul concluded, “but the value came from knowing which of those things should be done, why, and how to execute with clarity and purpose. That was judgment, the skill no machine could ever replace.”