In the early 2000s, robotics engineering wasn't specialized, forcing students to learn software, mechanical, and electrical engineering. This "jack of all trades" background taught rapid context-switching, systems thinking, and grit—core competencies for successful product managers and startup founders.
A product manager's primary role is not just managing roadmaps but injecting courage into the team. This means making unpopular decisions, like scrapping a project after months of work, to ensure the team is always building the right thing, even when it's difficult or requires challenging leadership.
Pivotal career moments often require reaching a point of conviction where you're willing to risk your job to advocate for the right decision. This "screw it" mentality is not about recklessness but about having such deep domain knowledge that you feel compelled to make the difficult call.
To manage non-deterministic AI products, Shopify created an internal tool where PMs grade AI-generated outputs. This creates a "ground truth" dataset of what "good" looks like, which is then used to fine-tune a separate LLM that acts as an automated quality judge for new features and updates.
At companies like Shopify, the best PMs quickly abandon the "CEO of the product" mindset. They are instead motivated by the "magical moment" of intense collaboration where ideas are built up and torn down collectively. The resulting solution, better than any one person could create, becomes an addictive high that retains top talent.
To persuade CEO Tobi Lütke on a controversial API versioning strategy, Vanessa Lee presented her plan on a simple, hand-drawn piece of paper. This low-fidelity format intentionally lowered the barrier for feedback, inviting the CEO to collaborate on the "how" rather than just approving a polished, final proposal.
To gain deep context without perpetual micromanagement, Shopify's VP Product runs intense, daily sprints with teams on key initiatives. This short-term, deep immersion allows her to understand the domain, build rapport, and then step back while still being able to provide educated guidance long-term.
Building non-deterministic AI products fundamentally changes the PM role. Instead of creating detailed, rigid specifications, the PM's primary task becomes defining and codifying "what good looks like." This is done by repeatedly grading AI outputs to train evaluation systems and guide the model's behavior.