Builder PM Framework
Diagnose your PM archetype and pivot from information relay to high-judgment builder.
Product management is bifurcating into two archetypes: the 'information mover' who translates and surfaces information up and down the org, and the 'builder' who makes judgment calls and creates directly. As AI automates information relay, the information mover role is becoming obsolete while demand for builders with strong judgment is at an all-time high. This framework guides PMs through diagnosing their current archetype, automating mechanical relay tasks, developing hands-on building capability, and sharpening the judgment muscle—evaluating what to build, which direction, and whether it's worth shipping—that AI cannot replace.
- Builders create direct connections between ideas and outcomes; information movers add latency and friction
- Judgment—evaluating what to build, which direction, and whether it's worth shipping—is the irreplaceable PM skill
- AI can automate information relay; it cannot automate taste and decision-making
- Responsibility without authority is the root of PM dissatisfaction; builders reclaim authority through direct action
- The cost of testing and changing is dropping 10–100x, making judgment more critical, not less
- Pace of continuous learning is now a competitive differentiator, not a nice-to-have
- Diagnose your current archetypeReview your last two weeks of work and categorize each major task as either 'information relay' (summarizing, translating, presenting to stakeholders) or 'judgment and building' (deciding what to build, making direct product calls, creating artifacts). If more than 50% is relay work, you are operating as an information mover.Pro tipBe ruthlessly honest. Status updates, meeting prep, executive decks, and research summaries are information relay even when they feel strategic.WarningDo not rationalize relay work as 'alignment' or 'influence.' Influence through direct building is what the market rewards in the AI era.
- Identify mechanical tasks to automateList every recurring PM activity that involves moving, reformatting, or summarizing information—writing research summaries, generating status reports, translating specs into briefs, preparing executive readouts. These tasks are your automation targets, not your differentiators.Pro tipA useful test: if the primary output of a task is a document that someone else uses to make a decision, it is relay work.WarningDo not let identifying these tasks become another relay exercise. The list should take 30 minutes, not two weeks.
- Automate your relay work with AI toolsImplement LLM-based tools to handle the mechanical information-moving tasks you identified. Use AI to draft status updates, summarize user research, generate spec variations, and prepare stakeholder materials. Reclaim that time for building and judgment.Pro tipStart with the relay task you dislike most. Automating what you hate first creates immediate motivation and relief to continue the transition.WarningGetting AI tooling 'set up perfectly' can become its own form of process theater. Ship the automation quickly and move on.
- Develop hands-on building capabilityStart directly building with AI coding and design tools without waiting for engineers or designers. Create a prototype, improve an existing feature, or fix a broken flow—own the artifact end to end. The goal is a shipped change, not a perfect spec document.Pro tipPick one small, real problem in your product and attempt to improve it using AI tools this week. The feedback loop of seeing something you built actually work is the core motivator that sustains the builder archetype.WarningBuilding for the sake of building is not the goal. Every session should be anchored in a specific judgment call you made first about what problem is worth solving.
- Practice the three-question judgment checkApply this decision filter to every product change: (1) Is this change good or bad for users and the system? (2) Is this direction sustainable and differentiated, or will it fragment the product? (3) Does this meet the criteria to be worth building and worth releasing? Write down your reasoning each time.Pro tipThe judgment muscle is built through repetition and post-mortems. Keep a log of your calls and review outcomes monthly to calibrate your instincts over time.WarningDeferring all decisions to data, user requests, or stakeholder pressure is the information mover pattern dressed as objectivity. The builder owns the call.
- Accelerate your continuous learning paceCommit to a weekly rhythm of learning new AI tools, building methods, and product approaches. Attend or watch live demos of what other builders are creating. The landscape is shifting every three months; staying current is now a core job requirement, not optional professional development.Pro tipSeeing peers build live—like a structured show-and-tell format—is the fastest way to update your mental models and surface tools you didn't know existed.Warning'I'll get to it later' is not a viable strategy. PMs who fall three months behind peers are already being seen as using deprecated approaches.
At a monthly meetup of 125 heads of product in San Francisco, Nikhyl asked members to show what they were building with AI. Every attendee had their laptop open, demonstrating AI-powered workflows—from chief-of-staff automation apps to agent-driven prioritization tools. The energy was completely different from meetings focused on process and roadmaps. PMs who had shifted to hands-on building described feeling satisfaction for the first time in years, pointing to direct feedback loops between their decisions and shipped outcomes.
Nikhyl describes taking breaks from his PM job to change light bulbs at home—because seeing a broken bulb fixed and the light come on was more satisfying than anything in his day job. The reason: PM work under a responsibility-without-authority model offered no direct feedback. You couldn't see something broken get fixed. Now, with AI coding tools, a PM can directly design, build, and ship—experiencing the immediate satisfaction of a concrete problem solved without waiting for a designer's backlog slot or an engineer's sprint.
Extracted from Nikhyl Singhal's insights on Lenny's Podcast, drawing from his 30 years building consumer products and his role leading the Skip community of 125+ heads of product at companies of all stages.