How to be a top PM today in the AI era

TL;DR

PMs win by making great decisions, not just coordinating tasks. The talk argues that the true craft is decision making and offers a three part approach: understand people deeply, diagnose root causes, and refine the details that make a product useful and delightful. AI can scale interviewing and insights, but human judgment remains key.

Transcript

Hey everyone, it's so great to be here and I love being here to get to talk about one of my favorite things ever, being a PM, product manager, and builder. And what's pretty incredible is that it's never been so interesting to be building products. And after 20 years almost of building, particularly in consumers, by the way, it's very upsetting to ... Read More

Key Insights

• Understanding people deeply is essential and can be done by uncovering the jobs users hire products to do, which reveals real needs beyond surface features.
• The core value of a PM is decision making, not merely organizing work, and becoming a student of what makes decisions good or bad is critical.
• AI can scale interviews and user insights via agents that replicate structured interviewing, enabling you to gather deep user input at scale.
• Diagnosing root causes requires a clear, prioritized view of problems, followed by iterative fixes and re evaluation to move the product toward market fit.
• Multimodal understanding and knowledge systems enhance what you can offer in a product by combining text, visuals, and context to support decisions.
• Early product launches reveal what works and what doesn’t, teaching teams to distinguish between features that sound good and experiences users actually value.
• Contextual accuracy and authority are essential for AI powered products, ensuring users trust the information and the path the product provides.
• A disciplined, three part approach—understand, diagnose, and refine—helps teams shape products that feel purposeful, useful, and delightful in the AI era.

Questions & Answers

Q: What is the main skill a PM should master in the AI era

The main skill a PM should master in the AI era is decision making. The speaker argues that with the ability to build almost anything using AI agents, the true value to a product team lies in making high quality decisions. This involves evaluating user needs, prioritizing options, and choosing the path that best aligns with user value and business goals. It is a learned craft that can distinguish great products from good ones by focusing on outcomes rather than simply executing tasks.

Q: How do you understand what users need according to the talk

Understanding user needs begins with deeply exploring what jobs users hire products to do. The method involves putting yourself in the user’s life circumstances, asking detailed questions about their day, surroundings, and constraints. This jobs to be done perspective helps reveal hidden motivations and critical moments that drive behavior, guiding product decisions toward real user value rather than surface level features.

Q: What role does jobs to be done play in product development

Jobs to be done plays a central role by reframing product development around the tasks users want to accomplish. By interviewing users and reconstructing their experiences during pivotal moments, teams uncover the true drivers of choice. This insight informs what to build first, how to measure success, and how to align features with the tangible outcomes users seek.

Q: How can AI scale user interviews without losing depth

AI can scale user interviews by using agents trained to conduct structured conversations that follow the jobs to be done framework. These agents can ask targeted questions, capture nuanced responses, and return organized insights at scale. While automation accelerates data collection, it must preserve the depth of understanding by probing about context, emotions, and situational details that matter for decision making.

Q: What is described as the PM’s craft beyond task management

The PM’s craft is described as mastering decisions rather than merely managing tasks. The talk emphasizes being a decision making expert who understands tradeoffs, user value, and market signals. This involves prioritization, risk assessment, and iterative learning to guide product direction toward outcomes users care about and that contribute to business success.

Q: What examples from Instagram inform the talk

From Instagram, the talk references experiences like launching Stories, reels, feed ranking, and messaging. These examples illustrate how early experiments can reveal what works, how users respond, and how to iterate toward features that become core experiences. They demonstrate the importance of learning from initial launches to shape broader product strategy.

Q: How does Google Search in the AI era fit into the framework

Google Search in the AI era fits by combining knowledge with user friendly experiences, such as multimodal responses and integration with maps and authority data. This aligns with the framework by prioritizing accurate, context rich interactions that answer user questions effectively. It demonstrates how AI powered features can meet user needs while maintaining trust and usefulness.

Q: What three part approach does the speaker propose for building great products

The three part approach proposed is: understand people deeply by uncovering their needs and jobs to be done, diagnose root causes by prioritizing problems and testing fixes, and refine the details to create useful and delightful experiences. This iterative cycle helps teams move closer to market fit while leveraging AI to enhance execution, not replace judgment.

Summary & Key Takeaways

• A product manager’s key skill is making high quality decisions that guide direction and impact. The talk outlines a three part framework: understand people deeply, diagnose why something isn’t working, and refine the product details to improve usefulness and delight. AI assists scalability but does not replace due diligence in decision making.

• The speaker draws on experiences from Instagram and Google Search to illustrate how teams learn user needs, uncover jobs to be done, and use iterative diagnosis to move toward market fit. He highlights the importance of accuracy, context, and user value in AI powered products.

• The talk emphasizes practical methods like jobs to be done, multimodal understanding, and knowledge rich experiences that combine data with real world context. The emphasis is on disciplined decision making and a culture that tests ideas, learns, and improves through iteration.