Why cultivating agency matters more than cultivating skills in the AI era | Max Schoening (Head of Product, Notion)

1. Cultivate Personal Agency

Develop agency by actively tinkering and making things, realizing that the world is built by people no smarter than you, which empowers you to change things rather than just follow existing structures.

2. Design in Code Medium

Designers and PMs should prototype and design in code, not just static tools like Figma, because it forces a deep understanding of the material and how agent loops work, leading to better product outcomes.

3. Focus on Tiny Core

Identify and perfect a single, exceptionally good “tiny core” superpower for your product, rather than continuously adding features, as this core is what makes great products truly successful.

4. Increase Product Exploration Shots

Leverage AI to rapidly explore multiple product ideas and prototypes (“shots on goal”) early in a project, as the initial 10% of work is now “free,” allowing for more experimentation and faster iteration.

5. Build Taste Through Iteration

Develop “taste” by consistently doing reps, getting feedback, and iterating on your creations, much like training a model. Additionally, engage in side projects and expose yourself to high-quality work to refine your aesthetic judgment.

6. Prioritize Obviously Good Quality

Aim for “obviously good” products that are few in features but exceptionally high in quality, relentlessly iterating on the core idea until it truly resonates with users, rather than focusing on feature count.

7. Embrace Malleable Software

Seek or build software that allows users to easily change and adapt it to their specific needs, fostering ownership over their computing life rather than being confined by rigid, pre-defined applications.

8. Use Jobs-to-Be-Done Framework

When building or reviewing a product, use the “jobs-to-be-done” framework to honestly assess what users truly “hire” the product for, stepping outside your employee mindset to evaluate it from a user’s perspective.

9. Avoid Frenzy Distraction

Avoid getting caught up in the Silicon Valley “frenzy” or fear of missing out; instead, focus on your true passions and interests, understanding that historical patterns often repeat, which can reduce anxiety about an uncertain future.

10. Don’t Over-Optimize Token Spend

In the early stages of AI adoption, prioritize exploration and learning over strict optimization of token spend, as focusing too much on cost metrics can stifle innovation and lead to misleading performance indicators.

11. Avoid First-to-Market Trap

Focus on building a product that is “right” and durable for the long term, rather than rushing to be first to market, as initial lead often doesn’t guarantee lasting success.

12. Embrace Exclusivity for Quality

For certain products, it’s acceptable and even beneficial to design for a specific, high-value “top of the class” user group, as this focus can lead to a truly excellent product, even if it means excluding a broader audience.

13. Understand Computer Fundamentals

Deepen your fundamental understanding of how computers work, as this foundational knowledge is surprisingly lacking even among professional programmers and can provide a significant advantage.

14. Beware Fake Legibility

As an executive or system designer, be cautious of creating “fake legibility” through status reports or metrics that simplify complex realities, as these often neglect the true on-the-ground dynamics and can lead to flawed decisions.