principle 1: user experiences are probabilistic, not deterministic
you can’t control how someone experiences your product. experiences are personal and subjective. all you can do is increase the likelihood that they’ll have a positive one. AI’s role is to optimize the odds—it doesn’t guarantee anything, but it can shape the environment to make a positive outcome more likely. it’s about shaping probabilities, not dictating experiences.
principle 2: AI is the gardener, not the architect
AI isn’t building rigid systems or telling users what to do. it’s creating a flexible, adaptive environment. think of it like a gardener, not an architect. the AI helps set the right conditions for users to succeed, but it doesn’t control how that success happens. over time, it adapts as users interact with it, growing and evolving alongside them.
principle 3: create a dynamic, learning environment
a great product is a living system. the product experience should constantly learn from its users, adjusting and improving based on how people interact with it. right now, we rely on human feedback to do this, but the future is a product that can learn directly from its users. it evolves, refines, and adapts in real-time, cutting out the middleman and making the feedback loop faster and more direct.
principle 4: co-regulate with the user’s nervous system
one of the most powerful things a product can do is sync with the user’s nervous system. your product should be able to regulate stress and calmness in the user. by paying attention to how users behave (like when they’re frustrated or taking too long on a task), the AI can adjust the experience to keep them either engaged or calm, depending on what’s needed. this creates a more personalized, responsive experience.
principle 5: start broad, refine subtle
when building a product, you start with big adjustments. you need to see what works and what doesn’t. but over time, as the AI learns more about the user, it should begin making smaller, more subtle refinements. these marginal gains are where the real magic happens, turning a decent product into a great one. it’s the little things that often make the biggest difference.
principle 6: reveal hidden constraints
users don’t always know what’s holding them back. sometimes they’re limited by things they can’t see—whether it’s mental, emotional, or something else. the AI’s job is to identify those hidden constraints and surface them in a way that helps the user move forward. by watching for patterns in behavior, AI can help users overcome obstacles they might not even realize they have.
principle 7: the product is always evolving
there’s no such thing as a finished product. users change, expectations shift, and the world moves forward. your product has to keep evolving. AI should be built to continuously learn and adapt, so the product stays relevant and valuable even as the world around it changes. it’s a constant cycle of evaluation and improvement.
principle 8: fail gracefully, iterate rapidly
not everything will work right away, and that’s fine. the key is to iterate quickly and learn from failures. even when things go wrong, your product should still provide value and recover gracefully. AI helps by speeding up the iteration process, making adjustments smarter each time. it’s about learning fast, failing fast, and improving faster.