GenAI is magical at intersectionality
“If a machine is expected to be infallible, it cannot also be intelligent.”
― Alan Turing
The Clicking Moment
It’s been almost a year since I started poking at ChatGPT, treating it like a search engine with conversational flair. It’s been a more human-feeling way to query the internet, but I know it had more to offer.
Then something shifted.
Instead of asking it to explain product management, I asked it to contrast different theories of product management. Instead of requesting a definition of cognitive framing, I wanted alternative theories for similar observations. That’s when it clicked.
What I’ve Discovered
The magic of GenAI isn’t in its ability to give me definitive answers—it’s in its capacity to hold multiple frameworks in tension simultaneously. When I ask about Simon Wardley’s Explorers, Villagers, and Town Planners and then ask how it could be leveraged in a design thinking exercise, or when I explore how Deleuze’s philosophical concepts of “bodies without organs” might reframe my workout routines, the model offers something a search engine can’t provide. It offers a way to wonder (and wander) through curiosities. It expands its value beyond an answer machine.
My best GenAI conversations follow this pattern:
- Start with a concept I think I understand
- Ask for alternative perspectives or adjacent theories
- Push for connections between seemingly unrelated domains
- Iterate until I find a new way to frame my thinking
My favorite example remains the simplest and most personal: my philosophical fitness journey. By asking an AI to help me explore how Deleuze and Guattari’s work might inform workout motivation revealed connections I never would have made on my own, and it’s been effective: I am infinitely more motivated to workout because I explored what it meant to me in a new way.
The Constraints That Matter
This only works within specific boundaries:
Topics I know well enough to correct hallucinations. I’m not learning new technical skills from ChatGPT; I’m exploring different ways to think about skills I already have.
Low-stakes exploration. Like my approach to work as a game, this is about expanding my thinking, not replacing it. I’m not making life-altering decisions based on AI responses—I’m collecting new lenses through which to view problems I’m already working on.
Intersectionality over authority. The model excels at finding connections between established ideas, not at generating novel insights. It’s semantic pattern matching, not wisdom.
When Voice Mode Becomes Thinking Mode
The most surprising discovery: sometimes my mind needs more interactivity than a podcast but my curiosity is too niche to have anyone to discuss it with. Voice mode ChatGPT has become my thinking-out-loud partner for organizing thoughts around concepts that exist at the intersection of multiple domains.
I always prefer talking to people—this isn’t replacing human conversation. But a random debate with a semantic model can help me organize thoughts in ways that pure introspection sometimes can’t.
An Updated Product Perspective
My shift in usage has fundamentally shifted how I think about GenAI as a product leader. We are still early enough where most organizations are approaching AI as a feature factory—adding chat interfaces, summarization tools, and content generators to existing workflows. But that’s the job to be done that differentiates its value.
The real opportunity isn’t in making existing tasks faster. It’s in enabling entirely new ways of thinking that were previously inaccessible. When I help teams think through product strategy, I’m constantly looking for those moments where we can reframe the problem space rather than optimize within it. GenAI excels at exactly this kind of reframing.
Instead of building “AI-powered search” or “smart document summaries,” what if we built tools that help knowledge workers explore the intersections between their expertise and adjacent domains? Or their work and our organization’s definition of value? The most valuable product experiences will be those that recognize this fundamentally different JTBD: not finding answers faster, but discovering better questions.
The Takeaway
GenAI works best when treated as a crossroads rather than an add-on or destination. It’s not about getting the answer; it’s about discovering which questions are worth asking next. The technology is magical at intersectionality precisely because it has been trained on the connections between everything. Use it to find the bridges between islands of knowledge you already possess, not to build new islands from scratch.