Using AI as a Systems Thinking Partner Instead of a Chatbot
Over the last couple of days, I’ve been thinking a lot about customer discovery, product strategy, and the growing role AI is starting to play in how designers and founders think through complex problems.
Yesterday, I wrote about something I noticed during Boulder Startup Week. A lot of founders were struggling with the same challenge once they moved beyond their own ideas and started trying to validate them with real people. They didn’t know where to find the right customers for discovery, how to move beyond “me-search,” or how to tell whether they were solving a meaningful problem versus hearing a handful of encouraging opinions.
The more conversations I listened to, the more it felt like the issue wasn’t a lack of ideas. It was a lack of systems for continuous learning, validation, and decision-making.
That observation became the signal I started exploring more deeply today.
One of the skills I’ve been actively trying to develop lately is thinking beyond interface design and more like a UX researcher, systems designer, product strategist, and PM. As the boundaries between these disciplines continue to blur, there’s increasing pressure to solve more operationally complex problems without necessarily having more time, resources, or larger teams to do it.
That means understanding much more than screens or features. It means thinking through workflows, operational systems, information architecture, decision-making processes, onboarding logic, and the boundaries between human judgment and AI assistance.
What stood out to me today was that I didn’t begin with a fully formed product idea. I started with a problem space and a signal worth investigating.
Normally, turning something vague like that into a coherent product direction would involve multiple discovery workshops, whiteboarding sessions, UX research exercises, architecture diagrams, strategy meetings, and rounds of documentation. Instead, I used AI as a structured thinking partner to help work through the ambiguity in real time.
The value had very little to do with generating interfaces, polished outputs, or code.
The value came from being able to pressure test assumptions and think through systems faster and more iteratively than I normally could on my own.
Instead of only thinking about features or happy paths, I found myself exploring ranking systems, object relationships, operational workflows, edge cases, onboarding flows, prioritization models, and conditional logic around how different parts of a system should behave under different circumstances.
More importantly, the process forced me to define assumptions much more explicitly.
It made me think more carefully about the difference between something being “interesting” versus actually painful enough for users to care about. It pushed me to question whether recurring conversations represented meaningful signal or isolated human context. It also made me think more critically about where AI should assist people versus where human judgment still matters most.
By the end of the process, I didn’t have a finished product.
What I had was structured clarity around the problem space and a much stronger understanding of the system I was actually trying to design.
I think this is one of the more important shifts happening right now across UX research, product strategy, systems design, and product management. AI becomes significantly more valuable when it’s used as a systems-thinking partner rather than just a content generator.
It doesn’t replace customer discovery, product thinking, or research.
But it does change how quickly you can organize complexity, challenge assumptions, and evolve ideas.