Problem that's driving me
I think the most interesting design problem is figuring out where AI should act and where it should ask as well as how do we translate existing patterns to cater for AI. It's the question underneath my work, the reason I'm building Flowprint, and topics I'm currently exploring.
Ideas I keep returning to
- Situated AI changes the trust contract. Traditional UX assumed predictable systems; AI products don't behave the same way twice, and our patterns haven't caught up - Designers are systems thinkers who've been handed pixel tools. The thinking layer is missing — that's what Flowprint is trying to fix - Reversibility is an underrated design primitive. Most "trust" problems in AI products are really "what happens if it's wrong" problems
What I'm reading right now
[The Alignment Problem] by Brian Christian on how AI systems learn the wrong things and what designers can do about it [Co-Intelligence] by Ethan Mollick on working with AI as a partner, not a tool The Mom Test by Rob Fitzpatrick: re-reading, because customer discovery is a muscle that goes slack
My values
Empathy. Transparency. Communication. They guide me in both professional and personal life.