
We talked to people about stuck moments. What surprised us: they weren't short on advice. They had plenty of answers already. What they needed was someone to help them think out loud, and a nudge to commit to one.

Our early concepts tried to deliver answers - flowcharts, AI suggestions, quick decision trees. None of them held up against what we'd heard. So we changed the frame: from giving answers to prompting a conversation.

Drop AI-style answers. The tool prompts; two people do the thinking.

Short, structured questions are easier to start on than open-ended ones.

A quick "okay, I'll do that" beats a deep analysis nobody acts on.
Within minutes, people went from "I don't know what to do" to a concrete next step. The second cube did more work than the first, it turned a prompt into something that felt like being listened to.
This designathon taught me something I didn't expect: ambiguity isn't always a problem to solve. Sometimes it's the whole point.
