-
Chat interfaces assume you know what to ask. Most people don’t.
Years ago, analyzing tens of thousands of search queries taught me something that still applies today: people often don’t arrive with a fully formed question. They have a situation. They need to orient before they can retrieve. Chat-first AI interfaces weren’t designed for that. Most still aren’t.
-
AI products don’t need better screens; they need better systems.
Most AI design conversations focus on interfaces. But when a product’s behavior is probabilistic, the hardest design decisions aren’t on the screen. They happen before anyone opens a design tool. Here’s what I think design teams need to do differently.
-
The conversation that should happen before any AI project kicks off
The people funding the project, the people building it, and the people whose workflows it changes all have completely different ideas about what success looks like. Nobody includes that in the project plan, and that’s usually where things start to go wrong.
-
Most AI projects don’t fail at the model
88% of companies are using AI in some form, yet only 6% can point to real business impact. The gap isn’t a technology problem, and closing it starts with understanding what’s actually getting in the way.