My AI app felt off, and here’s why. The model building it has never used a phone, so it can’t feel what’s in your hands, and whatever layout it gives you, it assumes it’s fine. The fix is to describe the physical reality it can’t feel.
The two zones
- Thumb zone, the bottom. Where your thumb reaches whether you hold the phone in your left or right hand. Every control the user taps goes here.
- Passive zone, the top 65 percent. No controls, just the information that builds trust, like the detected input and the transcript.
The three constraints
This is basically the whole prompt.
- There is a thumb. Design for one-hand thumb reach.
- There is about a second and a half. The interaction has to be that fast.
- A baby should figure it out. No instructions, just obvious.
Take things out
Every extra control is one more thing in the way. You get more value taking things out of the interface than keeping them in.
The prompt
Here’s my phone, an iPhone Pro Max. My thumb only reaches the bottom comfortably, left or right hand. Put every control the user taps in that bottom thumb zone. Use the top 65 percent as a passive zone with no controls, just the info that builds trust, like the detected input and the transcript. Make every interaction take about a second and a half, and make it so a baby could figure it out. And take out anything that isn’t pulling its weight.