From a voice message to production

A lot of my ideas come during my morning walk or at the gym.

I'm a designer, so those ideas often turn into things I want to change in our products. A different interaction. A new feature. Something that needs fixing.

But having the idea was never the problem. What I did next was.

An idea is easy to leave as a note. Turning it into something I can look at and test takes another step. That step used to mean stopping what I was doing, or waiting until I was back at my computer.

The screens were getting in the way

I'd pull out my phone and start typing. Trying to explain the idea properly, before I forgot it. Suddenly, part of my walk or workout was spent looking at a screen.

That went against what I wanted from that time. I wanted to be outside. To train. Not turn every idea into another reason to stop and type.

I didn't want to lose the idea either. So I changed the way I work.

Now I can send a voice message in a chat app. Say what I'm thinking, explain what needs to change, and keep moving.

I've built a workflow around Instinct, Notion and Claude that fits the way I think and build. It helps me turn ideas into something I can test, without needing to be at my computer for every step.

The useful part isn't replacing typing with talking, I still do both. It's shortening the distance between noticing something and having a first version to react to. A saved note still needs me to come back and turn it into work. A first version gives me something to question.

An idea from a walk

During a walk, I had an idea for the AI Mirror redesign: a painted hand reaching towards its ASCII reflection, their fingertips almost touching.

I asked for a draft in a local preview. When I got home, the draft was there.

Not a finished design. Not a change already live on the site. A first version I could look at, test and improve.

That's the kind of idea that is specific enough to describe, but still needs to be seen. How do the two hands sit together? Does the contrast work at the top of the page? Does it still make sense on a phone?

Those are design questions. Describing the idea doesn't answer them. Having a draft gives me somewhere to start.

Less describing, more trying

The useful part was coming home to something I could react to, instead of a reminder to start it later.

Getting a first version faster doesn't mean skipping the design work. It means I can spend less time getting the idea out of my head and more time deciding whether it works.

I can look at what is there, see what feels wrong, and give more specific feedback. Sometimes the first version will show that the idea needs changing. That's useful too. I'd rather learn that from something I can test than keep polishing the description in my head.

Voice is one way I can make that first step easier in my current workflow. I can explain a change while I'm away from my computer, without stopping to write a long message.

For a simple fix, I can describe it in chat and have it live about ten minutes later, without opening a computer. That's different from the AI Mirror draft in a local preview. A design like this still needs testing and refinement, and larger changes take longer.

I still decide what should be built, review the result and approve what goes live. The shorter path is to a decision, not around it.

If your team has ideas that get stuck before anyone can try them, we'd like to help you rethink that process at LayerX.

Designing how we work

I don't think every team needs to run product work through a chat app.

Your friction might be somewhere else. Repeating a handoff. Losing context between teams. Spending too much time turning a decision into something people can use.

Before adding AI, I'd start with that moment. Where does the work stop? What does someone have to repeat? What would help them take the next step without losing the context?

The answer might be a first draft, a proposed change or a clearer handoff. It still needs someone to judge it. The goal isn't to remove people from the process. It's to give them something useful to work with sooner.

Designing a better way to do that is part of being a designer too. We spend a lot of time thinking about the experience people have with a product. The way a team gets that product built deserves the same attention.

At LayerX, we offer AI consulting to help teams improve how they work. Bring us one process that wastes your team's time. Let's look at what could work differently.

Let's start with that, not with a list of tools.