CAMERON WESTLAND

My Corner of the Web

Why Your AI Startup Needs to Move Beyond Figma and Start Building

Date:

Let's talk about Figma and AI startups. Don't get me wrong, Figma is a powerful tool. It's great for inspiring teams and aligning visions. In many ways, it's like concept art in video game development - beautiful, aspirational, and great for setting the tone.

But here's the thing: concept art never tells you if a game is fun to play. And Figma prototypes can't tell you if your AI actually works.

A Cautionary Tale from the Trenches

Let me share a painful lesson from my own experience. In my last startup, we set out to build a chatbot/AI product manager. We poured our hearts into creating stunning designs (in Framer, but the principle's the same as Figma). The mockups were beautiful, pixel-perfect even.

But here's the kicker: we grossly underestimated the challenge of having an AI guide users through creating a product brief. Worse yet, we hadn't validated if users even wanted this in the first place.

The result? We had to shut down the startup. It's a tough pill to swallow, but I can't help but wonder: if we'd focused on building a walking skeleton instead of pixel-perfect designs, could we have identified these critical issues earlier and pivoted?

The Concept Art Trap

This experience taught me that we, as AI startup founders, often fall into the concept art trap. We craft gorgeous mockups, thinking we're making progress. But are we really? Or are we just creating digital paintings of what our AI product might look like?

Beyond Pretty Pictures

Don't get me wrong - vision is important. Having a clear, visual representation of where you're heading can be incredibly valuable for team alignment and investor pitches.

But at some point, you need to put down the digital paintbrush and pick up your keyboard. Because the real challenges in AI development aren't about UI design - they're about making the damn thing work.

The Walking Skeleton Approach

Instead of polishing that Figma prototype for the umpteenth time, why not build a "walking skeleton" of your AI product? It doesn't need to be pretty. It just needs to work.

In our case, a walking skeleton could have been a simple command-line interface that attempted to guide a user through creating a product brief. It wouldn't have looked impressive, but it would have exposed the core challenges we faced much earlier.

Minimum Viable AI

Think of it as creating the most basic playable level of your game. What's the simplest version of your AI that can provide value to users? Build that.

It might be ugly. It might be buggy. But it'll be real, and that's what counts.

A Call to Action

So here's my challenge to you:

  1. Use Figma for what it's good at - inspiring your team and aligning on a vision.
  2. But then, step away from the pretty pictures.
  3. Identify your core AI functionality.
  4. Build the simplest possible version of it.
  5. Get it in front of real users ASAP.
  6. Iterate based on actual feedback, not assumptions.

The Harsh Reality

In the world of AI startups, shipping working features trumps pixel-perfect designs every time. Your users don't care about your beautiful Figma files. They care about whether your AI can solve their problems.

Wrapping Up

Figma has its place in the AI development process, just like concept art has its place in game development. It can inspire, align, and excite. But it can't replace the hard work of building and testing real AI functionality.

So, by all means, create your concept art. Align your team. Get inspired. But don't mistake it for the real thing. The sooner you start building, the sooner you'll know if you're onto something truly game-changing.

What will you build today? And more importantly, what will you learn from it?