Your AI Is too smart for normal people ☠️

Why investing in your UX is the only way to help your product cross the chasm

Apr 03, 2025

👋🏻 Hi, it’s me Nickey Skarstad and welcome to Builders, where I break down what it actually takes to build products that people love (and actually use). No fluff, just hard-earned lessons.

Every time I use ChatGPT and select which model to use I slap my hand to my forehead. We love to move fast and ship stuff, but for the love of god, let’s make it easy to use. This is not it! 👇🏻 Today we’re going to chat about UX as a competitive advantage and why products like ChatGPT must invest here to compete in an already crowded AI marketplace.

The AI marketplace is already crowded

One thing we've learned from the rapid evolution of AI is that the best model today might not hold that title six months down the line. The AI landscape is moving fast, with new features and models emerging constantly, each bringing unique strengths and capabilities.

And let's be real: we're reaching the point where the actual AI models are becoming commoditized. GPT-4, Claude 3, Gemini, DeepSeek … they're all impressive, and they're all getting better at roughly the same rate. As capabilities converge, the tech itself becomes less of a differentiator.

So what happens when everyone has access to roughly equivalent AI models? The deciding factor becomes user experience!

UX as the ultimate moat

Remember the browser wars? Or more recently, streaming services? Once the core technology reaches a certain threshold, users choose based on the experience.

Some AI products are already figuring this out:

1. Claude's Custom Writing Styles – They recognized that people want AI that adapts to them, not the other way around. By letting users define specific writing styles, they've created a sticky feature that's hard to leave behind once you've invested in setting it up.

2. ChatGPT's Canvas – A brilliant feature that lets users visualize and tweak their conversations. It transforms how you interact with AI from a simple back-and-forth chat to a creative partnership that feels intuitive to how you’d work with a human editor. It’s hard to not use this once you start!

3. Google Notebook’s podcasting feature – They understood that creators need tools that fit their workflow, not just raw AI capability. The ability to quickly turn conversations into polished podcast content solves a real user need for creators and has downstream impacts for students, too.

These examples show that they aren't just competing on whose model scores higher on an arbitrary benchmark; they're competing on whose product feels most natural and empowering to use.

Other ways to build a moat with UX

Beyond simple UI, there are several other strategies builders can leverage to create a lasting competitive advantage:

1. Context – The more data and insights you gather about your users, the richer and more relevant the experiences you can create for them. Leveraging extensive context allows your AI to output better responses. I’ve spent months building a Project with context on Claude and it’s now really hard for me to use anything else.

2. Personalization – Deeply understanding people's preferences and behaviors enables you to tailor experiences specifically for them. Highly personalized AI experiences make your product uniquely valuable, creating a switching cost that folks are reluctant to give up.

3. Integration and ecosystem – Embedding AI seamlessly into tools your customers already rely on can make your product indispensable. A well-integrated ecosystem reduces friction and makes people less likely to want to leave, further raising the barrier for competitors. This is one thing Google is doing well and Apple is not doing at all (a huge risk IMO!).

Crossing the chasm requires great UX

The early adopters who flocked to AI tools in the beginning? They'll tolerate complexity. They'll figure out how to wrangle prompts and navigate confusing interfaces if the power is there.

But mainstream users? Not a chance.

Geoffrey Moore's "Crossing the Chasm" theory has never been more relevant than it is for AI products right now. The pragmatists who make up the early majority won't adopt AI tools unless they're easy to understand, reliable, and integrate seamlessly with their existing workflows.

Look at how Claude emphasizes simplicity in its interface, or how Copilot is being integrated directly into Microsoft products people already use. These aren't accidents; they're deliberate strategies to make AI accessible to the next wave of users.

What this means for Builders

If you're building AI-powered products or integrating AI into existing tools, take note: your competitive advantage isn't just about which model you're using it's about the experience you're creating around it.

The teams winning in AI aren't just picking the best model, they're asking:

How do we make the tech fully disappear into the background?
How can we reduce cognitive load for our users?
What unexpected moments of delight can we create?
How do we design for both first-time users and power users?

And crucially, they're designing for flexibility. As new models emerge and capabilities evolve, your product architecture needs to adapt without disrupting the user experience.

The bottom line

As builders, we often get seduced by technical challenges. But remember: most people won’t understand the model behind the curtain; they care about what they can do with your product.

The winners in the AI race won't necessarily be those with the cleverest algorithms, but those who create the most intuitive, empowering user experiences. They'll be the ones who make advanced technology feel natural, even obvious.

So next time you're making product decisions, ask yourself: "Are we building for other AI enthusiasts, or are we building for everyone else?"

Because that's where the real opportunity lies.Not just in crossing the AI chasm but in building the bridges that help others cross it too.