If you’ve noticed that websites have started to converge into a sans-serif beige haze, you’re not imagining it.

One of the biggest advantages of AI has been allowing non-technical people to code their ideas into real, monetizable applications. As we wrote in April, anyone can create an app in a couple of hours, using tools like Claude Code, Lovable, Replit or Base44.

These AI-designed apps have some telltale signs, and the devil is in the details: using similar design styles that look pretty but are dysfunctional. While apps may work on a small scale, these small details could become big problems when scaled up and commercialized.

Here’s how to tell if your app appears AI-encrypted and how to change it.

1. Regression to the mean, also known as painfully mean

The first sign: the app’s design is boring and plain.

Paul Bakaus, CEO of AI design startup Impeccable, said in a June 23 podcast interview with venture capital firm Andreessen Horowitz that AI giveaways, particularly for Claude Design, include beige or tinted backgrounds and sans-serif fonts.

He called it an “algorithmic Uniqlo or Ikea,” a design that’s not bad, but not necessarily unique.

Donghoon Shin, a human-computer interaction researcher at the University of Washington, published a paper on how vibration coding has led to design homogenization.

Shin told Business Insider that vibration-coded products tend to converge toward “a single, statistically average aesthetic.”

The distinguishing features: a muted color palette with lots of whites and grays, a single-brand accent color, standard sans-serif typography, and elements with rounded corners and drop shadows.

Sauvik Das, an associate professor at Carnegie Mellon University’s Institute for Human-Computer Interaction, told Business Insider that this was the “regression to the mean” effect.

People trying to make money with their AI-coded apps are realizing this too.

Priyanshi Bansal, a product manager in India, vibration-coded an app using Claude that helps people choose gifts for their loved ones. Early Users told him the app seemed like “a waste of AI.”

He said the first version of his app featured a lot of emojis, shadows and rounded edges.

“I’m now creating version two of the app, which is much better, particularly the user interface, since I’m a product designer, so I need to nail that down,” she said.

2. Pretty and dysfunctional

The second telltale sign: the website is super slick for a product that’s not yet available.

“You’ll see, for example, a very polished landing page for a product that’s still in an early alpha phase,” Das said.

It may look pretty, but many of the features don’t work intuitively, because AI designs for aesthetics and skimps on usability.

User interface and experience designers are trained to read human emotions, behaviors and intentions, said Ankush Samant, professor of digital innovation and design practice at the National University of Singapore. They know exactly how to design “the weight of a button, the pace of an onboarding flow, the tone of an error message.”

“AI tools are optimized for the happy path and tend to produce interfaces that appear complete until someone actually uses them,” Samant said.

Das added that vibration-coded products sometimes invite users to interact, even when That piece has no function. For example, hovering over an item gives it a slightly shaded outline and a size increase that suggests it can be clicked for more information, but clicking on it gives you nothing.

3. Error 404: Ignore edge cases

And finally, vibration coding tools don’t devote much attention to edge cases.

“Designers also spend a lot of time on what I would call edge state design (empty states, error messages, skeleton loaders, and offline states) which AI often leaves as an afterthought or skips entirely,” Shin saying.

Samant said the way an app bugs is how you can tell if the app’s creator has thought through the entire user experience or if it’s still in a demo phase.

“AI tools tend to skip them entirely or generate placeholder copy (‘Something went wrong. Please try again’) that eliminates the human voice exactly when users need reassurance most,” he said.

No doubt, new vibration coding companies are taking note of all these drawbacks.

San Francisco-based startup Base44 on Monday launched its own AI model, Base 1, which it hopes will reduce the AI ​​look and feel of vibration-coded products and produce better UI/UX than frontier models.

You vibrate coded an app. And now what?

In this house, we don’t shame vibration encoders. We ourselves are big enthusiasts.

But if you find that your app has some of the red flags mentioned above, here are some ways to change them.

Samant said the first step is to stop prompting aesthetics and start prompting decision-making.

“Instead of ‘make this look clean and modern,’ try: ‘This screen is where a user who is anxious about their data will decide whether to continue. What should I delete and what should the copy do?'” he said.

Secondly, vibration coding is not just about vibrations. Shin said users should provide as many details as possible to AI tools, such as design references, brand limitations and what the user doesn’t want to see.

And if you are thinking of expanding it, there is the possibility of hiring a professional to take the product to the next level.

Das said that while AI is a useful tool for improving user interfaces, it may not be the best for creating good UIs from scratch.

“I won’t say that a purely vibration-encoded application cannot achieve sustained commercial success, because the most important thing is product-market fit,” he said. “But I think you always need a good UI/UX designer to take a good product and make it great.”