People Say Taste Matters in the Age of AI. So WTF is Taste?

You’ve seen the posts. “Everyone can build now. The only moat left is taste.”

Okay, but what is taste? Everyone says it matters but almost no one explains what it actually is or how you get it.

WTF is Taste?

Taste is discrimination. It’s knowing the difference between good and not good. Between alive and dead. Between not quite there, finished, and overdone.

It’s knowing you shouldn’t one-shot a video of a dog on a skateboard and call it content. Or that the landing page AI created for you is just step 1, and now the real work begins.

Taste applies to everything we do – copy, design, and product.

It’s the feeling you have when you read a sentence and it feels clunky even though it makes sense. Or when a design feels off, even when the layout seems standard and everything lines up.

Sure, good is subjective. I’m not suggesting there is some universal standard you’re unaware of.

But we can agree that there is good copywriting and bad copywriting. That there are beautiful and ugly designs. And that some user flows will lead to more goal completions than others.

AI produces competent work – work that checks all the boxes, even looks good to the layman.

Having taste is recognizing the difference between work that just checks boxes and work that is worth shipping.

How to Develop Taste

How do you improve taste? How do you learn to spot the difference between good and bad work?

You need to spend time in the following areas.

Aesthetics

Aesthetics is a design that looks good. Think of a design or mock you looked at recently – maybe something you made with AI recently. Something was probably off. You felt it, even if you didn’t know the text was too small or the gutter between components was a couple of pixels off.

That’s taste.

You don’t need to be a designer to pick up on these things. You just need to understand a few principles.

Rule of thirds. Divides a space into thirds (think a tic-tac-toe board). The idea is to place the subject or focal point along these lines or at their intersections. My approach has always been to use 2/3 of an area, which satisfies the rule of thirds and ensures there is white space.

Use of white space. White space is “empty” space – but it’s not empty or dead space. It’s space that gives your copy or components room to breathe. White space enables you to create a focal point whereas cramming things together feels chaotic and unfocused.

Reading patterns. People read left to right, top to bottom, in an F or Z pattern. You want to think about this when placing your most important elements.

Color. Colors can evoke emotion. They can create contrast. They can make words or CTAs either stand out or blend in.

Alignment. Imagine a layout on a grid. All your components and copy should line up. This doesn’t mean everything lines up from top to bottom of the page. Different sections might have different layouts. However, every section will align with the broader design – i.e., align with the outermost container.

Again, you don’t need to be a designer. You just need to understand these principles well enough to develop the sense for when something is off. Once you do, you’ll notice that most things AI produces – one-shot outputs or outputs without context or training – are bland.

Logic

Logic is whether something makes sense. Whether it’s intuitive.

Not just visually – behaviorally. You click a button, something happens. You land on an offer page, there’s a clear next step. You sign up for something, you get the thing you signed up for.

When that breaks down, you feel it. You clicked a button and nothing happened. You found the offer but it wasn’t clear how to buy it. You signed up for a SaaS and spent ten minutes trying to figure out how to start.

When logic breaks down, you create confusion. Confusion leads to frustration and fear. Both lead to churn and abandonment – bad words for business owners.

How do you develop logic?

First, use things and think about why they work. Not just apps on the internet. Flip your light switch; turn your car on; preheat the stove; make a pot of coffee; turn your TV on and change the channel.

What did you expect when you flipped the light switch? When you turned the key (or pressed the “on” button) in your car? When you hit the power button on your TV remote? You expected something to happen – likely something to turn on or start.

This expectation is universal. When someone sees a button, they expect that they can press it and something will happen when they do.

Second, talk to people. You’d be surprised at how people use software or tools differently, or how they thought something should work. You’ll see this firsthand if you review findings from a heatmap test.

Third, think about the user journey. What problem are they trying to solve? What steps do they need to take? What can you do to make each step feel obvious rather than like work?

Goals

Goals cover a wide spectrum of outcomes. They made a purchase. Opted in to your newsletter. Completed your onboarding flow. Created their first task in your SaaS.

There is overlap between logic and goals, but it’s possible for something to be logical but fail to hit a goal.

Here is an example:

You have a sales page. The sales page is solid – speaks to the user’s pain points, shows them how the tool solves their problem, answers their questions about price and features, and has a big “sign up now” button. But the page also has tons of links to other pages and a full navigation bar.

Logical? Sure – but will it convert?

It will, but not as well. Every extra link and nav item is an exit ramp away from the actual goal – to sign up for the product.

One more example:

Say someone signs up for your SaaS. You have two distinct user types – someone using this kind of SaaS for the first time and someone migrating from another platform.

What do you build? One onboarding flow or two? Multiple paths? How do you make the next step obvious for one user without making it obscure for the other?

Taste is knowing how to weigh those needs, prioritize the one that drives revenue or retention, and then build a version that doesn’t sacrifice one user for another.

Why Taste Matters More Now

Taste has always mattered, but the stakes are different now because output volume has exploded. Thanks to AI, it’s never been easier to create and build. The challenge is that what AI produces is competent. Okay. Bland. Boring. Average.

Are you okay with average?

Taste is knowing the difference between average and great. And taste is not a skill or intuition that only some can have. You can develop it by studying what makes things aesthetically pleasing, logically sound, and effective at driving goal completion.

The next step is deciding what to do once you’ve spotted average and know what good actually requires.

That’s on you.