“A Good Baseline”: IS UX Writing Science or Art?

In the last 10 or so years, UX Writing has gained visibility and recognition as a discipline – and with that visibility has come a growing tension around how we see our craft. Is it more a creative task that is about empathizing with the user and addressing their emotions, all while, of course, following certain rules – or is it a technical, almost engineering-like task that primarily follows certain rules, and creativity is only the cherry on top?

And I know what you think:

“Both.”

Or “Well, sometimes it’s more the first, sometimes it’s more the second.”

But you know what? I won’t let you get away with that. Not today!

Because finding a clear answer to that question does matter, not just philosophically, but practically: It affects how we teach, hire, and evaluate UX Writers, and how we explain and advocate for UX Writers within product teams. It shapes whether our work is treated as strategic or ornamental, as intuitive or evidence-based. It shapes how we talk about UX Writing, what we see as good and bad UX Writing, or put differently:
It literally affects everything about our job.

And sometimes, the debate is inflamed not necessarily by well-balanced, rather academic discussions, but by bold (yup, let’s say “bold”) public statements like this one made by Carl Rivera, CDO at Shopify, which drew heavy criticism from across the UX community.

“I want to get away from terms that make our craft more science than art. AI enables anyone to make things usable: our job is to make them unforgettable. We hire for taste. For aesthetics. For a point-of-view. Anyone can generate a good baseline.”

This statement was mostly criticized for suggesting that AI has leveled the field of usability to the point where craft is reduced to taste, and that basically good UX can be generated by anyone. With that, it downplays both the complexity of conceptualizing digital systems and the expertise required to do it well. And most importantly, it implies that what makes our work valuable lies in aesthetics, not in function.

I personally deemed this statement dangerous, as it (and similar statements, too) might encourage decision-makers to wrongfully devalue UX work to something that can be outsourced to AI already, while the truth is that these tools are still far from understanding context, consequence, or user psychology.

However, is everything about Rivera’s perspective really wrong? Aesthetics do play a role in digital experience and brand building. And yes, there is a real danger that as AI becomes more integrated into product development, many digital products will start to look generic and similar.

And for that, again, what is UX, at its core?

Science or art?

Let’s dive deeper.

What Is Art? What Is Science?

To evaluate whether UX belongs more to the realm of art or science, we need to understand what each of these terms means. Looking at various definitions of the two, we learn that:

Art is, fundamentally, a form of self-expression. It is subjective, interpretive, and emotionally driven. It often exists for its own sake or to convey a personal or cultural experience. Its meaning shifts depending on context and audience, and it is judged by criteria such as originality, aesthetic value, and emotional impact. The personality and human experience of the author plays a huge role here. Art is also here to be thought-provoking, emotionalizing, and mood-shifting.

Science, by contrast, is the structured pursuit of understanding reality (as well as we humanly can) and developing solutions in a structured way. It relies on research, testing, analysis, and repeatable methods to uncover patterns and produce usable knowledge. Scientific thinking is systematic and guided by a commitment to evidence. Subjective viewpoints and biases must be uncovered and understood in order to create reliable solutions.

Reading those definitions, at least for UX Writing, the verdict is clear: UX copy is not written to express the writer’s vision, content is not designed to express the content designer’s perspective on life. Not at all. It is created to solve a problem for the user.

Let’s get this straight for a moment.

UX Writing Is Not About Self-Expression

The role of a UX Writer is not to showcase personal voice or creativity. It is to guide, support, and inform - often in high-stakes, high-friction, or time-sensitive moments. When a user resets a password, grants access to personal data, or tries to understand a pricing change, the language we provide must do more than sound good. These moments require cognitive knowledge, linguistic accuracy, awareness of technical restraints and dependencies, backend logic, and design requirements – not necessarily something like poetic instinct.

Heck, not even a broad vocabulary.

And most importantly, when done right, users won’t notice our writing. Just like in UX design, the results of our work come together to create an overall seamless experience.

No art to find here.

Right?

Art in an Artless Craft

With that being said, no, neither UX Design nor UX Writing are a form of art. Not even when we bend the definition of our crafts by a 180° to fit into Rivera’s apparent understanding of them. However, that doesn’t mean that UX is free from art entirely.

There is, for example, the importance of being aware of the current Zeitgeist: Zeitgeist, which according to the Cambridge Dictionary literally translates to “the general set of ideas, beliefs, feelings, etc. that is typical of a particular period in history,” does play a certain role in UX: it makes the products we build visually and verbally appealing to a broad audience in the now.

Human emotions, too, are essential ingredients in both art and UX: The words we write, the shapes we choose, the imagery we go for – all make the user experience, and every user experience is, at its core, a human experience – no matter what we choose to call it. And we can’t and shouldn’t take the human element out of our craft and reduce it to the application of mere analytics or mathematics.

In these regards, UX and art indeed have some aspects in common. But that only makes the comparison more dangerous.

The Empathy Trap – And the Danger of Framing UX Writing as “Art”

“Hiring for aesthetics” means you hire someone for building something aesthetically pleasing. Or put differently: Something the two of you - the hiring person and the hired person – both find aesthetically pleasing. A shared understanding of aesthetics, if you will.

There are many ways in which this approach becomes problematic: First, just because a hiring manager (or CDO) finds something aesthetically pleasing doesn’t mean the end users do. Second, just because something is aesthetically pleasing doesn’t make it a good experience (remember: seamlessness!). Third, putting aesthetics on the front stage of design and writing makes it impossible for teams and stakeholders to give and discuss feedback - let alone explain design or writing decisions. Under this premise, discussions in teams will look exactly like this:

“I designed it like this because it’s aesthetically pleasing.”
“Well, for me it’s not.”
“Okay, but for me it is.”

And consequently, user tests (if there even are some) will probably come down to NPS-ish “Do you find this aesthetically pleasing or not?”

The problem here can be summarized in one simple word: Subjectiveness. Subjectiveness disguised in a seemingly universally applicable principle - aesthetics.

And unfortunately, this problem is not new to the UX community. So far, it has only gone by another name: empathy.

(Hear me out on this one!)

We often hear UXers of all disciplines claim that our job is much about having empathy, and that empathy is a core skill of a great UXer.

If you ask me, nothing could be further from the truth.

According to the Merriam-Webster website, empathy can be defined as “being aware of and sharing another person's feelings, experiences, and emotions.” Sounds important for UXers, right?

Kind of. If only we were actually able to feel our users’ feelings accurately, which we’re probably not. If you work in UX, chances are you have a lot of experience with digital products. You know about typical interaction patterns, you know what a burger menu is and what happens when you click or tap it. That alone makes it hard for you to empathize accurately with your users.

But there’s more: How old are you, and how young or old is your youngest or oldest user? What’s your ethnicity? Your range of cognitive, physical, and mental abilities? One thing is certain: You can’t share all characteristics of all users.

That alone makes empathy a problematic guiding principle: Science has shown that we tend to empathize more with those who are similar to us – a risk summarized by the classic UX warning: “You are not your users.”

Therefore, empathy, just like aesthetics, is too subjective to serve as a North Star for UX professionals.

Conclusion: Art Can Refine, but Science Must Lead

Rivera’s vision of unforgettable experiences is ambitious, but it has very little to do with good UX practice. As strong as the urge might be for all of us to outsource the boring parts of product building to AI, pushing our expertise into the background to focus on aesthetics, in an overall mission to leave the science of UX to the machines and make our job about arts – that’s far away from the reality of both UX teams and users. It would hurt the collaborative nature of our work, it ignores the swarm intelligence of interdisciplinary product teams of designers, writers, devs, and researchers - and it heavily – I mean heavily – overestimates the abilities of any AI tool currently on the market.

Moreover, we need to keep in mind: what we find aesthetically pleasing, or what we think is “good art”, is not neutral. It’s shaped by cultural norms, generational values, and even social power structures – often without our awareness. What is considered elegant, witty, or “cool” to one may be confusing, weird, “cringe”, or even discriminatory to others.

If we prioritize aesthetics, we risk designing for people who look, speak, and think like us - and excluding those who don’t. This can lead to ageist, ableist, classist, or culturally biased patterns that undermine both inclusion and usability.

Here, scientific thinking helps guard against this by grounding decisions in behavior and evidence, not assumptions, intuition, or personal taste. That doesn’t mean that you can’t sprinkle some creativity on top of your solid user experience – it just means you need the expertise to decide where it adds value – and where it doesn’t.