UX's core mission is to advocate for change. From the common double - diamond to more complex models such as the analysis - synthesis bridge model, we view our job as bridging the chasm between the current "problematic" state and the desired "solution - oriented" future state.

 

However, that's the relatively straightforward part. The real challenge lies in actualizing that change.

 

Although our design tools grant us full control over graphical elements like rectangles, designers lack such power over people. "No one listens to us" is perhaps the most persistent grievance in the field. Ironically, the design field is moving in the wrong direction. By offloading the task of creating persuasive arguments to AI tools that are ill - equipped for this purpose, we are becoming less and less persuasive.

 

Technology won't rescue us here. Designers aiming to break free from merely confirming stakeholder biases must extend beyond STEM and start devising compelling arguments.

 

Since the 1960s, long before the term "UX" emerged, we've known that to address and solve the complex problems that are the very reason for design's existence, designers not only need to identify an issue but also to clearly state their position and back it up with solid argumentation.

 

The argument is crucial because the value of UX hinges on convincing stakeholders to act differently from what they initially intended.

 

Work that fails to achieve this has no value. Charles Lambdin illustrates this through the Sure Thing Principle: if our decision remains the same before and after research, then the research effort is wasted.

 

Let  be a decision and  be research. If , then  is waste. An interesting implication is that the more a decision changes, the more valuable the insight. We'll revisit this later.

 

If our audience doesn't require persuasion - if we're only validating what stakeholders already believe - we aren't providing new understanding. If our audience needs convincing but we fail to do so, then our insights won't be implemented, and the outcome is much the same.

 

Either way, we haven't added real value.

 

Even in the "design" aspect of UX - defining how a product functions rather than uncovering the why - we still have to persuade stakeholders. They'll never perfectly replicate the system we envision in their minds. Hence, we use diagrams, models, flows, and presentations to convey our ideas. Most of our outputs describe the user experience, yet the user isn't the intended audience.

 

These descriptions must be persuasive because design isn't the ultimate decider of what gets shipped. In the standard "three - legged" model of ownership, "feasible, viable, and desirable" aren't independent. They're always in conflict. There's no such thing as "infeasible" without the cost - benefit analysis of "viable" - a highly profitable business opportunity can justify unlimited technology investment.

 

If design can't make the argument that a great experience justifies extra effort from the other two "legs," our impact will be limited to the fringes of the business model and the technology stack.

 

Fortunately (or unfortunately), this doesn't mean being the loudest in meetings. An argument must work in various scenarios - some for when we're present, and others for when we're absent. To persuade in diverse situations, we first need to design the argument in its purest form - an essay.

 

 

I admit that few things are less popular than essays. I didn't like them in school either. I, like many students, just started typing until I reached the required page count.

 

But we all missed the point. The goal wasn't to write text but to practice constructing an argument.

 

Repeatedly, our teachers told us the formula: state your thesis, present the evidence, and explain how the evidence supports the thesis. Instead of asking the audience to trust us blindly, the essay format guides us to convince readers by showing how we reached the conclusion from the facts.

 

Issue, position, argument.

 

It's no wonder that humanities graduates land some of the highest - paying jobs across industries, often in leadership roles. The ability to analyze texts, extract relevant information, and combine it into a convincing proposal to solve tough problems is fundamental to how companies formulate strategy. If we accept Charles Stross's view that corporations are a slow, procedural AI, then humanities majors are the data scientists training this AI for the desired outputs.

 

It's well - known that a company's structure determines the products it creates. Over numerous decisions and releases, layer by layer, the system optimizes to produce what it's designed to make. When a designer suggests a change to that output, they're pushing against the collective force of a system that's set to produce a particular product.

 

If their main tool for this battle is a Figma file, they'll always lose. All but the most minor changes will be discarded not because they're unfeasible, unviable, or undesirable but because the path of least resistance leads elsewhere.

 

To create a different product, designers must change the existing system. In a corporation, this is achieved through persuading its members.

 

 

Sadly, the UX field, like many others, is moving away from changing the existing system and towards the opposite: speeding along the well - trodden path. Designers are asked to cut short the research process to focus on the "real work."

But this is the real work.

 

Relying on LLMs to speed up the process of gaining insights so you can "do design" is like irresponsible teenagers using LLMs to write mediocre essays faster. You get text in the shape of an essay but not the benefits of writing it. A report that quickly confirms all your assumptions, completely ignoring the potential for new insights.

 

Because an essay isn't just a vessel for an argument; it's a tool for constructing the argument. The most fascinating discoveries and impactful insights come from a hunch that two ideas are related and the urge to explore that connection further. When you can move from "customers do this and this" to "customers do this because of this," you uncover opportunities that others miss. The connection between these two points forms the structure around which everything else logically aligns.

 

This connection is valuable precisely because it's not obvious. Its novelty can significantly alter the decision - making process, but it also demands a more convincing argument. A deliberate approach to treating design as an essay is a catalyst for value - it enhances both the design's innovativeness (by helping us find distant data points) and its persuasiveness (by establishing a strong connection between them).

 

In a specific context, the actions people take to reach a goal may face obstacles - a problem hypothesis. Overcoming these obstacles improves key results, leading to business impact - a business impact hypothesis. A solution provides the means to overcome these obstacles - a solution hypothesis.

 

The hypothesis - driven design approach is powerful because it establishes causal relationships among inert facts.

 

When you use AI to summarize (or worse, generate) transcripts and responses, you're just getting text. An LLM can't draw inferences to discover interesting arguments or make them compelling because it's only shortening the input. It can't "think harder" but only produces a slightly different list of average ideas.

 

A well - developed point of view can never be commodified, and the skills to persuasively express it will always be a worthy investment.