The many deaths of UX design
How UX was proclaimed dead (again), what do AI and Elon Musk have to do with it, and what poses a real threat to the discipline of design.
tl;dr: UX is a young discipline in flux, which many have pronounced dead, thanks to incompetent leadership diminishing its importance, common misconceptions about its role, and the ever–changing scope of responsibilities.
In this article, I attempt to dissect the role of UX in the world dominated by AI and rockstar CEOs, debunk some common misconceptions about the responsibilities of UX designers, and demonstrate that although UX is not disappearing, it may be facing challenging times ahead.
Design is in danger.
“You don’t necessarily have to be a good designer in order to craft a slick–looking website”, reads a 2015 VentureBeat’s article, covering The Grid, an AI–powered website builder that took off and was quickly sunset afterwards, never to be seen again.
“You don’t need to be a designer to [create graphic assets for a website like a pro]”, writes Podia’s guide and proceeds to list out the tools that help put together mockups of all sorts. “You don’t need to be a designer to look professional”, promises CareerArc’s visual editor.
Sounds like you don’t need to be a designer at all. Period.
Sketch2Code will (soon) turn your drawings into HTML pages, ChatGPT will write your copy, DALE•E will generate the visuals, designs.ai will take care of the miscellaneous. Who even needs mockups when AI can generate entire websites? Worst case scenario, one of almost 3,000 premium WordPress templates will do the job.
Engineers are taking over.
In December of 2022, Twitter added analytics to tweets, showing how many people saw each particular post. The reaction was mixed, with many pointing how it has cluttered the UI, to which Musk responded with a poll, asking Twitter users to choose whether impression count should “stay on left” or “move to right”.
Michael Buckley in his article “How Elon’s Twitter takeover has disrupted the UX industry” calls it “aggressive field-testing” and claims Elon runs “a fast-paced and turbulent design process”. As per Buckley, CEOs are “likely concluding they don’t need a formal UX process to achieve their business objectives”.
It’s hard to say whether it will indeed be the case, but it surely is for the bird app. Ultimately, in a memo to his employees, Musk promised that Twitter will become
much more engineering-driven. Design and product management will still be very important and report to me, but those writing great code will constitute the majority of our team and have the greatest sway.
Public’s opinion on the matter appears to be divided, but people calling Elon’s poll an inquiry into the user’s opinion or an A/B testing, serves as a reminder that design is still perceived as synonymous with aesthetics, a subjective art of “making things look pretty”, and as such — dealing with preferences and tastes, and not deserving a seat at the table. This leads to a conclusion that the imminent rise of AIs will make designers obsolete.
The thing is, design is neither. The key to understanding why design in general and UX in particular remains relevant lies in comprehending two inconvenient truths: designers are not dealing with preferences, and experiments are conducted under controlled conditions.
Designers are not dealing with preferences.
It’s widely believed that designers are busy asking people what they prefer, but haven’t we all heard the fake Henry Ford’s quote about faster horses? Researches have been begging business owners for years: stop asking users what they want.
In 2009, Walmart lost a tremendous amount of money after having launched “an uncluttering project” and asking their customers whether they’d like “Walmart aisles to be less cluttered?”. An obvious “yes” costed the corporation a billion dollars: sure, customers were happy to see clean isles, but the sales quickly went down.
There are good examples when listening to what the customer wants leads to wrong conclusions: take a New Coke or a 1992 Shevy Caprice for example. Sometimes, running focus groups, testing in the usability lab, facilitating interviews may yield misleading results. Sometimes, ignoring what your customers say is the best course of action.
People always prefer status quo. The majority will always remain silent, and getting them to talk is a challenge. Researcher’s bias is an impactful factor that cannot be dismissed completely. In fact, there are a number of cognitive biases researchers are well-aware of: a Sampling Bias, for instance, when the selected users do not represent the customer. Surveying 100 adult males with no allergies to research food intolerances in female teenagers is a pointless endeavour.
Asking users for their personal preference is clearly not what research should entertain itself with — and luckily, it doesn’t. In Mental Models: Aligning Design Strategy with Human Behavior, Indi Yong writes, “Building products based on preference research is like building a kitchen from a stack of magazine clippings”.
By asking people to vote on the matter, Musk has fallen victim to three biasses simultaneously: asking for a preference in a pre–determined matter, asking a closed–ended question, and relying on the opinion of a limited group of individuals, not representative of the user base.
Something an experienced designer could have easily avoided.
Designers experiment.
Designers do not deal with preferences. Instead, they explore, analyse, and runs experiments under controlled conditions.
In Experimentation and Testing: A Primer, Avinash Kaushik writes that “80 % of the time you/we are wrong about what a customer wants”, and explains that experimentation and testing “help us figure out we are wrong”.
Controlled experiments date back to Sir Ronald A. Fisher’s experiments at the Rothamsted Agricultural Experimental Station in England in the 1920s, but remain the only way to test a hypothesis until today. Kaushik proposes three methods, A/B testing among them.
People call Elon’s infamous poll A/B testing as well, but the actual methodology is a bit more complicated than simply rolling out a feature without probing or establishing the need. A/B involves testing a control version against an alternative, occasionally with multiple variants (split, multivariate). Its results, despite obvious limitations, can be quite trustworthy and reliable.
The role of a designer is not to only to plan and implement the testing, but to interpret the results, which may pose quite a challenge. The job is to form a hypothesis based on the problem statement — something we are not yet capable of automating — and analyse the impact on business, product, and users.
So, Twitter should have launched at least 3 versions: one with the analytics in the left corner, one in the right, and one — without analytics, splitting the three evenly across all Twitter’s users. After some time, the team would look back at the results, recruit and interview some users, analyse the difference in the impact of each variant, and derive conclusions.
Meanwhile, someone would have probably noticed that analytics is not available on Mac and is still on the left on mobile. Oh, no, it’s on the right. No one really knows for sure.
So, what is happening to UX?
Nielsen–Norman Group defines the role of a UX designer as that with a “broad range of responsibilities: from designing prototypes to collaborating with subject-matter experts or carrying out qualitative usability tests”, but recently the role of a UX practitioner has become less and less universally defined.
In the red corner, those who think UX as a discipline will soon cease to exist. Back in 2014, Peter Merholz called for the UX’s responsibilities to be split between interaction designers (taking care of structural and interface designs of complex systems) and product managers (responsible for the strategy).
More Paul Wilshaw writes that “in ten years, UX will be dead”, predicting that designers will get closer to the developers and merge into teams of “design engineers”.
In the blue corner, those who see the death of UX (one way or the other) as a transformation. Colin Shaw, a customer experience pioneer, makes a case for CX taking the place of UX, highlighting the important difference between users, clients, and customers. To Shaw, understanding what the customer actually wants is a challenging task, but a key to success: design is not dead, it should revolve more around the customer than a user.
Paul McAleer in his 2015 post states that UX is transforming, outgrowing the fast-paced digital world, and predicts that the skills of UX designers will be needed in other industries:
“We can inform others. We can ensure those without voices are heard and respected and understood. In the end, UX is all about people”.
In the article about Elon’s Twitter, Michael Buckley concludes that UX is “doing the work of product designers to some degree”, and apparently, the merge is already happening, with the expectation from a product designer to “design for an experience for the user”.
Other practitioners seem to observe it, too: whether by claiming that “UX designers should be involved in business aspects more” or by admitting that the borders between product and UX designs are “super fuzzy”.
Either by merging with product design, splitting into multiple disciplines, like it had already happened with UX Writing and UX Research, or staying the same: UX is not vanishing any time soon.
Or is it?
The not so phantom menace.
In 1993, Adam Richardson published a paper titled “The Death of the Designer”, in which he stated that industrial design was in crisis, as designers did not have as much control over product as they claimed to have.
“Designers are part of a team involving others that have done much of the work before the designer begin”, writes Richardson. Form may follow function, but if the designer is excluded from the process of defining the function, their role begins to diminish. We may have reverted it by promoting the idea that design should drive the decisions as much as sales and business do: so much so that engineers began to feel excluded.
Does that mean that the death has been averted?
Well, wrapping up the paper, Richardson casually quotes Clive Dilnot who said that designers do not feel compelled to read philosophy, ethics, economics, social science, or even design history. If designers, who have to deal with all these disciplines, do not bother studying them, what are the implications?
Three decades later, who is to blame that a large portion of designers have been “dribbblified”, and are now designing to “impress their peers rather than address real business problems”? Well, the answer might as well lie in design education.
In the 1920s, Bauhaus changed the way art is being taught forever. A curriculum developed by Walter Gropius and other teachers featured a half–year Preliminary Course, which included theory of colour and form (taught by a famous Johannes Itten), principles of composition, exercises in drawing, and more. Completing a course was mandatory before the student could progress to mastering pottery, weaving, or drawing.
Fast–forward 100 years: Berkeley’s curriculum opens with learning about design thinking, empathy, and user personae, Interaction Design teaches “the importance of UX design”, and the vast majority of courses emphasise good pay, great career growth, and global opportunities.
Bauhaus wanted to nurture a generation of creators who would change the world, reflect the unity of all arts through total education and cross–discipline learning. Today’s UX designers are educated to put together personae and ask 5 “why” without having to learn anything beyond their immediate scope of work.
In 2021, 44,254 students graduated from 100 US–based bootcamps alone: it’s hard to calculate how many studied UX specifically, but Singapore’s own General Assembly boasts about having over 97,000 alumni globally.
The bootcampisation of design is an imminent threat that looms over the industry. UX Bootcamps model is fundamentally broken: built around generating revenue, it expands rapidly, promoting a quick and easy way into UX.
People who taught as such courses state: they are not always what they are advertised to be, some go to such extremes as to call bootcamps a scam straight away. Success stories of all kinds will mention graduates landing jobs, earning a 6-figure salary, and at the end of the day, being a happily employed UX professional, but how many will mention an outstanding curriculum or a wide range of disciplines taught?
Can greatness be expected of bootcamp graduates? Of course: bootcamps are merely a gateway, an introduction to design. They do, however, create an image of UX as an easy, straightforward craft one can manage in 4 to 16 weeks, thus contributing to the idea that unlike engineers, designers deal with matters that require nothing but common sense and a little bit of taste. As such, designers seem easy to replace: with an AI, a UI Engineer, or a manager prioritising features at a grooming session.
Design Thinking, Double Diamond, User Centred Design: designers are taught that every project is unique, but in the end, 90% of portfolios end up looking the same, and LinkedIn is riddled with “Which UI is more accurate”. In the end, new UX designers are out in the wild, looking for jobs with their portfolios filled with perfect cases: barely able to explain what double diamond is (and isn’t), knowing next to nothing about how computers work, but having classy personae and slick Dribbble–worthy UIs on their Webflow personal pages.
There’s still hope.
That all being said, UX is far from being dead. Some of its aspects are threatened by AI (although it’s not going to steal your job and burn your house), some by incompetent leadership, it may need a rename, a rebranding, a restructure. However, as a young discipline that doesn’t know yet what it is, it is more than alive.
UX is in flux. Unicorns are almost extinct, design systems are taking over, and the job is getting harder. The future of UX education must ensure that it combines the right amount of practice with the right amount of relevant theory — by standardising the definition of UX and a set of required and desired skills, developing knowledge beyond UX tools, and providing practical experience that’s relevant to the world.
Incorporating UX in other academic disciplines could be the way to go: UX designers coming from or transitioning to other disciplines could become domain experts capable of conducting research, ideating, designing, testing concepts.
UX is inherently multidisciplinary, it can be adapted to fit almost any discipline that deals with communication, technology, design, or psychology.
To survive, UX needs a different type of designers: passionate, informed, well–rounded. UX designers with decades of experience and current bootcamp students should srive to develop a broad mindset, go above and beyond, master and challenge the frameworks, learn and bend the rules. Beyond the realisation that design is far more than your daily craft lies the key to salvation for the entire industry.
Building a stronger cohort might change the general perception as well: from pixel–pushers and beautifiers to strategists, researchers, human behaviour specialists, in charge of product success and customer satisfaction.
People who come from other backgrounds and bring domain knowledge, and professional designers who take great interest in culture, philosophy, and history, consider local context, contribute to it. People who are equally concerned with the customer and the business.
People who have earned the seat at the “big kids table” for the entire industry.
