Joannes Vandermeulen: How UX design copes with the AI revolution and the "design winter"

CN
1 hour ago

Author: Techub News Compilation

Introduction

In February 2025, at the "UX Beers" event held by the Belgian design agency Namahn, senior design consultant and Namahn co-founder Joannes Vandermeulen delivered a speech titled "How Data, Computing Power, and Algorithms Impact User Experience." In the face of increasing global uncertainty, ecological crises, geopolitics, and intertwining technological revolutions, Vandermeulen candidly stated that the user experience (UX) industry is experiencing a "design winter." Drawing from over twenty years of industry observations, particularly the disruptive impact of artificial intelligence (AI), he pointed out possible breakout directions for designers — shifting focus from pixel and process design to conversational AI design. This speech was not only a profound diagnosis of the current state of the industry but also a forward-thinking reflection on the transformation of future design paradigms.

Summary

  • The UX industry is undergoing a "design winter," caused by factors including the commodification of design, increased client maturity, AI drawing all attention, and questions about the contribution of some digital products to social welfare.
  • AI's "creativity" and "statefulness" (memory of context) are revolutionary, and designers must take them seriously, placing human-centered AI at the core of their concerns.
  • The future focus of UX design will shift from screens (UI) to conversational design, involving the design of AI personality, role, behavior, and the exchange of control between humans and machines.
  • In response to changes, design agencies need to reposition (such as focusing on complex decision support and infrastructure), while individual designers need to learn new skills, including understanding AI technology principles, mastering conversational design tools, and enhancing emotional intelligence and cultural sensitivity.

Facing the "Design Winter": The Current State and Challenges of the UX Industry

Joannes Vandermeulen clearly articulated that the term "design winter" encapsulates the general sentiment within the current user experience design industry. He noted that this situation is caused by multiple factors.

Firstly, UX design itself is being commodified. Unlike the early years when practitioners were few, there is now a large number of UX designers, with associated methodologies, courses, and knowledge systems highly standardized and easily spread. While this has propelled industry maturity, it has also led to UX services appearing "ordinary" in the eyes of some clients, even being viewed as a removable "luxury" add-on. He illustrated this with an example of Harley-Davidson's motorcycle sales being halved; under economic pressure, non-core "luxury" consumption is the first to get cut.

Secondly, client maturity is on the rise. Many companies have established internal design teams capable of handling a significant amount of UX work, which poses a direct challenge to design consultancy agencies. Clients are gaining a deeper understanding of design methods and values, and their demands for design partners are increasing.

Deeper reflections come from within the industry. Vandermeulen mentioned that the ethical implications of some digital product designs are being scrutinized. For example, Aza Raskin, the inventor of the "infinite scroll" feature, has publicly expressed regret over this design. This raises the question: does human-centered design sometimes contradict certain inherent characteristics of the digital world? The real contribution of UX design to global welfare needs to be re-evaluated.

Lastly, the most formidable force currently cannot be ignored — artificial intelligence (AI). Vandermeulen described AI as "sucking all the air out of the room," becoming the focal point of all discussions. AI's exhibited "creativity" (often referred to as "hallucination") and powerful content generation capabilities are fundamentally changing product forms and interaction methods.

In the face of these challenges, Vandermeulen believes that design agencies must rethink their positioning. Namahn, where he works, is repositioning itself as a decision consultant in complex circumstances, focusing on two directions: one is assisting leaders in long-term strategic decision-making through systematic design methods; the other is designing user interfaces for real-time control of complex systems (such as factories, airspace, pipelines), focusing on the infrastructure field. He emphasized that AI will play an increasingly important role in both areas, and therefore design agencies must "seriously embrace AI."

The Core of the AI Revolution: Data, Computing Power, Algorithms, and "Statefulness"

To understand the impact of AI on UX, Vandermeulen guided the audience to examine the three cornerstones driving the AI revolution: data, computing power, and algorithms. The scale is astonishing: large foundational model providers are extracting every ounce of information from the internet; hundreds of thousands of chips (such as Nvidia chips), each worth as much as a small car, are being deployed for computation; algorithms manipulate hundreds of billions, even potentially trillions, of parameters. This scale is as unfathomable as the universe.

So, what initially shocked practitioners about AI? Vandermeulen believes that, aside from the "magical" experience that is "advanced enough to seem like magic" and the unprecedented speed of its adoption, a key breakthrough is "statefulness," meaning AI's ability to remember the context of conversations. Computer scientists refer to it as "statefulness," while linguists might call it "anaphora" processing capability. Early AI conversations were disjointed, while current large language models have an expanding "context window" or "memory" that can hold an entire book or even more content. Vandermeulen predicts that in the future, it may be possible to incorporate a person's entire life experiences into the context, allowing conversations to reference exchanges not just from a few minutes ago but also from two years prior, introducing a new dimension of interaction.

Faced with such powerful technology, what should designers do? Vandermeulen pointed out that the foremost task is to familiarize themselves with the concept of "human-centered AI." He recommended resources from the Stanford Institute for Human-Centered AI and mentioned that the European Union's AI Act also calls for the development of human-centered AI, emphasizing "alignment" — ensuring AI's behavior aligns with human needs and values, including safety, explainability, and transparency. He also recommended the local AI nonprofit research organization FARI in Brussels as a resource to understand AI ethics and governance.

Moreover, designers need to pay attention to cutting-edge trends, especially "agent AI." Future AI will not only be conversational partners but also “agents” capable of proactively executing tasks. This indicates a fundamental shift in interaction paradigms.

From Screens to Conversations: The New Frontier of UX Design

Vandermeulen's core assertion is that the next frontier of UX design is not on screens. Screen design (pixels and processes) will not disappear, just as broadcasting continues to exist in the podcast era, but it will no longer be the sole, and could potentially no longer be the primary focus. The new battleground is "conversational AI design."

This means that the design object shifts from visual interfaces to the personalities, roles, and interaction patterns of conversations. Designers need to consider: What role do you want AI to play? Is it a teacher, coach, student, peer, friend, opponent, or challenger? The market has already seen the emergence of "AI friends" with tens of millions of users, creating a massive new field.

How do you design a good personality for conversational AI? Vandermeulen posed a series of questions: How do you want it to treat users? Should it provide flattering feedback like ChatGPT ("That's a great question!"), or be more calm and pleasant like Anthropic's Claude? How do you establish trust? How does AI communicate the user's reasonable expectations of it? In human-machine conversations, how is control exchanged and shared? Is it a "co-pilot" mode, or can AI take a more proactive role in questioning and actions (many AI assistants have recently become more "bold")? This involves extensive manual fine-tuning rather than a fully automated process.

To excel in conversational design, designers can draw inspiration from other disciplines: sociolinguistics (studying language use in social contexts, related to cultural sensitivity), affective science (a branch of psychology), pragmatics (examining how context influences semantics), and psychotherapy. He mentioned the historically famous chatbot ELIZA, whose simple rules based on Carl Rogers' psychotherapy Q&A model once captivated people as it consistently kept the conversation going through questions. Additionally, a sense of humor (understanding context and delivering amusing jokes) is a hallmark of advanced intelligence and poses a significant challenge to AI conversational design.

On the tool level, Figma is obviously not suitable for this. Vandermeulen pointed out that foundational model providers like OpenAI, Anthropic, and Google have tools to shape AI personalities, one of their core methods being "reinforcement learning from human feedback." Currently, there are thousands of human trainers using complex tools to fine-tune models, giving them a more friendly and conversational "personality". He predicts that in the next 1 to 4 years, simplified versions of such tools will emerge in the market, allowing UX designers to participate directly in shaping the personality of conversational AI.

Q&A Highlights: Skills, Paradigms, and the Future of Education

During the Q&A session, the audience engaged in deep discussions with Vandermeulen, further expanding on the speech's themes.

One audience member raised the point that current UX designers serve as a bridge between the technical (engineers) and human needs. As the interaction paradigm shifts from "human-computer interaction" to "humanoid interpersonal interaction" (conversation), will the types of practitioners change in the future? Is there a greater need for emotional intelligence (EQ) rather than traditional design skills? Vandermeulen agreed, stating that conversational design indeed requires higher emotional intelligence than visual design, and the required skill sets and sources of inspiration will shift. However, he also emphasized that the tension between designers and engineers (technical implementers) will always exist and is healthy, as engineers' creativity can also inspire new ideas.

Regarding the trajectory of technology development, one audience member drew historical comparisons: new technologies often start with text interfaces (like terminals) and later develop into graphical user interfaces (GUIs). Is investing heavily in "humanized terminals" (conversational interfaces) worthwhile, or should we anticipate the next paradigm shift? Vandermeulen believes that multiple modes will coexist and complement each other in the long run. Language predates writing; AI's voice capabilities have already enabled direct conversations. In the future, users will converse with computers through voice, observe content presented, or show real-world scenarios to receive help. The visual system's large information-carrying capacity and clear advantages will not disappear. Therefore, conversational AI design and screen design will have a complementary relationship.

Finally, when asked which seven courses he would include if invited to redesign a syllabus for design schools, Vandermeulen provided his suggested list:

  • Deep understanding of AI technology principles: Designers don’t need to become statisticians or developers but must be able to communicate effectively with engineers to establish credibility.
  • Tools and practices for conversational AI design: Master the methods and tools for shaping AI personalities.
  • Philosophical and ethical thinking: Covering deep topics like consciousness and the impact of AI, helping designers form their own viewpoints.
  • Continuously follow the latest tools: Instructors need to update the curriculum annually, familiarizing students with rapidly evolving new tools like Claude and Notebook LM.
  • Design ethics: This is an essential core course.
  • Role creation and narrative: Learn from the film and theater industry about character creation, scriptwriting, and drama studies, used to build AI personas.

Vandermeulen concluded with a predictive article he wrote in 1994, reminding everyone that while predicting the future is easy, taking action at the right time is incredibly difficult. The world is highly unpredictable, and technology is evolving rapidly; what we see today is "the worst AI ever," because it will only get better. For UX designers, embracing change, continuously learning, and boldly stepping into the new realm of conversational design may be the key to navigating the "design winter" and welcoming the next spring.

免责声明:本文章仅代表作者个人观点,不代表本平台的立场和观点。本文章仅供信息分享,不构成对任何人的任何投资建议。用户与作者之间的任何争议,与本平台无关。如网页中刊载的文章或图片涉及侵权,请提供相关的权利证明和身份证明发送邮件到support@aicoin.com,本平台相关工作人员将会进行核查。

Share To
APP

X

Telegram

Facebook

Reddit

CopyLink