User Experience and Artificial Intelligence
UX design remains at the core of all digital service development. But how has user experience design actually changed over the past decade? And what role does artificial intelligence play in all of this? Here is a look at how the field has evolved and what to expect from its future.
It is rare for a technological innovation to conquer the world overnight. Breakthroughs often mature for years within expert circles before reaching mainstream popularity. This applies to the current AI boom as well. Although OpenAI's journey began in June 2018 with the release of its first Generative Pre-trained Transformer model, GPT-1, it was still an academic proof of concept at the time. The real explosion, which by the most colorful estimates changed the course of humanity, occurred four years later on November 30, 2022.
On that day, the company released ChatGPT, built on the GPT-3.5 architecture. The secret to its success lay in its exceptionally accessible and intuitive interface – in other words, its user experience. Suddenly, AI was no longer just a sci-fi antagonist or a data scientist's tool, but a tangible, conversational assistant that reached over 100 million users in just two months, making it the fastest-growing consumer application in history.
This explosive growth triggered an intense arms race among tech giants, with billions of dollars invested in developing ever larger and more powerful models. It came as a complete surprise to many when DeepSeek, a Chinese startup founded in July 2023, released its own open-source model. It was able to compete in performance against American giants, despite being developed at a fraction of the cost (under 6 million dollars) and with less powerful hardware, such as NVIDIA's H800 chips.
As the technical performance of AI models has converged, the focus of competition has shifted fundamentally. It is no longer enough to achieve the best results on standardized benchmarks. What matters now is the holistic user experience: how quickly the model responds (latency), how much information it can process and remember (context window), and above all, how reliable and error-free it is.
This strategic shift became evident in August this year when OpenAI released GPT-5. Unlike previous updates, its biggest leap was not in raw power but precisely in usability. GPT-5 introduced a unified, adaptive system that uses a real-time router to automatically select either a fast base model for simple queries or a deeper, more resource-intensive “thinking” model for complex reasoning. This removed the need for users to manually choose between different models – a classic UX problem that had plagued earlier versions, and one that even OpenAI CEO Sam Altman criticized as too complicated.
At the heart of the release were also AI agents designed to perform complex, multi-step tasks autonomously. Instead of the user guiding the AI step by step, the agent can set up its own workspace and use a browser to carry out tasks. For example, you could ask an agent to plan a wedding, and it could search for a suitable venue, compare suit rentals, and draft invitation cards on your behalf.
Will AI Replace UX Designers?
The discussion about AI-driven job losses is not unfounded. In the first half of 2025 alone, technology companies laid off nearly 94,000 employees, and many roles are threatened by the efficiency AI brings. Tech giants such as Amazon, Salesforce, and Google have all carried out significant reductions. For example, Salesforce cut 1,000 jobs while actively hiring new employees for AI-focused roles. Amazon CEO Andy Jassy stated openly that the company expects AI to reduce headcount in the coming years.
The story is not that straightforward, however. For instance, Ikea deployed an AI chatbot called Billie for customer service, but instead of simply laying off employees, it retrained thousands of call center workers as interior design advisors. This provides a vital counter-narrative and demonstrates that change can also mean the evolution of roles, not just their disappearance.
At the core of UX design are skills that current AI models find difficult, if not impossible, to replicate.
- Empathy and qualitative understanding: A designer's work begins with a deep understanding of user needs, motivations, and frustrations. This requires the ability to put oneself in another's shoes – a skill based on subjective, lived experience, not statistical probabilities.
- Strategic problem-solving and communication: A UX designer acts as a bridge between users, business goals, and technical constraints. This requires negotiation, persuasion, and strategic thinking – advanced human skills.
- Ethical judgment: In the age of AI, the importance of ethical questions such as privacy, non-discrimination, and transparency is greater than ever. UX designers are in a key position to act as user advocates and ensure that technology serves people responsibly.
AI will not replace the designer, but rather transform their role. As routine tasks such as wireframing and data analysis become automated, the designer's value shifts increasingly toward strategic thinking, in-depth research, and ethical guidance. AI commoditizes the mechanical aspects of design, compelling professionals to focus on their uniquely human skills.
The use of AI also involves fundamental, systemic risks. One of the most well-known is hallucination: a phenomenon where a language model produces content that sounds convincing and grammatically correct but is entirely inaccurate, nonsensical, or fabricated. This is not an “error” in the human sense but a consequence of the model's probability-based nature – it predicts the most likely next word, not truth. Because the output is presented confidently, it can easily lead to misinformation spreading in critical fields such as healthcare and finance.
An even more serious long-term threat is model collapse. As the internet fills increasingly with AI-generated synthetic content, future AI models will inevitably be trained on their own existing outputs. This creates a degenerative cycle: first, models begin to forget rarer information and the “tails” of data distributions (early collapse), and eventually their output becomes increasingly generic, mediocre, and erroneous, until their quality finally collapses (late collapse). Without careful preservation of human-curated data, AI can essentially cannibalize itself and consume its own future.
UX Design Trends
Let us leave the existential reflections behind and look at the current trends in user experience design. Although it may sound cliché, all design decisions in service development reflect broader societal developments.
One of the most significant current trends is the dramatic decline in human attention span. According to the pioneering research of UC Irvine Professor Emeritus Gloria Mark, the average attention span on a single screen has dropped from 2.5 minutes (150 seconds) in the early 2000s to just 47 seconds today.
Attention span on a single screen
| Metric | Period | Duration |
|---|---|---|
| Average time spent on a screen | Early 2000s | 2.5 minutes |
| Average time spent on a screen | 2023 | 47 seconds |
| Median time spent on a screen | 2023 | 40 seconds |
This “47-second reality” forces designers to strip away everything unnecessary and design for cognitive efficiency. The key trends of 2025 are direct responses to this challenge:
- Instant value: A service must deliver its benefits immediately, without friction or waiting. TikTok's auto-playing feed and one-click checkout processes in e-commerce are prime examples. Every extra second or button press risks losing the user.
- Strategic minimalism: Clean and intuitive design is no longer just an aesthetic choice but a functional necessity. It reduces cognitive load and helps users focus on the essential task within their limited attention window.
- Hyper-personalization: With AI, services can tailor content and recommendations to user preferences, increasing relevance and engagement. Netflix recommendations and Amazon product suggestions are classic examples.
- Accessibility is non-negotiable: Legislation such as the European Accessibility Act (EAA), which came into effect this year, makes accessibility a legal requirement.
The Future of UX Design in 2035
I have now worked in user experience for about a decade. It is surprising that despite constant change and the growth of AI, the field still looks very much the same as when I started. This is not self-evident: while a website designed in 2015 could well still be acceptable today, the same could not have been said about a site built in 2005 just ten years earlier.
This is largely due to the smartphone becoming established as the primary usage environment. The shift from desktop-centric (2005) to mobile-responsive (2015) design was a fundamental paradigm shift. Now that the foundational structure of digital services has matured, the real competition is increasingly about a deeper understanding of human needs.
I cannot say, however, how long this plateau will continue. Will future UX design involve AI designing services for AI? And what role will virtual reality (VR) and augmented reality (AR) play? If I had to guess, I would see AR having a greater impact on the field. Virtual reality's ability to simulate environments is impressive, but clunky hardware and its reality-isolating nature likely make it a “niche technology” for specific purposes like gaming. AR, on the other hand, layers digital information over the real world, using smartphones, earbuds, and lightweight glasses. It is easy to imagine that, especially combined with AI, this could — for better or worse — come to affect our daily lives a great deal.
All in all, we live in interesting times. It may be that the most successful services of the future will emerge from a symbiosis in which human empathy, creativity, and ethical judgment guide the machine's vast computing power to create services that are not only functional and efficient but also genuinely human and meaningful.
