AI Redefines Design: Towards a Revolution in Creative Practices

A Changing Practice

Design is undergoing a profound transformation, driven by advancements in artificial intelligence. Creation processes are significantly shortening, and many tasks once reserved for junior designers are now automated. Speed has become imperative, and time-to-market is shrinking as AI technologies progress.

AI facilitates prototyping, making this stage accessible to a broader audience. Some experts even predict that the practice of design could disappear, viewing it as redundant in a world where interfaces have become commodities. However, rather than fading away, AI offers a unique opportunity to return to the fundamentals of design: enabling users to achieve their goals with satisfaction. The current challenge is to focus on essential outcomes while using these new tools to honor the fundamental promise of design.

A Bit of Context

Over the past decade, design has increasingly been associated with user interface (UI) and visual aspects. Yet, these are merely superficial layers, representing communication and part of interaction. Since the emergence of human-computer interaction (HCI), the main challenge has been enabling humans to interact with machines, which were then limited.

Design's mission was to translate the complex language of machines into understandable metaphors, such as desktops, folders, or files. These concepts, although simplified, facilitated human interaction with technology. Computers do not actually operate this way, but this simplification made technology accessible and intuitive.

A Complex and Holistic Discipline

Design is both a complex and holistic discipline. With the increasing complexity of digital products in the 2000s and 2010s, the discipline fragmented, giving rise to specialists for each layer. This led to confusion where UI, the most visible layer, was mistaken for the entirety of design.

The tension between UX and UI has been a topic of debate for years, causing frustration among designers. The strategic centrality of design has eroded, with UX often becoming a byproduct of business goals rather than a driving force. This fundamental shift has transferred control of design from the hands of designers to algorithms, automated tools, and business decision-makers.

The Downside of Specialization

Specialization has created knowledge silos, where UX designers focus solely on translating information into maps and wireframes, leaving UI specialists to "dress" them. This division has led to a more visually-driven approach rather than one focused on meaning.

In an agile context, UX has often been limited to specific journeys rather than the overall experience. Good design requires collaboration among all elements, from information architecture to visual interaction, grounded in research and experimentation. Empathy remains key to understanding user needs.

Reversing the Paradigm

Interestingly, machines now seem capable of understanding our language, completely reversing the paradigm. In the future, we will need to understand the value of spending time reimagining such metaphors and whether they will remain relevant. If the machine can understand us, should we still focus on making it comprehensible to us?

In this regard, it is worth reading Maximillian Piras's article on the evolution from command line to graphical user interface (GUI) and then to conversational interfaces, and what the next layer of abstraction means for designers, along with Jacob Nielsen's reasoning on how autonomous agents will shape user experience.

It is noteworthy that all AI models currently available to the public still rely on good old metaphors to solve a complexity problem and make themselves intuitive to use. The chat interface is, in fact, a technological regression. In the 90s, we had CLIs (Command Line Interfaces)—just text, then we had GUIs (Icons/Windows). Now, with AI, we have returned to text (LUI), but we have dressed it in the metaphor of conversation to make it understandable.

What Can We Do?

In the meantime, two interesting developments are emerging:

• Designers can truly focus on the systemic and service level, rather than getting bogged down in details or partial journeys. We can concentrate on complex journeys involving multiple systems and touchpoints, essentially focusing on what was called Service Design. People use products for the outcome, not for the product itself, and the unique product is generally just part of the service.

• There are many products where it is clear that those who built them believed that a clean user interface and decent usability were enough to ensure a great experience. The problem is that one often ends up with very clean and polished interfaces over extremely complicated journeys.

A Concrete Example

A real example is the incredible and commendable effort made in Italy to digitize services for residents. In recent years, they have invested heavily in digital transformation, and now you can do almost everything online. They have also created an agency to establish a solid design system to improve accessibility, usability, and, in general, simplify interaction with public administration.

(Ironically, accessibility is presented as a key goal, yet all guides are in downloadable PDFs, which are notoriously worse than HTML versions in terms of accessibility.)

Recently, I had to deal with some of these services, and here’s an example: you can access your personal data using your Digital Identity Document (Carta d’Identita Elettronica — CIE), however, to do so, you need to obtain a PIN and a PUK number, half of which arrives when you try to access it for the first time, while the other half was on the paper applications you submitted when requesting the document.

Do I still have that number? Was I aware of the importance of such a number when I applied for the card? And what about elderly people, who are generally the ones who need access to certain services the most? (for example, pensions, health records, etc.).

Speed as a Key Factor

It is undeniable that the interface will remain essential, and we must ensure that each part is learnable, understandable, usable, satisfying, etc. Here, we have another opportunity: we can create prototypes that we can test in a few hours rather than several days, thus accelerating the process and allowing us to gather even more data to design better products, rather than spending days on Figma simulating all scenarios of a single journey based solely on assumptions.

We can obtain data more quickly, iterate faster, test across multiple connected journeys, and collaborate with engineers to build prototypes that are almost production-ready. The key is not just the acceleration of the process, but the amount of data we can gather and use, thereby reducing the feedback cycle. I believe that user recruitment, historically a pain point in most cases, will become even more critical.