In 2025, Jenny Wen, design lead for Claude at Anthropic, declared that the design process is dead. The NNGroup responded that it isn’t dead, just compressed.

To die it would’ve had to be alive. In most places where design is practiced around the world, a formal process never existed. The real ratio of designers to developers is 1:100 in most cases. An established design process has always been a midsummer night’s dream, not a reality.

So which process died exactly?

The design process is dead!

After the repercussion of the Anthropic design lead’s words, I went looking to better understand the argument. On Jenny’s site, I found the post that started the debate, a reflection on the Double Diamond. Between 2014 and 2020, in Silicon Valley, design became about The Process. Portfolios stopped showing the work and started showing artefacts like journey maps, personas, flows, user stories. Proof that the process had happened. Along the way, designers stopped caring about the thing itself, about what people saw and felt when using what was made. They ended up serving the process instead of serving the work.

Up to this point, it’s a valid observation. Sometimes we get used to a certain interpretation because it’s what the market expects. But I always understood it as interview training, not as exactly what managers would expect from me on the job. 2020 was when I formally entered the design path, and I remember learning the Double Diamond as one of the design processes, not the only one. When building the portfolio, what I was taught about process wasn’t so much whether I followed it to the letter, but whether I understood why I was doing what I did and documented design decisions. That documentation was more about knowing how to explain to others (because communication is a key skill for designers) than about following a specific methodological model in an orthodox way.

The problem is the ornamental dive that comes next. She starts from two reasonable observations, that portfolios became too orthodox and that the design process is not linear, and somehow arrives at the conclusion that “the design process” (as if there were only one) is dead and we should just follow intuition. But how do you build intuition without experience? That’s the NNGroup’s point (who are also Californian). Without process, what’s left? In Jenny’s discourse, what’s left is chasing after engineers to “learn something new”, as if that were very different from what’s already done. Selling this as innovation and visionary thinking is reinforcing the status quo that has progressively devalued the design profession.

Long live the design process!

“NN” are the initials of the founders, Jakob Nielsen and Don Norman, who popularised ergonomics and work psychology techniques applied to computers. Don Norman was VP at Apple, called his role “User Experience Architect”, and the term took off. Jakob Nielsen created the 10 usability heuristics and is still a relevant voice on the subject. The design process they advocate is User-Centric Design.

Their response to Jenny is conservative (after all, they sell UCD courses, they have to protect the business), but it has a valid point. Experience and exposure to design work are in fact necessary to develop intuition. Experienced designers don’t “skip process” — they have an internalised repertoire that allows them to make decisions more quickly. The intuition Jenny Wen alludes to is the result of years of practice, not a substitute for it.

However, the NNGroup’s presupposition is fundamentally the same as the Anthropic design lead’s — that there is a single, ideal approach to design and that this approach works for everything.

Which process of which design

The debate over the death of the design process reflects the specific reality of the US digital product industry. It’s a valid discussion within that context, but one that, given the visibility of those leading it, ends up being received as if it concerned everyone. Design exists and has always existed beyond that frame.

For Mauricio Mejía, professor of design at Arizona State University, there is no single design method or prescriptive design, as suggested by models like Design Thinking and the Double Diamond. In fact, these moulds distort what designers actually do, which is highly adaptive. In practice, design processes are messy but rigorous, reordered with every project. There is no single manual, there is a playground. In that case, the Anthropic vs NNGroup debate loses its point because there is no single process to die or be resurrected.

Karri Saarinen, CEO of Linear, draws on designer and architect Christopher Alexander to define design as the search for a fit between form and context. Alexander, whose work influenced both architecture and software design, understands context as the system of forces that constitutes the problem. When those forces remain unresolved, design has failed.

“The assumption behind them is clear: that design is the act of producing. That is the misunderstanding. The hard part of design is rarely generating the form. It is understanding the problem well enough to know what and how something should exist at all.” — Karri Saarinen, Output isn’t design

For Karri, design is the act of understanding, not of producing. Just as writing organises thought, working visually organises the comprehension of the problem. Moving things, testing relationships and refining structure is part of how clarity emerges. Delegating this to AI generates output, not understanding. AI can help prototype, explore, investigate. But design demands judgement, conversation, tension and time. The risk is confusing generated form with solved problem.

This movement is the design process, which always takes the shape of the context. The discussion that dominates the internet today, however, carries the opposite premise.

Universal design and Californian ideology

The premise that there is a right and universal way to do design, based on a sequential methodological presupposition, almost like an assembly line, was built by the Ulm School and adopted by Silicon Valley in a practically direct manner.

In California, the method gained speed (delivery cycle deadlines), market validation (if it scales, the design was right) and depoliticisation (the method is neutral, the designer is an operator, not someone who makes political choices). Barbrook and Cameron called this the Californian ideology, where technology liberates, the market resolves, and politics is friction. The Double Diamond is a child of this marriage, a method that presents itself as universal within an ideology that doesn’t recognise itself as ideology.

But this context represents less than 5% of the places where design is practiced in the world. In Brazil, Latin America, Africa, Southeast Asia, in agencies, consultancies, NGOs, government, there was never a Double Diamond. There were people doing design as best they could, with whatever resources they had, often with good results, often without. Even ESDI, the first higher education institution for design in Brazil, founded in Rio de Janeiro in 1963 directly inspired by Ulm, did not reproduce the model as a rulebook. The Double Diamond, to die, would’ve had to be alive.

That’s why the statement from Anthropic and the response from the NNGroup are unsurprising. Both are rooted in the productivist logic of design typical of Silicon Valley. That is, production at scale takes precedence over understanding the problem. The stance of designer as operator, not as critical thinker, is evident in the Anthropic design lead’s reaction to criticism. Instead of reflecting on what she said, she stated that “the market isn’t ready” for what Anthropic proposes. Put that way, it sounds like the proposal is visionary design. The numbers, however, suggest a less encouraging picture.

The impact of AI on the design job market

From 2022 onwards, as AI tools became widespread, the design profession lost value and jobs. Jeff Sauro, author of Quantifying the User Experience and founder of MeasuringU, coordinates the biennial UXPA survey with 444 respondents across 37 countries. In 2024, the survey showed that UX staff losses returned to levels not seen since the 2009 financial crisis, and for the first time in the entire historical series, hiring and losses were tied at 35% to 35%. The field stopped growing. Research was the hardest hit area, with nearly a quarter of companies eliminating dedicated researchers. In 2025, a partial recovery of positions began, but one that favours senior professionals and excludes juniors.

The industry spent a decade improving the ratio (IBM from 1:72 to 1:8, Uber grew its design team 70x) and is now reversing that trend with layoffs concentrated in design and research, while hiring AI engineers. More broadly, layoffs.fyi recorded cumulative layoffs of over 800,000 positions across engineering, design, management, product and other functions. The argument behind the second wave of layoffs, which began in 2024, is “AI restructuring”. In practice, companies are reorganising teams to be lean and work with AI tools, including design. This is a structural change with no end date in sight.

Conclusion

The design process didn’t die because it never existed in the singular. What exists are processes, plural, adapted to specific contexts, resources and problems. Declaring the death of something most people never practiced, at a moment when the profession faces the biggest crisis of its recent history, isn’t intellectual provocation. It’s a failure to read context, precisely what designers should know how to do.

The debate between Anthropic and the NNGroup is legitimate within the frame of those who live it. The problem is when that frame presents itself as universal, and when the ideology behind it doesn’t recognise itself as such. There’s more to say about that. It’s for a next text.

References