The oldest designed object we know of is a hand axe. Acheulean, roughly 1.7 million years old, shaped from stone by someone who looked at a rock and saw a tool inside it. There was no aesthetic intent. There was no user research. There was a problem — cutting, scraping, surviving — and someone figured out that if you hit a stone at the right angle enough times, you could make it do what you needed it to do.
That's design at its most elemental: shaping material to serve a purpose. And for most of human history, that was enough. Function was the whole game. You made things that worked, and whether they were beautiful was irrelevant because beauty wasn't the point. Staying alive was the point.
I think about that hand axe a lot, because the entire history of design is really just the story of how we got from "does it work" to "does it work, and does it feel right, and does it serve the person using it, and does it make the business money." That progression took 1.7 million years, and the last and most consequential leap — the one that defines the profession I've spent my career in — happened in roughly the last twenty.
The Industrial Revolution gave us mass production, and mass production changed the question. When you're making one chair for one person, functionality is all that matters. When you're making ten thousand chairs for a market, you start thinking about appeal, about differentiation, about what makes someone choose your chair over the one next to it. Aesthetics were still secondary to practicality and cost during this period, but the seed of something important was there: design was beginning to serve commerce, not just survival. And once design served commerce, design had to justify itself in commercial terms. That relationship has never gone away. Every design job that has ever existed was funded because someone believed it would contribute to a business outcome. The form of that belief has changed over two centuries. The underlying economics haven't.
Art Nouveau, in the late nineteenth and early twentieth century, pushed that further. Organic lines, nature-inspired forms, the idea that everyday objects — furniture, architecture, household items — deserved to be beautiful. This was the first time design treated beauty and functionality as things that could coexist rather than compete. It mattered not because of any single object it produced, but because it established a principle that would shape everything after it: the experience of using something is part of what makes it good.
The Bauhaus school formalized that principle. Founded in Germany in 1919, Bauhaus argued for the unity of art, craft, and technology, that good design should be functional and visually honest, and that it should be accessible to everyone, not just the wealthy. "Less is more" became more than an aesthetic preference; it was a philosophy about how objects should relate to the people who use them. The Bauhaus influence ran through furniture, housing, typography, and graphic design, and it laid the groundwork for everything the modernists would build on.
Then came Dieter Rams, and the line from Bauhaus to the present day became visible. Rams spent decades at Braun producing consumer electronics that embodied ten principles he articulated with a clarity that still holds up: good design is innovative, useful, aesthetic, understandable, unobtrusive, honest, long-lasting, thorough, environmentally friendly, and involves as little design as possible. That last one is the one I keep coming back to. As little design as possible. Not as little thought, not as little care — as little unnecessary intervention between the person and the thing they're trying to do.
Steve Jobs saw Rams' work and recognized something in it. Put a Braun T3 pocket radio from 1958 next to a first-generation iPod from 2001 and the lineage is unmistakable: the same simplicity, the same quiet confidence that the best design disappears into the experience of using it. Jobs brought that philosophy to software, and in doing so he helped create the conditions for an entire profession to emerge.
The rise of digital product design happened fast, historically speaking. The internet created the need, and two structural shifts turned that need into an industry. The first was the SaaS model — subscription-based software that lived or died by whether people kept using it. When your revenue depends on retention, the quality of the experience stops being a nice-to-have and becomes the product itself. Suddenly, design had a direct line to the business model: better experience, lower churn, more revenue. Companies didn't hire designers out of principle. They hired designers because the numbers made the case. The second structural shift was agile development — rapid iteration, continuous deployment, feedback loops measured in weeks instead of years. Agile needed designers who could keep pace, who could think in systems rather than static deliverables, who could iterate alongside engineers rather than handing off polished specs and hoping for the best.
For a while, this worked. Design had a seat at the table — or at least a seat adjacent to it. Companies hired UX designers, then interaction designers, then product designers, and the titles kept shifting because the discipline itself was still figuring out what it was. But the reason those seats existed was always the same: someone with budget authority believed that investing in design would produce a return. The moment that belief wavered, or the moment someone else could approximate that return faster, the seats started disappearing.
Apple's influence on this period is hard to overstate. When Jony Ive took over Apple's Human Interface team in 2012 and launched iOS 7 the following year, killing skeuomorphism and replacing it with flat, minimal, depth-focused design, the entire industry followed. Apple was the standard by which every digital experience got measured, and what Apple did, everyone else scrambled to match. Google responded with Material Design in 2014, bringing its own metaphor of digital surfaces that rise to meet your touch, and for the first time major platforms were competing on design philosophy itself, not just features.
The neuomorphic reaction came next, soft shadows and subtle depth cues trying to split the difference between flat and skeuomorphic. Instagram went from leather-stitched borders in 2010 to clean minimalism by 2024. The visual language of digital products evolved in public, in real time, and designers were at the center of it.
And then, in a move that should probably serve as a cautionary tale, Apple circled back. In 2025, they introduced Liquid Glass across iOS, macOS, and every other platform they control — a translucent, refractive design language that makes every element look like it's rendered on a wet sheet of glass. The reaction was immediate and almost unanimously negative. Users complained about readability. Accessibility advocates flagged contrast issues. Developers pointed out the performance overhead. And designers, the community that had spent a decade following Apple's lead, looked at it and largely agreed: this was design for design's sake. Visual novelty disconnected from any user need, any behavioral insight, any business rationale. It was craft without purpose, and the backlash was a reminder that craft without purpose is exactly how a profession loses credibility with the people who fund it.
I bring up Liquid Glass because it's a perfect example of the trap this book is about. Apple has the resources, the brand equity, and the installed base to survive a polarizing design decision. Most design teams don't. When the profession's most visible company ships something that prioritizes aesthetic ambition over usability and business outcomes, it reinforces every skeptic's belief that designers care more about how things look than whether they work. And in a market where design headcount is already under pressure, that's a reputation the profession can't afford.
Then, in late 2022, generative AI arrived in a form that regular people could actually use, and the timeline accelerated in a way nobody in the profession was prepared for.
Within eighteen months, AI could generate interfaces from natural language prompts, synthesize user research, produce functional prototypes, write production code, and create visual assets that were, if not great, good enough to ship. The production layer of design work — the layer that had justified most design hiring for the previous decade — went from taking days to taking minutes. Companies didn't stop needing design. But they started questioning how many designers they needed, and that question landed in a market that was already tightening. Design hiring flatlined. Teams got consolidated. The "seat at the table" started feeling less like a permanent arrangement and more like a seat that could be optimized away.
The cultural reaction has been interesting to watch. There's a visible counter-movement happening right now that rhymes with something we've seen before. Dumbphones are selling in numbers that would have seemed absurd five years ago. Light Phone, the credit-card-sized device that does calls and texts and basically nothing else, can't keep up with demand. Vinyl record sales have outpaced CDs for three years running. Film photography is back, not as nostalgia but as a deliberate choice by people who want friction in their creative process. Journaling apps are losing users to paper notebooks. There's a growing market for tools that do less, slower, on purpose.
It looks a lot like Art Nouveau. Not the aesthetic — the impulse. Art Nouveau was a reaction to industrialization, a demand that mass-produced objects still carry beauty and human intention. What we're seeing now is a reaction to optimization, a demand that the things we use still feel like they were made by and for actual people. The language is different but the underlying need is the same: when the dominant system prioritizes speed and scale, a meaningful number of people will pay a premium for something that prioritizes presence and craft.
But here's where the parallel breaks down, and this is important. Art Nouveau was reacting to factories. Factories didn't go away. The artists who rejected industrialization didn't reverse it; they carved out a space alongside it, and eventually their ideas about beauty and human-centered intention got absorbed back into industrial practice. The Bauhaus wouldn't exist without Art Nouveau, but Bauhaus worked with industrial production rather than against it. The same thing is going to happen with AI. The dumbphone movement won't reverse the fact that three billion people interact with AI-assisted interfaces every day. The analog backlash won't undo the reality that generative tools have permanently changed the economics of production. AI is here to stay, and the question for designers isn't whether to engage with it. The question is whether we're going to let other people define what AI-driven experiences look like, or whether we're going to do what designers have done at every previous inflection point in this history: step into the new medium and shape it.
This is a once-in-a-generation opportunity, and I don't use that phrase lightly. The last time the design profession had a chance to define an entirely new category of experience was the early smartphone era, roughly 2007 to 2012, when the fundamental patterns of mobile interaction were still being invented. We're in a similar window right now, and the shift is arguably more fundamental than the one mobile introduced.
Here's why. Every piece of software you've ever used, up until very recently, was deterministic. You pressed a button, you got a result. The same button, the same result, every time. The interface was fixed. The layout was fixed. The content was authored by a human and stored in a database and served back the same way to every user. Designers controlled every pixel because every pixel was predetermined. The entire discipline of product design was built on this assumption: that you are designing a specific, knowable thing that will behave the same way for everyone who uses it.
Generative UI breaks that assumption. Software is becoming non-deterministic. Interfaces that assemble themselves in response to context, content that's generated rather than retrieved, experiences that are different every time someone opens the app because the system is constructing them on the fly based on who the user is, what they're trying to do, and what the model thinks will help. This isn't speculative. Vercel is shipping frameworks for it. Every major product org is prototyping it. Within a few years, the majority of digital interfaces people interact with will have some generative component, and many of them will be substantially or entirely generated.
To make this concrete: imagine a grocery app. Same user, same home screen, three different layouts depending on whether the system infers you're restocking staples, planning meals for the week, or shopping on a budget. The produce page reshapes itself when your browsing history suggests you're meal-planning — recipe modules surface, quantities adjust, complementary ingredients appear alongside the items you're already looking at. Your cart doesn't just hold what you've added; it notices you're two ingredients short of a recipe you've been building toward and offers to finish it. A substitution ribbon appears at checkout, not because the system is guessing, but because your past behavior shows price sensitivity in this category and the confidence score is high enough to intervene. And if you dismiss the suggestion, that dismissal becomes a signal that recalibrates the next interaction.
None of this is templated. The modules, the copy, the calls to action — they're assembled at runtime from a governed component library, shaped by an intent classifier that's reading behavioral signals and a confidence scorer that determines how aggressively the interface should reshape itself. The design system isn't a style guide anymore; it's a governance layer, the set of rules that ensures a generative engine produces experiences that are on-brand, accessible, and coherent even when no designer has seen the specific screen a user is looking at. Research out of Google and Microsoft suggests that for repeated, transactional tasks like this, adaptive interfaces outperform conversational ones. Showing is faster than asking. The best AI-driven experiences won't make you talk to a chatbot to get what you need. They'll rearrange the world you're already looking at.
This changes what it means to be a designer in ways the profession hasn't fully absorbed yet. You're no longer designing screens. You're designing the rules and constraints and values that a system uses to generate screens. You're not deciding what a user sees; you're deciding what the system is allowed to show, and under what conditions, and within what guardrails. The craft shifts from composing specific layouts to defining the possibility space of acceptable outcomes. If that sounds familiar, it should — it's the same concept from the global and local maximum framework, applied to the interface itself.
The conventions for this kind of design barely exist yet. What does trust look like when the system is generating rather than retrieving? How do you maintain brand coherence when no two users see the same interface? How do you calibrate intervention — when should the system reshape aggressively, and when should it hold back? How do you test an experience that's different every time? How do you measure whether a generated experience is good when "good" depends on context the designer never saw? These are design problems, and they're wide open. The designers who step into them with business fluency, technical curiosity, and a willingness to work in territory that doesn't have established patterns yet are going to define the next era of this profession. The ones who sit it out, waiting for the hype to pass or the deterministic world to return, are going to find that the world moved on without them.
But here's the part of the history that matters most for this book, and it's the part that rarely gets told as clearly as it should.
Through all of this evolution — Industrial Revolution to Bauhaus to Rams to Jobs to SaaS to generative AI — the discipline kept expanding what it meant by "design" without ever fully resolving an internal tension that was there from the beginning.
User experience design, as it came to be practiced, focused on the user. That was the whole point — understand people's needs, preferences, and behaviors, then shape the product around them. UX sat firmly inside the user needs circle, and it did that work well. Research, usability testing, information architecture, interaction patterns — these were real skills that produced real improvements in how people experienced digital products.
But UX design, by definition, was solving for one side of the equation. The business had needs too — revenue, growth, market position, operational efficiency — and for a long time those needs lived in a separate circle. Product managers handled the business side. Designers handled the user side. And the overlap between them was where the interesting problems lived, but the profession hadn't quite built the language or the frameworks to claim that overlap as its own territory.
Here's what I've learned in twenty-two years: businesses don't hire designers to make things beautiful or even to make things usable. They hire designers because they believe design will drive outcomes they care about. Sometimes that belief is explicit and well-measured. Sometimes it's vague and aspirational. But it's always there, and the designers who understand what the business is actually trying to achieve are the ones who survive budget cuts, reorgs, and shifting priorities. The ones who don't understand it are the ones who wake up wondering why their team got halved despite doing great work.
Product design, as a term and a practice, emerged to name this reality. Product design considers the business context and designs for behaviors that drive both user and business value. It's the difference between "make this usable" and "make this usable in a way that also moves the metrics the business cares about." Product design is moving a button and knowing the potential business impact of that change. It's understanding that revenue is a lagging indicator, that the leading indicators are behavioral — how many items a user adds to a cart, how often they return, what they do in the first five minutes — and that design has more direct influence over those leading indicators than almost any other function in the organization.
That's the business case for design. Not "we make things pretty" or even "we advocate for the user." The business case is: we understand how user behavior connects to the numbers on the P&L, and we know how to move those behaviors through research, experimentation, and intentional design decisions. When designers can make that case clearly, they don't just get hired. They get resourced, they get promoted, and they get listened to when the roadmap is being built.
This distinction between UXD and PD isn't just semantic. It's the fault line that explains a lot of what's happening in the profession right now. The designers who built their identity around user advocacy, which is noble and important work, are the ones most vulnerable to the shifts this book describes. User advocacy didn't stop mattering. But it turned out that the organizations making decisions about headcount and investment were measuring something different, and the designers who could tie their work to those measurements had more leverage than the ones who couldn't.
I give a version of this history in a presentation I've delivered to teams and at conferences, and the part that always generates the most conversation is the Venn diagram. Two circles: business needs and user needs. UX design lives in the user needs circle. Product design lives in the overlap. The visual is simple, almost too simple, but it names something that a lot of designers feel without having the framework to articulate it.
When I started in this field twenty-two years ago, the profession was still mostly operating in the user needs circle and fighting to be heard by the business needs circle. The "seat at the table" era was really about getting close enough to the overlap to influence it. And we got there — slowly, unevenly, but we got there.
What nobody planned for was that the overlap would keep expanding while the exclusively-user-focused territory would shrink. As product management matured, as engineering velocity increased, as AI made production faster, the organizations that once needed a dedicated function to represent user needs found other ways to approximate that representation — rougher, less nuanced, but fast enough to ship. The exclusive territory of UX, the part of the user needs circle that doesn't touch business needs, became harder to fund, harder to staff, and harder to justify as a standalone investment. And when a function can't justify its investment, it gets cut. That's not cynicism. That's how organizations work.
Product design, on the other hand, the work that lives in the overlap and speaks both languages, didn't shrink. The designers who could walk into a quarterly business review and explain how a redesigned onboarding flow increased activation by 15%, and what that meant for annual recurring revenue, those designers kept their seats. They kept their headcount. They got asked to do more. The gap between those designers and the ones still framing their value as "I advocate for the user" widened every year, and then AI accelerated it.
That's the ground this book stands on. The history of design is a story of expanding scope: from pure function to aesthetics, from aesthetics to usability, from usability to experience, from experience to business impact. Each expansion added a layer without fully replacing the one before it. And at every transition, the designers who understood only the previous layer found themselves struggling to articulate their value in the new context.
We're in another one of those transitions now. The difference is that this one is moving faster than any of the others, and the tool driving it doesn't just change what designers make. It changes how fast everyone else can make things without them. The business case for design hasn't disappeared. If anything, the potential impact of good design decisions has never been higher, because the systems we're designing now operate at scales and speeds that amplify every choice, good and bad, in ways that static products never did. But the bar for what counts as "good" has shifted from craft quality to business fluency, and the window for defining what AI-driven experiences look like is open right now, not forever.
This book is about how to read that window clearly, step through it intentionally, and do the kind of work that justifies not just your seat at the table but the table itself.
That's where Chapter 1 begins.