Designing in the Age of AI Agents
What Changes, What Doesn’t
❝We become what we behold. We shape our tools, and thereafter our tools shape us.❞
— Marshall McLuhan1
Nearly every project I touch these days involves AI. We’re not just using AI to speed up workflows, sense-check ideas, or summarise notes — we’re designing and building AI-powered features right into the products themselves.
It’s even made its way into hiring. In recent design interviews, the topic almost always comes up: “How are you using AI?” “Is it replacing parts of your process?” “Should I be worried?” (That last one is usually unspoken but easy to spot in the eyes.)
It’s fair. The landscape is shifting, fast. AI is not “on the horizon” anymore — it’s sitting in the meeting, already commenting on the doc. So let’s pause, look around, and ask the real questions: What’s actually changing? What’s still essential? And what kind of designers will thrive in this weird, wonderful new world?
🔧⚡⏳
What’s Changing — Tools, Tempo, and Expectations
Designers have always adapted to new tools. But this time, the tools talk back. They generate screens, write copy, analyse feedback, and even suggest improvements — sometimes without waiting to be asked. Handy, right? Slightly unnerving? Also yes.
A few things that are clearly in flux:
Tools are becoming teammates (sort of). We’re no longer just using AI — we’re prompting it, editing its ideas, arguing with its tone choices, and occasionally wondering if it’s gaslighting us. (It’s not. Probably.)
Work has gone turbo. One prompt now does what used to take hours. That’s amazing until you realise your boss now expects three versions by lunch, plus one “playful” variant just for fun.
Quantity isn’t the problem anymore. When you can generate 50 layouts in 5 seconds, the real challenge becomes figuring out which one doesn’t look like it was made by a caffeinated squirrel.
In short, we’ve sped things up — but we haven’t necessarily made them better. That’s still our job.
🏛️✨🔑
What Doesn’t Change — and Actually Matters More Now
For all the disruption, some things haven’t changed at all. In fact, the most important parts of design have only become more critical.
Let’s not forget:
Good designers frame the right problem. AI is excellent at spitting out solutions. But unless someone’s asking the right questions, we’re just getting faster at going in the wrong direction.
Taste still matters. Aesthetics, judgment, cultural nuance — that comes from years of experience, not machine learning. AI can remix what already exists, but it can’t invent new meaning. Yet.
Design is still human. It’s negotiation. It’s facilitation. It’s managing conflicting needs, blurry goals, and last-minute changes from Steve in Legal. No amount of automation will handle that gracefully.
So no, the fundamentals aren’t going anywhere. If anything, they’re the anchor we need in a sea of AI-generated enthusiasm and questionable typography.
🧠🛠️🚀
The Emerging Design Skillset
Designers don’t need to turn into prompt engineers or machine learning experts. But we do need to get smart about how AI works — and how it doesn’t.
Here’s what’s rising in value:
AI fluency. You don’t need to write Python, but you should know what a language model is before you ask it to rewrite your onboarding flow.
Data awareness. If you’re designing with or for AI, you’re dealing with training sets, probabilistic outputs, and feedback loops. A bit of statistical common sense goes a long way.
Designing for unpredictability. AI-powered systems don’t always give the same output twice. That means we’re not just designing UI — we’re designing behavioural guardrails.
Ethical literacy. AI can do a lot. That doesn’t mean it should. Designers need to be the ones in the room asking uncomfortable questions — preferably before the product goes live and ends up on the front page of Reddit.
Curation and sense-making. When AI can generate endless options, the value isn’t in creating more. It’s in knowing what not to ship.
These skills are less about complex tools and more about adaptive thinking. The kind of thinking that keeps you relevant — and useful — no matter how many bots show up to the brainstorm.
Design has always been about change
⚖️🤖❗
Design Ethics in the Age of “It Was the Algorithm”
Let’s talk ethics — not the “don’t use Comic Sans” kind, but the big stuff.
AI systems are making real decisions: what we see, what gets recommended, what’s hidden, and what’s nudged. Designers are now part of those decisions, whether they like it or not.
So, a few things we can’t ignore:
Bias doesn’t vanish — it scales. If your data is skewed, your AI product will be too. Fast. That’s not a bug. That’s just math with consequences.
Opacity is a design problem. If a system makes a choice and users don’t understand why, that’s a UX issue. You don’t get to say, “Oh, the model did it.”
Neutrality is a myth. Every automated decision has a value baked into it. Pretending it’s neutral means someone else made that call without you noticing.
Dignity is the new north star. Whether we’re designing AI chatbots, recommendation engines, or invisible nudges — we’re influencing real human behaviour. Let’s try not to be creepy.
Ethical design isn’t a footnote anymore. It’s a core competency.
🎨🧩📈
Redefining the Role: From Pixel Pushers to Strategic Shapers
The more AI handles the executional side of design, the more we’re left with the hard, strategic stuff. You know — the messy, political, business-critical work that tools can’t automate (yet).
Designers are becoming:
So, yes, the job is changing. But it’s not shrinking — it’s expanding. You’re not being replaced by a robot. You’re being invited to design the future with one.
😏🎤💡
Final Thoughts
Here’s the thing. You’re not obsolete. Not even close. Unless your entire design practice is pushing pixels in alphabetical order — in which case, maybe do a course.
The rest of us? We’ve got this. Design has always been about change. We survived the move from print to digital. We survived the flat design wars. We even survived that phase where everyone put gradients and blobs on everything.
So let’s treat AI the same way we treat any trend or tool: with curiosity, caution, and just enough skepticism to keep things interesting. The goal isn’t to beat the machines. It’s to design better alongside them — and sometimes in spite of them.
And if one day the AI overlords really do take over, I say let them handle stakeholder feedback. We’ll see how long they last.
NK
☉
Marshall McLuhan was a Canadian philosopher, media theorist, and professor, best known for his pioneering work on how media and technology shape human perception and society. Born in 1911, he became internationally famous in the 1960s for coining phrases like “the medium is the message” and “the global village.” McLuhan argued that the form of a medium —not just its content— has a profound impact on how we understand and interact with the world. His insights were strikingly ahead of their time, predicting the rise of the internet and the cultural shifts that would come with it. Today, he’s regarded as one of the key thinkers who laid the intellectual foundation for media studies, cultural theory, and even aspects of digital design. His work reminds us that technology doesn’t just serve human needs — it transforms them.
Designers as systems thinkers: Designers today can no longer focus only on isolated screens or individual interactions. AI-powered products often operate within complex ecosystems—connecting data, users, algorithms, business goals, and technical infrastructure. Being a systems thinker means understanding these interconnected parts and how changes in one area ripple across others. Designers must map out these relationships, anticipate unintended consequences, and design experiences that work smoothly across channels and contexts. This mindset helps prevent siloed thinking and fosters holistic solutions that stand the test of time and scale.
Designers as problem framers: Before AI can generate a thousand ideas or automate tasks, someone needs to define the right problem to solve. Designers are increasingly called on to step back and ask: What is the real challenge here? What assumptions are we making? How do users genuinely experience this issue? Problem framing is about digging deeper than symptoms to uncover root causes, often navigating ambiguous or shifting goals. This skill is crucial in an AI context, where generating endless solutions is easy—but choosing meaningful ones requires clear, strategic thinking.
Designers as value translators: In organisations adopting AI, designers frequently act as bridges between diverse teams: engineers, product managers, marketers, legal, and executives. Each group has its own language, priorities, and metrics. Designers translate user insights and design principles into terms that resonate across departments, helping everyone understand how design decisions impact business value, user satisfaction, and ethical considerations. This role elevates design from a purely creative function to a strategic partner in shaping product direction and success.
Designers as ethical challengers: With AI’s power come significant ethical dilemmas—bias, privacy, transparency, and unintended harm, to name a few. Designers must be the voices raising these issues early and often, challenging assumptions that “the algorithm knows best” or “it’s just data.” This role involves advocating for responsible design practices, questioning how decisions affect diverse user groups, and pushing for transparency and fairness. Ethical challengers help organizations avoid pitfalls and build trust with users, ensuring technology serves people—not the other way around.
Designers as cross-functional glue: Design in AI-driven projects requires seamless collaboration across multiple disciplines—data science, engineering, user research, business strategy, and more. Designers who can weave these threads together become the glue that holds teams aligned and moving forward. They facilitate communication, mediate trade-offs, and keep user needs front and centre amid technical complexity and shifting priorities. This role demands empathy, diplomacy, and a broad understanding of how different functions contribute to the product’s success.