Navigating the AI x Design frontier
We're all figuring this out together—and that's exactly the point.
If you're reading this as a UX/product designer, chances are you've felt a knot in your stomach when yet another damn AI tool launches, promising to revolutionize design workflows! 🙄 Likely you've also wondered if your carefully honed skills are fast becoming obsolete, or if that junior designer enthused with AI will somehow RAG and MCP past your decades’ of hard-fought experience. 😬 Sigh…
The AI replacement anxiety is real.
It’s OK…you're not alone feeling this way, even across other disciplines/fields (Eng, Product, Marketing, Sales, so on). We’re all trying to make sense of AI in the face of what the hypers overly glamorize or the CEO-driven memos demand.
We're pioneers now. The AI tools emerging today are definitely rough around the edges, the best practices don't exist yet, and everyone—from design leaders at Fortune 500 companies to indie consultants — is making things up as they go along. Seriously.
Your willingness to experiment, learn, and adapt matters more than title or years of experience. Like I often say, “everything is a prototype”. Now more than ever.
🧭 Some principles to guide the way
As we navigate this “Wild West” frontier, three principles can serve as a compass:
Progress over perfect: The goal isn't to master GenAI design tools now. It's to make incremental growth in your learning journey. Crawl, walk, run. Every small interaction or trial effort teaches you something about how such tools can enhance your work — or not! You can still decide that. 🤓
Everything is a prototype: Treat every GenAI-assisted design session as an experiment. That persona generated by AI? Cool. It's just a starting point, not gospel. That layout variation? Fine, it's inspiration, not final output. This mindset removes the pressure to get it "right"to focus on learning at your pace…regardless of what those fear-inducing Tech CEO memos say! (they’re just publicly working out some insecurities, so ignore them 🙄)
We're in this together: There's really no prize for being the first to crack GenAI design workflows perfectly — partly because "perfect" doesn't exist in a landscape this dynamic. Literally every week something new (and kinda kludgy) appears! Instead, share your experiments and learn from others doing the same! 🙌🏽
BTW I highly recommend checking out Xinran Ma’s free Maven talk in which he goes through his GenAI tool workflows — with visual diagrams of when he uses which tool. Very approachable starter!
🎭 Four roles for AI to play in your workflow
As you experiment with GenAI tools in your design workflow, it might help to think of AI as playing four distinct roles, as a launch point:
Operational support: Handling admin tasks like scheduling, coordination, and documentation that eat up your creative time.
Supplemental “intelligence”: Helping you prepare better, catch blind spots, and identify risks you might miss on your own. More in the spirit of raising considerations to factor into your thinking; but you decide if they’re worth the weight or priority value. 😁 Your agency persists.
Creative “collaborator”: Serving as that supportive aide/staffer that generates options, offers alternative POV, and helps you iterate further, and quicker. Remember, this is mainly about inspiration & perspectives, not replacement or finality. And again, you’re still the creative lead and final decider (as I discuss further below)…
Synthesis engine: Processing large amounts of info while offering convenient summaries that inform your decision-making — while taking many grains of salt, of course! 🤨 Question it with healthy skepticism and be wary of the sources, biases, etc. Then compare it with your own reflective summary that you wrote separately. What’s the overlap or what’s missing?
🍥 The irreplaceable design mind
Let's take a moment to articulate what makes designers uniquely valuable, and why that value potentially increases in an AI-assisted world. Design scholar Richard Buchanan, former head of Carnegie Mellon's School of Design, identified four core abilities that define all designers in his landmark “Governing Ideas” document for CMU, back in 1993 — 32 years ago! 😳
These abilities remain not just relevant but essential amidst GenAI potentiality.
✨ Creating concepts: Designers must be inventive in conceiving the possibilities of a product.
AI can generate countless variations (which are largely based upon pattern recognition & probability models), but it takes human creativity to imagine what should exist in the first place. AI can provide some starter material to work with (to get off the “blank page” faster), but you're still the visionary deciding what problems are worth solving.
⚖️ Judging options: Designers must be able to judge which of their inventions are viable in the contingent circumstances of manufacture and human use.
AI can simulate scenarios, but it takes human wisdom — through “lived experience” — to understand what will actually work when it meets messy complicated reality. Your actual experience navigating technical limits, user psychology, and real-world situations remains irreplaceable.
✅ Deciding a direction: Designers must synthesize knowledge across many fields —psychology, engineering, business, culture—and draw reasonable, pragmatic conclusions, or make decisions about the plan of a product.
AI can surface information from these domains, but it takes human intelligence(s) to make meaningful connections and informed decisions about product direction & outcomes.
👉🏽 Choosing the solution: Designers must evaluate options & select the appropriate (“right”) solutions based on values, preferences, and goals.
AI can analyze data and suggest optimizations, but it takes human judgment to weigh competing values and make choices that reflect what matters most to users…and businesses.
These four abilities (creating, judging, deciding, choosing) embody the core of design thinking that no GenAI tool or LLM can replace fully, effectively now.
“The real subjects of the new intellectual free trade among the many cultures are our own thought processes, our processes of judging, deciding, choosing, and creating.” — Herb Simon, The Sciences of the ArtificialThey require complex contextual understanding, ethical reasoning, and the ability to balance competing human wants/needs — also known as politics. 🙃 Fun!
→ Moving forward with confidence
So yes, AI helps generate more options, synthesize datapoints, and move faster. But the true value of designing lies in the human judgment behind critical choices—the envisioning, the decision-making, the prioritization of values. That remains uniquely yours. The designers who succeed will most likely be those who help steward AI forward, integrating it thoughtfully into their practice while holding strong to their core design values. ⚡️
So start small. Pick one stage in your process, experiment with one AI tool, and capture your insights — preferably your own authentically raw thoughts, before polishing them with ChatGPT! 😇 Share what works or what doesn’t, because everyone — including the GenAI toolmakers — is navigating this in real-time.
As I’ve written before, we’re not passive passengers in this journey. We are the stewards. The future of design is being defined now, and we have the power to shape it. So step forward with confidence, and be that designer who leads a path forward. 🙏🏽
What AI experiments are you running in your design process? I'd love to hear about your wins, failures, and everything in between. Hit reply and let's learn from each other.