The AI design partner I really want
Moving AI beyond assistant or accelerator towards a real "partner"
I've been deep diving into v0, Lovable, Replit, and Cursor for the past few weeks— we're talking like 4+ hours a day of pure AI-assisted design experimentation using some old dusty mocks/sketches that never shipped. 😅 It's been equal parts exhilarating & exhausting, like holding a rapid-fire convo with someone who speaks nicely capable design, but randomly hallucinates entire user interactions from nowhere. 🙄 Sigh…
But beyond constant tweaking/refining, here's what I’m wondering:
If GenAI is going to be my design partner — not merely a fancy quick prototype generator — what would that actually be like? 🤔
Not some hypey marketing pitch version, but a bonafide, back-and-forth, challenging partnership that pushes me to be better as a designer at my craft. It’s almost cliché to say most GenAI tools feel like eager interns. 🙄 They'll generate whatever you ask for, nicer and faster than you imagined, but they won't tell you where your concept is critically flawed. They won't ask the uncomfortable questions that matter, in effect telling “truth to power” — and yes, you hold the power, not the AI! ⚡️
I want an AI partner that functions more like the best design critique sessions I've ever had — the ones that left me kinda uncomfortable (and yes, maybe testy) but highly informed, with my work nicely improved. Something that challenges me in a Socratic way: "Why are you solving this particular problem? What happens if your assumption about user behavior is wrong? How does this design decision serve the actual business need versus just looking slick or cool in a demo?"
When I'm deep in a design flow, especially with some GenAI tool amplifying my output, I’ve noticed it’s easy to become seduced by what's possible rather than what's necessary. That “shiny object syndrome” is real, and AI can make everything shinier! 😅 Now, I need a partner that pulls me back to those fundamentals:
Are we still solving the user's primary problem/pain point, or have we drifted into feature creep?
How does this iteration actually improve on the last one, beyond surface aspects — information hierarchy, interaction flow, usability & accessibility, etc.
What cornerstone research insights are we supporting or ignoring with these choices?
The GenAI tools today speedily excel at tactical execution, but struggle with strategic thinking, I’ve found. 🤨 (BTW, if you know of better tools or prompts to enable strategic thinking — please do share!) They can generate nicely stylized components but can't tell you why one approach serves long-term business value better than another — that might involve a million tokens’ worth of context perhaps! 😅
I want a partner that helps me connect design decisions to impactful, prioritized strategic outcomes… and something that can surface technical constraints before they become expensive problems, or explain how a design choice aligns with three-year marketing positioning plan, but might limit future pivots.
At least raise those considerations proactively so I can better prepare and shape a cohesive, anticipatory design direction or roadmap that satisfies stakeholder demands…and anxieties!
Here's what I keep coming back to: in my experimentation with GenAI design tools, the tools I've been using are amazing at generating & iterating, but they're quite passive when it comes to critical thinking. They'll execute my vision with much fine-tuning needed, but not interrogate whether it's the right vision to begin with.
Maybe the AI partner I want isn't one that has all the answers, but one that asks better questions than I'm asking myself — to stimulate my design intuitions. 🤓
So what would push you to grow as a designer? What kind of AI partner would make you uncomfortable enough to level up your thinking, not just your output? ✨