Using LLMs to generate, critique, and refine UX design directions, IA, flows, layouts, and wireframes

I've been using AI in my design process daily for a while now. Genuinely useful. But something keeps bothering me.

It'll suggest a flow, a layout, an interaction pattern, confidently. And I sit back and look at it and think: if a real user actually got here, this is not how it would go.

AI doesn't know your users. It doesn't know where people actually get confused, what they skim past, what makes them quietly leave a flow halfway through. It's pattern-matching, not watching real behavior.

And founders moving fast with AI (which, fair, time is real) are sometimes shipping that gap straight to production without realizing it's there.

Half of what I do in a UX audit now is exactly this, going back through AI-assisted decisions and asking, would a real person actually experience this the way it's expecting them to? Often, no.

I run UX audits for exactly this reason, catching the gaps between what AI assumed and what your users will actually do, before it costs you drop-offs or support tickets.
_________________________________________________________________

I'm Wuraola Olaibi, a product designer based who builds products with AI in my workflow.

How I use AI in my daily workflow to move faster!

The AI Step: I use Figma Make to generate a rough layout instantly. It saves me time on brainstorming and gives me a quick starting point. I just need to understand the project and also the direction the client wants!

The Human Step: I redesigned it to match the design style and what was wanted.

So AI does the most part, and I do the cleaning because I am a UI designer (I love to have clean interfaces).

What do you think of this workflow, and how do you design lately?

Before: Figma -> Webflow/Framer. Now: Ideate with Claude -> Fine-tune in Figma -> Take back to code (all 50x faster). It's genuinely insane!

Plan mode

Plan mode is a new opt-in mode that helps you shape that direction before generation starts. Make takes a look at your project, asks a few clarifying questions, and drafts a plan you can edit, refine, and approve before anything gets built.

It's most useful for complex work—multi-section layouts, detailed specs, design imports—where getting aligned upfront leads to noticeably better results. For simpler prompts, you can skip it and build directly.

Turn it on via the dropdown in the prompt box or the /plan command. Because plan mode does extra work upfront, it uses more AI credits than a standard build—you'll see an estimate before you commit.

Web search & fetch

Make can now pull live context from the web mid-build. Search broadly or fetch a specific URL to ground builds in current content. Tool-call approvals let you review before anything enters your session.

Queued messages

Stack follow-up instructions while Make is still generating. Edit or delete them before they commit, and they'll send automatically once the current build finishes.

My design workflow looks different now.

Brainstorming → Claude
Breaking down UI references → Gemini
Wireframes → Lovable
Working prototypes → Antigravity

What used to take days now takes hours.

I'm not using AI to replace my design thinking. I'm using it to move faster so I can spend more time on the decisions that actually matter.

The designers who figure this out early are going to be very hard to compete with 🤔

What's in your AI stack right now?

AI tools are everywhere right now.

But some of them don’t actually fit into real design workflows.

So instead of listing everything I’ve tried,
I broke down the 9 top AI tools I actually use in my UX workflow, from idea to prototype to code.

Not hype.
Not theory.
Just what’s working for me right now.

One thing I’ve learned:

AI doesn’t replace design thinking.
It removes friction so you can focus on what actually matters.

#AItools #AI #UXDesign #Designworkflow

I used AI design tools every day for 60 days.
No hype. Just real work.

Tested:
→ Figma Make
→ Google Stitch
→ Framer AI
→ Claude skills

Used them on actual product flows — not fake concepts.

What worked 👇
— Instant first drafts ⚡
— Quick layout ideas
— Helped when I was stuck
— Decent starting copy

What didn’t 👇
— No real user understanding
— Breaks flows beyond 1–2 screens
— Zero product thinking
— No consistency across screens

Here’s where it got frustrating:

I wasn’t designing anymore.
I was fixing.

And worse — I was fighting code 👨‍💻
instead of thinking about actual design problems.

Faster output… ⚡
but slower progress.

That’s when it clicked 💡

AI is great at generating screens.
But product design is about decisions.

Now I use it for:
— Quick drafts
— Exploring directions

And I keep the thinking:
— What problem matters
— Why this flow exists

AI won’t replace designers.
But it will expose the ones who don’t think.

Use it as a tool.
Not as your brain.

I recently tried designing a website with AI as the starting point.

The idea was intentionally simple and generic:
a camping / camper‑friendly website.
A topic where AI should perform well.

I started the process the right way not with visuals, but with thinking.
I talked with ChatGPT about the project:
- what problem the website should solve
- what kind of experience I wanted to create
- the goals of the business
- who the users are and what they expect

The result was surprisingly good.
AI helped me structure my thoughts and turn them into a clear design brief something that usually takes quite some time. That part genuinely pushed me forward.

I then took this input and moved it into an AI web design tool, expecting the visuals to naturally follow.

And that’s where things became… tricky.

The result was okay, but very generic.
I tried refining it with more prompts, more instructions, more iterations but instead of getting better, it slowly became constrained and messy. The design started to feel stuck.

So I stopped.

I took the AI-generated design and moved into Figma, this time designing manually, without AI.
That’s where everything clicked again. I refined layouts, hierarchy, spacing, interactions and finally reached a result I was happy with.

My takeaway?
AI can push you very far, very fast especially in:
- problem framing
- structuring requirements
- defining goals and constraints

But when it comes to design decisions, clarity, and nuance, AI can also slow you down or trap you in generic solutions.

AI is not a replacement.

It’s a powerful tool if you know when to use it and when to step away.
And that’s probably the most important design skill today.

#AI #UXdesign #UIdesign #DesignToday

Best AI design tools in 2026 - these are the 7 AI tools I use to run my design studio every single day. Not demos, not sponsored picks - real tools that ship real client work.

I tested over 50 AI design tools and narrowed it down to 7 that cover every part of the web design workflow: ideation, layout, assets, visuals, design, building, and the one tool that connects everything together. Each tool earned its spot by surviving real client projects at my studio, Klime.

When I use AI in my design work, I’m not outsourcing my thinking. I’m speeding up the parts of the process where speed is actually valuable, generating options, exploring directions, getting a quick read on structure, so I can spend more time on the parts where my judgment matters. That’s a meaningful difference.

I use Claude for research. I’ve trained it over time to understand the tone and voice of the products I work on. When I save a screenshot of something that inspired me, I send it over and it helps me remember why I saved it, what the design is actually doing, and how it connects to whatever I’m building now. That’s not impressive. It’s just useful.

I’ve also given it instructions for UX writing — the tone, the constraints, who the users are. When it gives me copy suggestions, I don’t paste them straight in. I read them against the product’s voice and ask whether they’re actually clear for the user. Sometimes I run tests to find out.