AI is not coming for your job. Your complacency is. The tools have changed every five years since I started — Photoshop to Sketch, Sketch to Figma, Figma to whatever fresh hell is next — and every single time the same designers panic while the same good designers adapt and keep shipping.

I've watched grown adults spend more hours tweeting about whether Midjourney can replace a brand designer than actually doing brand design. That's the problem right there. You're so busy being terrified of the machine that you forgot to get better at being human.

AI can spit out a thousand logo options in ten seconds. Sick. Can it sit across from a founder who has no idea what they want, pull a coherent vision out of that mess, and make them feel like you read their mind? No. It bloody well cannot.

Your taste. Your judgment. Your ability to look at something and say "this is rubbish, bin it." That's the job. That was always the job. The medium changes. The standard doesn't. Stop doom-scrolling prompt engineering threads and go talk to a real user. Remember what thinking feels like without autocomplete.

Dieter Rams wrote ten principles of good design in the 1970s. They're perfect. They're timeless. And roughly 98% of working designers have never actually applied a single one of them in production. We print them on posters. We cite them in case studies. We do not follow them.

"Good design is as little design as possible." Now open your latest Figma file and count the bullshit. The decorative gradient nobody asked for. The icon that means nothing. The toggle that toggles another toggle. Yeah. You know.

Minimalism is hard because it requires you to say no. No to the client. No to the PM. No to your own ego when it wants to add one more thing. Every element you remove is a fight. Every pixel of whitespace is a negotiation. Rams made it look effortless because he spent decades getting there. We want the output without putting in the years.

The best work I've done in 13 years is the stuff where you can't see the work. Nobody emails you saying "wow that negative space is incredible." They just use the thing without thinking. That's the job. Invisible design. If Rams walked into most design teams today he'd walk straight back out. Fair enough honestly.

I've had this conversation maybe 300 times. Client looks at the design. Points at the gap. "Can we put something there?" No mate. That space is doing more heavy lifting than the CTA you want to cram into it.

Whitespace isn't absence. It's presence. It's the pause between notes. Every great piece of design — from the original iPod box to the Nothing Phone packaging — uses negative space as a structural, deliberate decision. Not laziness. Architecture.

The urge to fill comes from fear. Fear the design looks "unfinished." Fear the CEO thinks you slacked off. But stuffing more into a layout doesn't make it more useful. It makes it a car crash. You're not designing a newspaper from 1997.

Next time someone asks you to fill the whitespace, pull up google.com circa 2004. One field. One button. Two trillion dollar company. The space was not wasted.

1972. Massimo Vignelli redesigns the New York subway map. Geographically it was a disaster. Central Park was a square. Brooklyn looked like it floated to New Jersey. The East River went on a diet. New Yorkers absolutely lost it.

But it worked. As a piece of wayfinding, it was a masterclass. It prioritised the user's actual task — getting from A to B on a rail network — over geographic accuracy that nobody needs when they're underground. You don't need to know Manhattan's exact shape when you're transferring at Times Square. You need the coloured line.

This is the tension we live in. Accuracy vs clarity. Completeness vs usability. The map got replaced because the public couldn't get past the distortions, but Vignelli's thinking changed how every designer since approaches information. Sometimes you distort a little to communicate a lot.

Every dashboard I design, I think about that map. What does the user actually need right now? What can I simplify or kill entirely? The answer is always more than you think.

Custom tokens. Semantic colour naming. Responsive spacing scale. Auto-layout on every frame. Beautiful documentation page. Used by exactly zero people after week two.

Design systems fail for the same reason diets fail. High initial motivation, unsustainable execution, and six months later everyone's duplicating frames and eyeballing padding again. I've built systems for five-person startups and 500-person orgs. The failure mode is always the same: nobody owns it after launch.

Here's the dirty secret. Design systems aren't a design problem. They're an org problem. They need governance, ownership, and someone whose literal job is maintenance. Most teams can't or won't resource that. So the system becomes a museum — nice to visit, irrelevant to actual work.

Start way smaller than you think. Five components. One colour scale. Consistent spacing. Done. A system that covers 60% of cases and gets used beats one that covers 100% and gathers dust. Ship the minimum. Maintain it properly. Add things when the pain is obvious. And for the love of god stop naming your design system. It doesn't need a brand. It needs to work.

I used to design interfaces I wanted to look at. Now I design interfaces that disappear. That shift took eight years and a lot of user testing sessions where people completely ignored the thing I was most proud of.

When someone completes a task without friction, confusion, or conscious thought about the interface — that's the highest achievement. The design was so good it became invisible. Nobody writes you a Slack message saying "the padding on that card was perfect." They just finished their task. That's the win.

This is fundamentally at odds with how our industry works. Dribbble celebrates visual spectacle. Awards celebrate animation. None of them celebrate "user completed checkout in three seconds and didn't notice the design." But that is the actual goal.

Aesthetics matter — they build trust, signal quality, create emotion. But when they come at the cost of usability, you've made a painting, not a product. Last time I checked, paintings don't have conversion rates. Design for forgettability. I'm serious.

Sketch died because it couldn't collaborate. Not because Figma was better — early Figma was rough as guts — but because Figma understood that design had become a team sport while Sketch still thought it was a solo act. The tool that wins is never the most powerful. It's the one that fits how people actually work.

So will AI kill Figma? No. AI is a feature, not a product. The thing that kills Figma will be whatever correctly identifies the next shift in how design teams operate. And I think that shift is the death of the design-to-dev handoff.

The handoff is where design goes to die. Pixel-perfect Figma files get interpreted by developers working on completely different mental models. The result is always a compromise. The tool that makes design and code the same artifact wins everything. We're seeing early versions — tools that generate production code from design, tools where designers work in code without knowing it.

Figma knows this. That's why they're pushing dev mode, variables, code connect. Whether they get there fast enough is the question. History says incumbents rarely do.

Oliver Reichenstein said "web design is 95% typography" in 2006. Nearly two decades later designers are still spending most of their time picking colour palettes, tweaking gradients, and arguing about border radius. Meanwhile the type is 16px system font with default line height. Mate. Come on.

Typography is the foundation of visual hierarchy. Nail it and everything else falls into place. Get it wrong and no amount of colour theory saves you. A well-set piece of type on a white background will always outclass a poorly-set piece on a gradient mesh with glassmorphism and parallax. Always.

My process: start every project in black and white. No colour. Just type, spacing, hierarchy. If it doesn't work without colour, it doesn't work. Colour is seasoning, not the meal. You wouldn't pick paint colours before building the house.

Pick a typeface with intention. Set it with care. Give it room. That alone puts you ahead of 80% of the industry. I guarantee it.

"So you make websites?" Sort of. "Like an artist?" Not really. "But you use a computer?" Yes. "So you're a developer?" No. "What do you actually do?" ...I make things easy to use and nice to look at. "Oh like decoration?" I need a beer.

This conversation has haunted me for 13 years. The industry hasn't helped either. We rebrand ourselves constantly. Graphic designer. Web designer. UI designer. UX designer. Product designer. Design engineer. At some point we'll run out of prefixes and just use emojis.

The core problem is that good design is invisible. People notice bad design instantly — confusing nav, broken flows, ugly layouts. But good design feels... normal. Expected. So the better you do your job, the less anyone understands what your job is. Beautiful paradox.

I've stopped explaining. Now I just say "I solve problems you didn't know you had, with tools you've never heard of, for people who'll never thank me." Gets a laugh usually. Sometimes a concerned look. Either way, conversation moves on and I can get back to arguing about 4px vs 8px grids.

Agency life: brief, research, ideation, concepts, refinement, presentation, revision, revision, revision, sign-off, delivery. Six weeks minimum. Usually twelve. Sometimes six months for a damn logo.

Web3: you have three days. Community wants to see progress tomorrow. Founder changed the roadmap at 2am. Token launches Friday. There's no brand guideline. No style guide. Barely a brief. Go.

And honestly? Best thing that ever happened to my work. Not because chaos is good — it's obviously terrible — but because it forced me to strip design back to fundamentals. When you can't hide behind process, you rely on taste. On instinct. On the decade of reps that tells you "this works" before you can explain why.

Web3 taught me speed without sacrificing quality. Taught me to ship imperfect and iterate publicly. Taught me that a beautiful product nobody uses is worse than an ugly product everyone uses. Most importantly, it taught me design is a service, not an art form. You serve the user, the product, the business. In that order.

Every designer should spend six months in a high-velocity environment. It'll either break you or forge you. The ones it forges are dangerous.