Thoughts on design, technology, and the things in between.

Some thoughts should not disappear

This journal is a place to give ideas a little more room. What you find here is not just commentary, but perspective. Ongoing reflections shaped by practice, observation, and a bit of curiosity.

↓ Highlights

Infinite Options. Zero Filter.

AI creates options. Strategy decides what becomes real. Generation stopped being the bottleneck a while ago. Nobody rebuilt judgment to match. New article is about that gap — what lets brand design drift into the wrong direction, why "possible" isn't a strategy, and the three questions I use to filter instead of just generating more.

↓ Latest articles

More Use, More Anger: The Mood on AI Is Turning

More use. More anger. The mood on AI is turning, and it's not because the technology got worse. It comes down to what happens once the demo ends and a business has to make the thing actually hold. I wrote about where the anger is coming from, and what it's revealing about companies who thought a tool would fix a structural problem it never touched.

The real cost of replacing humans with AI

Klarna fired 700 people to prove AI could do the job. A year later they were hiring for the same roles. That's the part every "replace to save money" pitch skips. It counts the salary you remove. It never counts what you have to add back in: oversight, review cycles, the person catching the edge case nobody wrote a policy for.

Taste Is Not Enough – The Visual Designer in 2026

AI can now read a layout, describe the hierarchy, generate something passable in seconds. That doesn't kill visual design. It kills the cover that vague craft used to hide behind. If you can't explain why something works, a model can fake your output without your reasoning attached. The real test was never whether a machine can parse your work. It's whether a human still wants it.

Built for AI. Broken by Drift.

You built the foundation right. Clean tokens. Precise documentation. Brand logic translated into rules an AI can actually use. And the output still drifted. Not obviously. Not all at once. Just close enough to pass review, repeated enough times to quietly shift everything. That's the failure mode nobody talks about. And it has nothing to do with your foundation.

Adjusting on the Go – Digital Production Reality These Days

Five teams. One product. Five different directions, and nobody planned it that way. That's digital production right now. Design first, code second stopped being the rule a while ago, it just took longer for anyone to say it out loud. Teams prototype in code before design opens a file. 

Keep up with tech – The Relevance Window Is Shrinking Faster Than We Are

The time it takes to study a new technology now exceeds that technology's relevance window. That's not a motivation problem. That's just the math. New article on why chasing the pace is the wrong game, and what to do instead.

Designer's Tool Stack – a staffing exercise, not a shopping list

Design Track and Code Track on my team never share a single tool. I stopped starting with tools and started with the workflow instead. Drew it out, broke it into the activities it's actually made of, then assigned one tool to each. Not the other way round. Worth running this exercise on your own workflow before the next tool renewal lands on your desk!

From Human-Readable to Machine-Parseable

Bold but approachable. Premium, but not distant. Human-first." Beautiful. Completely useless for execution at scale. There's a translation layer missing between brand strategy and the systems that execute it. New article on what that looks like — and why nobody owns it yet. 

Designed by Humans. For Humans.

Most agencies call it transformation. Faster pipelines, AI-powered workflows, new platform, big stage moment. Under the hood, it's the same business logic at higher speed. The question nobody asks first: how is human behavior actually changing, and what does that mean for what we build? Knowing what not to build is a human judgment call. So is knowing what your customers actually need before you reach for the tool.

Human-Led Design, Systemised for AI.

Most AI workflow conversations start with the toolchain. Which tool, which pipeline, which model. Wrong starting point! The bet worth making: human-led design, systemised for AI. Taste embedded in the system before generation begins. Not as a principle – as a production architecture.

Before the System, Fix the Brand

A design system won't fix an unclear brand. It will scale the confusion faster. Before tokens, before components, before governance, there's a question that has to be answered: what does this brand actually stand for? If nobody in the room can answer that clearly, execution is the wrong next step.

One Step Ahead. One Layer Behind. The UX knowledge gap.

UX has always encoded knowledge for humans. Designers can read between the lines. Agents can't do that. They only know what you've codified. And most design systems, brand assets, and product documentation were never built to be codified. They were built to be read. That's the gap. 

Faster output to Bigger mess: AI will not fix your design system

AI will not fix your design system. It will expose it. Design-to-code sounds exciting: Figma to code, tokens to components, Storybook, GitHub, faster production. But automation only scales what is already there. 

The future of the Internet

Nobody is coming to your website! Not because your design is bad. Because the audience has changed. AI agents now research, compare, and buy. That's not a future scenario. The traffic numbers are already moving. New article on what this means for channels, advertising, digital products, and the designer's role.

AI vs. Design – Don't fight the wrong fight!

Everyone's debating whether AI will replace designers. Meanwhile, the real cost is happening somewhere else entirely. The firms that win are not racing to generate assets fastest. They're the ones who understood what mattered before they started designing.

↓ About the author

Micha Brandt is a Design Lead based near Düsseldorf, with over twenty years in digital product design. 

He builds digital products as well as the systems behind them — governance, structure, and the workflows that connect design to development, especially at enterprise scale and across multiple brands. 

This journal is where he writes about that work and the questions around it: what makes a design system actually get adopted, where AI fits into design and development pipelines, and what craft means once you're building for scale rather than a single screen.

↓ Archive

AI & Future of Work

Design Systems

Design Thinking

Industry & Change

Craft & Creative Culture

Leadership & Experience

©2026. All rights reserved.