I made a game about UX and politics
I've been doing some career reflection and decided to practice. Not with a course, not with a book, but, you guessed it, with AI, by throwing myself at hard scenarios and seeing what came out.
One conversation led to another. Business pressure scenarios, then ethical dilemmas. The "your Head of Product is horrified, the CEO loves it, and your research says it's a bad idea" kind. The kind you only get good at by living through them. Or, apparently, by simulating them obsessively with Claude at 11pm.
Somewhere around the third scenario, I typed something I didn't expect to type: "This is by far the most useful and entertaining AI chat I've ever gotten."
Let's make a game out of that!
Play it here uxatthefirm.com
What design education can get better at
Most ways of learning UX are about craft. How to run a research session. How to structure a journey map. How to present to stakeholders without your voice going weird.
What they don't teach (at least I myself never saw that in any curriculum) is the political part: you're in a room and the Head of Product says "I don't care about the research, I care about the revenue target" and you have to figure out in real time what to do with that.
That skill gets built through painful experience, usually at the cost of relationships, projects, or your own confidence.
I kept thinking: what if you could practice that without the consequences?
The game
The concept is simple. You load the app, you get some scenarios, including (but not limited to):
A CEO who loves a dark pattern.
A Growth team stepping on your territory.
Stakeholders who've already decided what they want, and are hiring you to draw it.
You write your answer in a text area and an LLM reads what you wrote and gives you the same kind of feedback you'd get from a very direct designer, what you got right, what you missed, what a more experienced person might have done instead.
Disclaimer: it's the answer of an LLM, please don't take the feedback face-value, but more as an indicator.
I only wanted to make a create-react-app
At first, I expected to learn something about React. I didn't expect to learn something about design education.
The scenarios are the hardest part to get right. A good scenario has enough pressure that it forces a real decision, but not so much that there's only one objective answer. It has to feel like something that could actually happen, because the whole point is to build pattern recognition for the real world.
What I found, writing them, is that the best ones come from real situations. Things I've seen, heard about, or been in. The scenario where the data says one thing and the business says another. The one where the right UX decision would slow down a launch the company can't afford to delay. The one where you're pretty sure the feature is harmful, but everyone else thinks you're being precious.
That felt worth sharing!