UX AI Design Anti-Patterns

Here’s a AI UX translations tutorial for Starters, Junior, Mid-level and senior UX designers and researchers,…

Listen to AI UX Speak at work, at home and on the road! It’s free.

Listen to AI UX Speak 001. It’s free.

This quirky AI audio-guide is like a podcast on caffeine. It runs for a solid 18 minutes. It is perfect for those who can’t decide between a quick coffee break or a deep dive!
It simplifies and demystifies the elusive UX design speak that often sounds like a foreign language to practitioners and users alike.
It hilariously breaks down the peculiar dialects of UX design collaborations. This is especially vivid when AI contributes its insightful, albeit automated wisdom into the mix.

In this well-researched and somewhat chatty deep dive, the AI dishes out on how designers banter both amongst themselves and with stakeholders, all while playing a game of “Who’s Who?” across varying experience levels! With relatable examples sprinkled in, it sheds light on the confusing meanings behind phrases that make you wonder if you accidentally signed up for a secret society.

I have a hunch this will be a handy tool—like a GPS in a corn maze of design communication and collaboration, which can sometimes feel like a game of charades.

Curious if you want to hear more of these laughs in the future?
Don’t hold back—share your own Speak in the comments below!

There is a systemic AI prompting reason for the symptomatic, recurring picture’s depiction of fake happiness at work. Why should they all look so happy?

Not to speak about their age, their skin’s, exquisitely detailed plastic texture and fake expression of emotional intelligence.

Feel free to add your own UX Speak below in comments! It will fill us with joy.

The REAL Enterprise AI Design Anti-Patterns: A Translator’s Guide

Ever wondered what UX designers and researchers actually mean when discussing AI projects? Let’s decode the corporate-speak across designers’ experience levels, shall we?

STARTER LEVEL to Junior level

Politely Clueless:

• “I’ll make AI more user-friendly” = “I’ll add a chat bubble icon and call it a day”
• “We’ll focus on AI transparency” = “I’ll put ‘AI-powered’ badges everywhere”
• “I’ll create an AI interaction flow” = “I’m going to copy ChatGPT’s interface”

Naively Optimistic:

• “I’m just the AI UX designer” = “Please don’t ask me about prompt engineering”
• “Let’s leverage AI design trends” = “I saw this cool thing on Midjourney once”
• “I’ll make it human-centered” = “I’ll add loading animations to hide the API delays”

MID-LEVEL

Tactically Aware:

• “We need user research on AI adoption” = “I need data to prove this AI feature is confusing everyone”
• “Let’s explore AI affordances” = “Nobody knows when they’re talking to a bot vs. a human”
• “I’ll prototype AI interactions” = “I’ll fake it with hardcoded responses until we get API access”

Pragmatically Cynical:

• “We should run an AI design sprint” = “Let’s spend a week discovering why this LLM won’t solve our problems”
• “Let’s test different AI feedback loops” = “Users hate all versions, but we need metrics”
• “We need to improve AI transparency” = “Legal made us add 47 new disclaimers”

SENIOR LEVEL

Aggressively Experienced:

• “Let’s rethink our AI integration strategy” = “This is the worst implementation I’ve seen in my 15-year career”
• “We need a holistic AI system approach” = “Stop slapping LLMs onto everything that moves”
• “Let’s discuss AI governance frameworks” = “Your DIY LLM will cause a PR nightmare”

Destructively Honest:

• “We should align AI with user needs” = “Users don’t want AI, they want things that work”
• “Let’s evaluate the AI platform requirements” = “Your startup’s ‘AI agents’ are just if-else statements in a trench coat”
• “We need to consider AI edge cases” = “This will hallucinate in exactly 42 different languages”

The Real Talk Corner:

• “The AI needs more context” = “Your training data is garbage”
• “We’ll implement RAG architecture” = “We’re admitting the LLM can’t remember anything”
• “Let’s create an AI feedback system” = “We need proof when this thing goes off the rails”

Remember: When a UX designer says “It depends” during an AI project review, they mean “This whole thing is a house of cards.”. They mean “This whole thing is a fragile setup.” They might be too tired to explain why. When they suggest “exploring innovative AI interfaces,” they’re really saying, “Please, for the love of usability, do not mimic ChatGPT’s design.” They want to revamp it for better usability.

P.S. If your UX researcher says they need to “validate the AI user experience,” get ready. You will receive a 100-page report. It will explain why users are both fascinated and terrified by your “revolutionary” AI assistant. This assistant can’t reliably add two numbers.

The key to (some) success? Stop treating AI like magical fairy dust and start building actual integrated systems that solve real problems. But hey, what do we know? We’re just the designers who have to make this stuff usable for actual humans. ?

Listen to Speak AI UX design 001, available for free.

Interact!

Tell me what you think, this will make that thing, well, better.

Follow the Speak Podcasts here