ChatGPT Has a Serious UX Problem
If you’re a UX - UI - customer wizard, listen up: Your moment in the AI spotlight has arrived. Grab your pointy wizard hat. We have work to do.
ChatGPT has a major user interface problem.
This is not just a tiny hiccup. It might be THE roadblock standing between mere mortals and widespread AI adoption.
(BTW, this is why I’ve built a new Generative AI course. In all my research and operations work, I’ve discovered something big. Here it comes.)
The stumbling block to generative AI is not “figuring out the technology.”
It’s lowering the behavioral barriers between humans and AI.
But we need your help, User Interface Ninjas!
You have saved the world before. Remember the World Wide Web?
Consider that the internet existed before the whole WWW thing. Early internet users had to navigate through command-line interfaces and text-based systems. It wasn't until graphical web browsers came along that the internet became accessible and useful for the general public. (We owe you a debt of gratitude, UX gurus!)
THE GREAT AI PARADOX
Everyone's buzzing about how close we are to AGI (Artificial General Intelligence). We're supposedly on the brink of handing over entire companies to our AI overlords.
And yet...
Tons of humans barely use generative AI. Sure, the usage stats look impressive, but let's be real - they're as inflated as, like, a seriously inflated balloon. (Note to self: work on analogies.)
Why? Because...
INTERFACE IS EVERYTHING
Picture this: You're watching "Iron Man," and Tony Stark is chatting with JARVIS like they're college buds. Or in "Her," where Theodore falls crazy in love for an AI that sounds and acts like a real person.
Now, look at ChatGPT.
It's supposed to be this incredible, human-like intelligence (human LIKE not human LEVEL - don’t @ me). But what does it look like? A boring old search bar. Talk about a letdown.
Don't think visuals and UX matter? Let me introduce you to Hank the Pencil.
THE BALLAD OF HANK THE PENCIL
A professor once put googly eyes on a pencil and introduced it as "Hank." The class giggled, then started asking Hank questions about pencil life. Everyone was having a grand old time. What’s it like to be a pencil? Do you have any kids?
Then, without warning, the professor snapped Hank in half.
The class gasped in horror. They got legitimately mad at the professor.
It was just a pencil with googly eyes, folks. But for a few minutes, it was personified. And that made all the difference. The students literally felt something.
Which do you think would be easier? Talking to a pencil lying on the desk? Or talking to Hank the Pencil? Propped up with googly eyes? (If you’ve seen “Castaway” and Wilson, you know the answer.)
THE UX CHALLENGE: MAKING AI FEEL HUMAN
So, here's our dilemma: We're supposed to talk to AI like it's human to get the best results. But it doesn't look human. It doesn't sound human. It's just... there. A cold, impersonal text box.
This is where you come in, UX superheroes. Your mission, should you choose to accept it, is to bridge this gap.
(Do I have to give the disclaimer that AI isn't human? That's on you, my people.)
Let’s dive in.
THE UX REVOLUTION: 4 STEPS TO AI ADOPTION
Personify the AI: How can we give ChatGPT a personality without going full uncanny valley? Maybe it's an avatar, maybe it's a voice. Get creative!
Natural Interaction: Design interfaces that encourage conversation, not just query-response patterns. Think chat bubbles, think emojis, think GIFs (okay, maybe not GIFs).
Feedback Loops: Create visual or auditory cues that make users feel heard and understood. We're talking subtle animations, sound effects, or even haptic feedback.
Contextual Design: Develop UIs that adapt based on the user's task or emotional state. Serious research mode? Calm, minimalist interface. Creative brainstorming? Bring on the playful elements!
THE VOICE FACTOR
Voice interaction might be the big step here. Think about it - in "Her," Samantha was just a voice. But she felt real, reactive, emotional.
We're not far from that reality. But it raises questions:
Is it practical for everyone to be talking to AI all the time?
How do we handle privacy concerns in public spaces?
Can we create voices that are pleasant but clearly artificial to avoid the uncanny valley?
Ethically - are we treading in murky waters here?
THE BOTTOM LINE
The AI adoption problem isn't just about technology. It's about psychology. It's about making people comfortable interacting with a non-human intelligence.
And that, my friends, is where UX comes in.
So, to all you UX magicians out there: We need you. The future of AI adoption is in your hands. No pressure or anything.
Thanks in advance.
This course is designed to be interactive, engaging, and yes, even a little bit challenging in the best way possible.
Booking Conor for an Event
I’m thrilled that I get to do a ton of keynotes and lead workshops for companies. If you’re looking for a highly practical, framework based approach, feel free to reach out at www.ai-mindset.ai/speaking
