Summary: Generative AI (Gen. AI), like ChatGPT, introduces novel interaction patterns in human-computer interaction (HCI). These patterns prioritize fun and user enjoyment, significantly shifting from traditional HCI approaches. Bring these UX patterns to your AI, UX, or Service Design project.

Remarkable UX patterns used in AI you can adopt

The following interaction patterns with ChatGPT mark a shift in how we think about UX Design challenges.

In this part 1 video on Remarkable AI UX patterns, Frank Spillers covers essential aspects of Generative AI interactions, many of which borrow from Emotion Design.

See What does emotion have to do with design? 

• Fun as interaction pattern: ChatGPT’s playful interactions offer a delightful user experience. Its ability to generate non-stop content makes work more fun than traditional interfaces. AI initiating user tasks departs from do-it-yourself interfaces that have users configuring, searching, and creating from scratch.
• Automating workflow: Gen AI lessens the user’s workload. The user directs the AI’s performance instead of doing all the work, as has been the case before Gen AI interfaces. Gen AI automates tasks like content creation and making sense of the context and meaning of information. This cognitive amplifier allows users to focus on higher-level thinking and decision-making rather than mundane tasks. The user can go back to focusing on what they want instead of on operating a UI.
• Fluid human-computer interaction: Gen. AI represents a paradigm shift in human-computer interaction- the field from where UX comes. It moves beyond rigid command-response models to more fluid, conversational interaction. This makes task accomplishment more imaginative, emotional, and human-like, the way brains work.
• Designing for Pleasure: Incorporating pleasure and fun into design is crucial. Enjoyable interactions increase user satisfaction and engagement. Designers should prioritize these elements, ensuring designs streamline tasks and deliver a pleasurable user experience.

ChatGPT’s interaction patterns emphasize fun and fluid ease of use, changing how we interact with computers. This approach can significantly influence future HCI designs, where user delight is as important as functionality.

Transcript:

“There are some remarkable things to notice about generative AI. When I was first learning AI in 1997, I stopped my Professor one day, who was a top AI thinker, and I asked him: “Could we embed AI in a virtual environment as an agent?” He turned around and said, “No! That’s preposterous!” and since then, I have been thinking about how achievable or how that might play out from a UX perspective. These days, it’s not a huge idea. Its pretty straightforward, so there’s no innovation there.

See Generative AI agents in VR

But what I’d like to talk about is what you can notice from generative AI… The part that my professor did not see was the >>fun<< part. Now generative AI and ChatGPT, in particular, who I think is a market leader in this area, at least are showing their interaction patterns are much more progressive than, say, Google Bard…(at the time of filming), so generative AI makes it feel like…not work! And that’s a really important thing, what do they do? They do fun! I would recommend a book to you called Funology but I won’t because it cost about $175 normally should be about $20 or $30 … And it’s been like that since I’ve been teaching Emotion Design for over 15 years.. Every time I look. The book is still really expensive, and the reason is that Funology, or being able to bring fun to your design, is a really, really important thing…

See The Science of Fun: How Fun Helps Improve Your Design?

So that’s the first thing to notice is to make it fun to look at how ChatGPT works… even when you give it feedback… When you give negative feedback, it automatically generates an alternative response, which is sometimes better, and it even asks you: “Was this better?” So it’s acting more like a Customer Service agent, like a real person.

Unique characteristics of ChatGPT

It’s got manners; It’s been socialized; They’ve tried to remove bias; They’ve tried to remove harm and tried to remove discrimination… and, no it’s not racist, it won’t help you make a bomb, or bring down the government. So it’s been trained in the same way that an employee wouldn’t help you do those things… …or treat you in that way. The second aspect of this fun nature is that it “doesn’t feel like work”. You go to Microsoft Office; even Google search feels like work! You’re typing in, you’re having to look, you’re having to click on links… You know Microsoft Office before AI is just like a canvas that requires you to output- that’s work: typing and thinking and typing… So we’ve moved into this new paradigm with AI, which is not typing and thinking necessarily but playing: making an “offer.”… It’s more like improvisation or “improv” where you make an offer: you say, “Hey, tell me about this or write this for me”… or: It’s really great at writing customer service complaint letters by the way… If you have a bad airline experience, as I had, They’re really good, at least (for an American context). They might not work in other situations, as colleagues told me. So, the fun aspect of it is critical.

The second one is in this “not feeling like work” is collaboration. You can lean on– not only funology, which comes from emotion design– but also collaborative UI’s and the concept of sociability– which is a really, really important concept of how social interaction occurs. Now we’re seeing sociability and collaboration as well as fun as default to the UX paradigm that generative AI, at least ChatGPT, is using…

Applying these unique differentiators to your design

So we can start using that in our own designs. We can make our own designs, and if they have AI, great! So they should be relevant first of all, but make it fun; Make it collaborative, make it helpful. Take the work off– you know, we’re just constantly logging in, filling in forms, searching, and comparing: It took me 5 to 10 minutes just to search for a photo on my desktop, and generative AI could have done that in seconds: looked at all my files and pulled up the images of whatever that object was I was looking trying to remember. So take the work off the user, that’s one of the ux principles. We’ve done a terrible job of it in UX… Everything is]work for the user– in other words: stress, strain, friction… And we’re trying to make it easy to use, but while easy to use is important… Fun is actually one vehicle to do that…So you can look at gamification as a way to do that… You can look at fun, or “funology” and “pleasurability”.

See Pleasurability and Emotion Design