#38 The User Experience of AI Products on Future of UX
13:47 Share

How do you design great user experiences for AI products? This is the question of today's episode.

So there are a lot of discussions around how to use AI for your workflows, what are the best tools to be more productive and what kind of like which tools are really helpful for us as UX designers. But there's also a second super important thing, which is how do you actually design and create amazing user experiences for AI products, AI features, AI software, right?

So hello and welcome my friends to the future of your ex, your podcast where we dive into the future of user experience design and make sure that you stay up to date with everything that's happened in the tech and in the design industry.

So my name is Patricia Reiners. I'm an UX and innovation designer from beautiful Berlin. I'm running my own business here, focusing on all kind of future UX topics, consulting clients, doing my own courses and workshops and also sharing content on social media. You find me at ux.patricia53 to just say hi. I am always super happy when people are connecting and just like saying hi.

So in this topic, we will talk about some really exciting things, but also some challenges when it comes to user experiences for AI products. I'm also going to bring in some additional examples and perspectives. So I would say, let's get started. To give a little bit of background information, I would say we start with the current state of UX in AI. When I'm saying AI, I mean AI products.

software AI products like ChatGPT, Midjourney, all these AI tools. And I think what is pretty, pretty obvious is that AI is everywhere, you know, from chatbots to image generation to text generation to Figma plugins.

But I would say besides its rapid adoption there is a glaring issue and this is usability. For instance, you know when we are thinking about mid-journey there are a lot of really big problems when it comes to basic usability.

So you focus on like the tech savvy people, people who are willing to take a lot of hurdles, who are not, you know, who take the time to install it and then find a workaround if things are not working as well. But there are also a lot of people, users who are not that tech savvy, who are having a lot of challenges and are really in a usability nightmare.

I would say the good thing is that at the moment or based on a lot of research and years and years of web design experiences and sharing insights and learnings, you would think that there are a lot of usability and design principles, right? You would think that in the age of innovation, basic usability principles would be given and

But what I'm seeing is that they are often overlooked. And when I'm talking about basic usability principles, one thing that I think is very, very helpful is, for example, from Jacob Nielsen, he came up with the 10 heuristics of interface design. They are still very valued, although he came up with these principles years ago, but

But they are still valid. I can also link them in the show notes so you can have a look. And they apply for traditional design or for web design, app design, but also for AI products. And there I want to dive a little bit deeper into one topic that I find pretty interesting. That is mid-journey.

For everyone who hasn't used Midjourney, Midjourney is a tool that you need to download through or that you can use through Discord. Discord is something like Slack but for techies or for gamers. And there you have an option to chat with a chatbot and the chatbot creates these images for you based on the prompts that you enter.

There are a ton of horrible usability issues that I am seeing currently in the app. A lot of consistency issues, accessibility issues, it's super hard to also really find the content that you created and also a lot of things that are not consistent when we think about

all the settings and you can control right so they have different kind of measurements for any kind of things there's a huge challenge to create

a prompt that really makes sense and I think that's really confusing. Also Jacob Nielsen who came up with the 10 heuristics of interface design shared a super interesting article on LinkedIn where he talked or where he went through the mid journey experience and shared a few things that are not working so well from his side. I think this is definitely really interesting.

And what we always need to keep in mind is that there isn't a dilemma of the early versus the late adopters. Early adopters are the guinea pig of technology. I would definitely include myself in that area. They're willing to overlook these flaws.

But what about the mainstream audience when we think about our parents, our neighbors, you know, people who are not that tech-savvy? They won't tolerate a clunky interface. So imagine your mom, you know, trying to navigate a poorly designed AI health app. It's not just frustrating, it also could be very dangerous. There are new UX challenges introduced by AI. Yes, something very exciting is coming or is already there.

AI isn't just bringing old problems. It's also introducing new ones. For example, the whole topic about ideation abundance. So it hasn't been easier to create content. Content creation is the huge thing, right? Like you just need to enter a prompt and then you get content for mid-journey. You enter a prompt, you get a lot of images on mid-journey. And this is exactly what this is referencing about. Ideation abundance, right?

With AI you can generate endless design variations. But the big question is: How do you manage them all? It's like having a Spotify playlist with thousands of songs but no way to sort or favorite them.

I think this is a really good example because this is what is happening. Like a lot of content, but a big challenge to really organize them, sort them and help yourself or help the user to make a judgment. What is actually the best solution?

So, creating content has never been easier: chat GPT, mid-journey, etc. But getting an overview of the results and judging that and understanding or getting your view on what's the best outcome, that's a huge challenge. So, let's look at some real-world examples. AutoML is a very powerful tool, but its interface is overwhelming for newcomers.

Although ChatGPT has a very intuitive interface and I would say of all great usability, it is still challenging to get the desired outcome. The term prompt engineering is there for a reason. I did a whole podcast episode about that topic, so feel free to check that out. And you strategically really need to frame the prompt and also know what the desired outcome should be.

On the flip side, there are also tools like Canva or Adobe Firefly, who have integrated AI much, much, much more seamlessly, offering design suggestions without complicating the UI. And for us, also really interesting to have a look at the companies or the tools who are already going in the right direction. And the most interesting thing, you know,

like Canva or Adobe Firefly or these things they already present to the user what is possible so the user doesn't start with a blank page but they already get recommendation and have some kind of like a starting point everyone knows that it's much easier if you get some kind of like background information and tasks and things that are possible instead of just starting with the blank paper

Another example that I find super interesting and I talked a lot about that topic in other podcast episodes or on Instagram is

is TextFX from Google. It is another very interesting example because this tool is a text writing tool that offers already different functions. You don't start on a blank page but you start with a function like POV. You evaluate a topic through different points of view. So I enter sandwiches for example just like one word and it results is that

um you know it creates different kind of sentences that i can use for my creative process example sandwiches are a great way to use leftover ingredients can also adjust the temperature of how creative the results will be with just a slider and then in the end you can go through the different results just by like just like a page you know you're going through the different pages

You don't only have the answer of sandwiches are a great way to use leftover ingredients, but also sandwiches are the perfect food for a quick easy lunch or sandwiches are the gateway to culinary exploration and so on. I really like that example and I can highly recommend to check this tool out. Not a lot of people know this tool, have tried it out, but I think it's an amazing example of how you guide the user through the process.

Let's go to the next important thing that I want to mention, and that's the need of UX investment in AI. You would be surprised how few UX professionals are employed by leading AI firms. And it's like building a car with no one in charge of making sure it's comfortable to drive. Companies need to invest more in UX, period.

I'm already seeing that this is going to happen as soon as more research and testing is going to happen. And at that point, these companies are going to hire UX designers. And what will be important for these kind of roles? To find people who already have experience working for and with AI tools, who are open, who have the right mindset, and who understand the section.

So where do we go from here? I would say AI needs to be more human-centric. For example, you know, future AI could understand user emotions, which is definitely something that's coming through facial recognition and also adapt its responsive accordingly. The possibilities are endless, but they require a strong UX foundation. To sum it up, I would say great UX in AI is not just like a nice to have thing, it's also a must-have.

And as designers, it's our job to bridge the gap between the technological innovation and the human needs. And what we are already seeing is when we just have a look at the current situation is that a lot of these AI products are really lacking basic usability capabilities.

Why? Because there are not enough UX designers involved in the process. If there are, the products are much better. If we look at Canva or Adobe, we're doing it much, much better in general. So also these companies who

are lacking usability are going to hire your ex designers focusing on ai in the near future so if that's an area that you want to learn more about if you are excited about that there are probably a lot of interesting opportunities coming up for you in the near future right

because we are already seeing AI integrated here and there but that will change. AI will be some kind of like a layer and integrated in basically everything and that will also change the way how our interfaces work. Super exciting times ahead!

Okay, friends, thank you so much for joining this episode. If that was helpful to you, feel free to give this podcast a five-star review. This is so helpful for me to keep the podcast growing and get it more visibility. I would be super happy to have your support here at that point. And overall, thank you so much for listening. If you have any questions or want to reach out, you can find me on Instagram at

and I would say my friends thank you so much for listening and hear you in the future