#58 Text-to-Interface AI Tools: Threat or Treat?
11:14 Share

Hello and welcome back to the Future of UX podcast. My name is Patricia Reiners. I'm an innovation and UX designer from Zurich, Switzerland. And in today's episode, we will talk about text to UI tools. There are some interesting tools out there and also some interesting updates. So we will do a quick deep dive into what it means if you can use text prompts to generate whole interfaces.

but before we are diving into the topic i have a quick announcement i prepared a free ai mini training for you it's a little bit longer than 30 minutes so pretty quick and comprehensive in the mini training you learn three strategies to leverage generative ai to really keep up to date and to use all these new tools

you can find the link to the mini training in the description box it's for zero euros so you don't need to pay any money for it

And the cool thing about the training is that it, I already mentioned, it guides you through the three strategies and it also comes with a little workbook. So there's an exercise that you can do to really get started. Download the free workbook, do all the exercises there and use this as a perfect starting point if you feel a little bit overwhelmed with all the tools, everything that's going on.

I think that's a perfect starting point for you. So as I mentioned, look in the description box and do the training right after you listen to this episode. So in this episode, we want to talk a little bit about text prompts to interfaces.

We all know tools like ChatGPT or Midjourney or other AI content generation tools. And these tools usually work like text-to-text. ChatGPT, for example, are all text generation tools, large language models. You enter a text prompt and what you get out there is a text prompt. Or if it's a multi-model, you can also generate audio or images, for example, with ChatGPT, which is a multi-model.

can generate text, text to text or text to images because of the DALI 3 integration. And now there are also other interesting tools like text to interface or text to task.

And text-to-interface is pretty interesting for us as designers, especially all the digital designers, all visual designers. So you basically enter a prompt like "write me" or create a interface for a library app for example and then the tool creates the separate screens for the library. And maybe some of you might remember the tool Galileo AI.

The tool shared a video, I would say around a year ago, that went kind of viral. A pretty interesting tool where you could see people, you know, like writing different prompts and then the interfaces were created within seconds basically. And I almost forgot about this tool until I saw a post on LinkedIn and someone posted about this tool and tried it out. So I immediately wanted to check this tool out and see what it has to offer.

Galileo AI is not the only tool out there. There are also tools like UISR for example or Visily. I will link all the tools in the description box so you can check them out. And I just wanted to see a comparison and see what really works so what I could integrate in my own AI workflows.

So Galileo AI especially caused quite a stir with this promo video that I mentioned and now this tool is finally available to the public and I tried it out and I must say it's not bad. I can definitely recommend to check it out you can find the link in the description box and the exciting part is that you can before it creates the interface for you

you iterate a little bit on the interface. So you start with a first prompt like I did an example around a tracking app, like a productivity app. So my prompt was create a productivity app with tracking functions. And then it continued with a text prompt. So, okay, do you mean that I should create a home screen, a profile screen, a setting screens and another screen?

something like that and then i needed to say okay and now i i want to iterate i don't need the settings screens i want to have this combined with the profile screen for example or i say okay perfect go so you iterate via text in the first in the first part of the process where you really continue or where you really focus on the the concept the first the concept and then it creates

the interface or the UI design, which I find pretty interesting compared to the other tools like UISart or Visily. You start right away. So you describe everything and then it creates the content for you. And Galileo AI does it very differently with doing a lot of iterations first, really creating this concept and then creating the designs.

So I think this is pretty interesting. And after that, it creates the single screens for you and you can iterate these screens. You can do changes via a text prompt by just like, you know, saying, oh no, I want a red button there in the center, for example. And then it does these changes or you can continue in Figma.

So what I find pretty interesting is to see how the workflows are changing here and it's only a matter of time before these tools, whether it's Galileo AI or another tool, really allows us to also feed our own design library and then design screens based on these design library elements basically.

I quickly want to go over the things that I personally really like about Galileo AI or all these tools and some things that I think are a little bit difficult at the moment because there is there are quite some discussions around this. I shared a post on Instagram the other day and

And yeah, a lot of interesting discussions happen. Some people are a little bit scared of, oh wow, this is basically like taking over my job, everything that I'm doing. Others are saying, oh no, it's completely crap what's coming out there. And my perspective is a little bit in the middle. So let me guide you a little bit through the pros and the cons of these tools.

The first thing that I really love about Galileo AI especially is that you iterate before the tool starts to create something. So you knit down the concept together and I think this is super interesting and also very important of the process, right? So first the concept, then the design. The second thing that I really like is that you can iterate on the screens with text prompts. So it's not...

like it is and you need to use it or don't use it, but you can iterate via prompts.

The third thing that I really love is that you can continue to work or copy those screens into your Figma file. This is probably what most people are going to do, right? Like doing all those changes in the Figma file, which is much more convenient and much easier than iterating via prompts, if you are a designer, right? If you are not a designer, then you need to continue with the prompts.

Some of the things that I don't like about these tools, or this tool especially like Galileo AI is,

Sometimes you have very very long loading times. So sometimes it takes forever, you're not 100% sure if it's still on track or if something happens so you need to go over it again. I assume that they are going to fix that. There's probably a super high demand at the moment. But definitely a little bit annoying when you are used to a fast-paced design world.

The next thing that I think is definitely not ideal is if you compare the designs to the design set you as a designer will probably produce, of course they are not that great. All the designs won't win a design competition. But I also think they are not bad either. So when it comes to very basic prototypes, something that you create in a workshop, something where you need something very quick, that doesn't need to be perfect,

I think it's a great tool because it's not that bad either and you can iterate. The third thing that I find a little bit problematic is iterating with prompts. Although I think it's amazing that you can do that, but if you are a designer, you want to iterate quickly. For you, it's much easier to continue in Figma.

and just leave it as it is and then copy all of the things to Figma and then do all the changes there. So I think iterating via prompt is quite tedious and not the best way to iterate because it's very difficult to describe every single thing that you would like to change.

And there's another resource for you that I find pretty interesting, again from Jacob Nielsen. I'm talking so much about this man, but he shares a lot of great content out there and I want to link an interesting article that he wrote where he's really advocating for a mix of prompt design interfaces and GUI, so a mix between both.

you have your interface where you can do all the changes but you have also prompts for very basic or for your starting points for example. And I totally agree, I think that would be the perfect thing. Galileo kind of tries it at the moment. Of course they can't do the iteration, like they can't integrate like a screen or something like that, although they might. But I think for now it's much easier to continue in Figma, like the tool that we are all using anyway.

I'm also pretty interested to hear your point of view, what you think about these tools. Would this be helpful for your own workflow? Do you think it's a little bit scary, something that you don't really want to use? I personally think it's a very important reminder for designers to really stay on track and really understand that things are moving quite fast.

and if today you decide oh this tool or none of these tools are great for me they are not perfect you might need to rethink this tomorrow because they are also improving and things are changing pretty pretty fast so for us it's important to keep one eye on the tools on the things that are out there and and always also iterate our own workflows and the way how we design

And now the quick reminder that you can sign up for the AI mini training. It's for free, so you don't need to pay any money for it. You can find the link to the AI mini training in the description box. Thank you so much for listening and hear you in the future.