GenAI Tools, I Love You — From a UX Designer
In this article, I want to talk a little bit about my personal journey and experience with AI-driven tools and how I have incorporated them into my design process. This is one of my favorite topics to talk about and I hope you get something insightful out of reading this article.
Contents
- Navigating the intersection of design and AI: Reflections on Midjourney and ChatGPT
- Exploration, exploration, exploration
- Learning from others
- My favorite AI tools and how I integrate them in my workflow
- How I plan to keep experimenting and learning
Navigating the intersection of design and AI: Reflections on Midjourney and ChatGPT
Last year and the year before were very interesting years for designers.
During the introduction of Midjourney and ChatGPT in 2022, a lot of speculation took place about wether or not us, designers, were going to lose our jobs and be replaced by AI or (some are still arguing this) by AGI in the future.
For me, the truth is that there’s no way we can 100% accurately predict what’s going to happen with AI and design in the long run. What is certain, is that there’s a clear advantage, for those willing to see it like that, over how we can use these tools to make our jobs easier and improve the efficiency of our processes.
Exploration, exploration, exploration
Since 2022 many more AI-driven tools have come out. It definitely can be hard to try to keep track with all of them and try to get over that learning curve fast (if they’re still on beta versions and are more complex or just have a bad UX this can make it worse). I’ve been there.
After exploring a couple of them, I’ve found that most of their outputs aren’t good enough to include in a real project.
Ultimately, I’ve stuck with the ones that I have felt more comfortable using and that provide outputs that work for me and my workflow. I would try to encourage other designers to explore a little bit these other non-mainstream AI creative tools and see which ones work for them.
Learning from others
This week I watched the recording of a webinar session that the UX Design Institute of Dublin recently had with Nick Babich, a Principal UX designer working at Brain Technologies creating Natural AI, an app that calls itself the world’s first generative interface, were “you no longer go to Apps, Apps come to you. Simply say what you need and the right app forms itself around your words”.
The talk was very inspiring, but what called my attention was the way in which Nick has been able to integrate ChatGPT and Midjourney into his design process; it was not so different from mine, we actually use them in the same way.
This was vey comforting to me because right now, we are at a point (I think) were there isn’t a “formal” or stablished way for designers to use these tools and I felt a little lost at times because of this (at the beginning I like to stick to stablished standards when working with new technology.)
We are learning how they work and how to use them as we go. I have to remind myself that it’s OK, and that it makes sense because we are still at the early stages of the so called AI revolution, and that’s the way these things usually go.
Here is the recording of Nick Babich’s webinar, I highly recommend you watch it.
My favorite AI tools and how I integrate them in my workflow
The ones that have worked (so far) for me, are four:
- ChatGPT
- Midjourney
- Galileo.AI
- Runway.ML
- (Excited about Sora by OpenAI)
I want to dive a little deeper into the details of how I use each one of them. I have to mention as well that I keep reading about new tools, that I’m still learning and I make a ton of mistakes along the way, so in no way am I saying that this is THE best way to use these tools as a designer.
I feel like my approach has been more about allowing myself to explore, have fun and experiment with them and keep making mistakes, as long as I can keep improving upon them.
ChatGPT
This LLM has been my go-to when it comes to getting good results fast for my design projects. I use it mostly to enhance my creativity and improve my efficiency.
The trickiest part might be the prompt creation but there are a lot of resources out there were you can find it easier to learn how to prompt engineer and get the best results out of ChatGPT. I save all the prompts that give me good results. Some of my ChatGPT use cases are:
- Get a general understanding of certain target users before I start with UX research.
- Organize and categorize large amounts of data.
- Get design feedback.
- Brainstorm and get creative by exploring different design solutions.
- Get educated about topics that I don’t understand or that I don’t know about yet, and many more.
Midjourney
This image generation tool is very popular as well. What I love about it is that you can pretty much create any type of image you want. I like to have fun and imagine sci-fi futuristic scenarios and illustrations. The possibilities are endless.
/Imagine an image of a small spacecraft sitting at the bottom of the ocean
in Atlantis, in the style of digital fantasy landscapes, alena aenami,
frank mccarthy, 4k, heavy lines, violet and blue,
jakub różalski --ar 16:9 --s 150 --c 10 --v 6.0Midjourney is mostly an inspiration and experimentation tool for me. Some of the use cases:
- Imagine UI interfaces
- Play around with visual design
- Generate product images
- Generate illustrations
- Imagine movie stills
- Photography images and more.
Also, X has the best threads on how to fine-tune Midjourney, it’s one of my favorite sources to get prompt ideas from.
Galileo.AI
I was so excited about this tool when I first read about it on X back in 2023; if I remember correctly, the CTO shared a tweet about how (very soon) they were going to launch an AI tool that would allow designers to generate complete Hi-Fi designs just by using a prompt.
I waited for a couple of months after registering to become a beta tester, but forgot about it after six months of not receiving any updates about it on my email.
Fast-forward to February 2024, decided to check their website and they were live! So I signed up to try it. I have to say that the interface is very friendly and resembles that of ChatGPT, so it’s very easy to use.
The designs that it generated were pretty generic, I imagine most of these tools will keep getting better over time but for now I only use it to get inspiration and sometimes to generate a Mid-Fi mock ups of screens really fast.
A really nice feature is that you can copy the generated designs directly to Figma. Makes the editing process super easy.
Runway.ML
I use Runway.ML as a complementary tool to Midjourney. I think it’s the best application out there to animate images (until Sora arrived?), and I like to use to bring life and movement to my crazy Midjourney creations.
I have only used the free version which allows me to generate five-second videos, but if you upgrade your subscription you can get up to sixteen seconds. They have a film festival for Runway animated movies, which I think is incredible.
I plan on generating and animating my own short-film soon.
How I plan to keep experimenting and learning
I’m very excited about what’s to come for Generative AI and design, I’m sure it’s going to keep getting better. I think us designers will be able to have a lot of fun with the things that these tools will allow us to do and create.
I want to keep exploring and experimenting. I also want to keep getting better at using the tools that have worked for me.
By researching and paying attention to what my favorite GenAI experts and enthusiasts are saying, I think I’ll manage to adopt the right products into my workflow.
