At JetStyle we experiment with everything that emerges in the industry of AI and share our insights in the blog.
Just 1.5 years ago the designer community was motivated to learn new ways to generate AI-based images. It’s no longer a secret these days: you’ll find plenty of tutorials for all the most popular neural networks.
Today we’re already researching ways AI influences customer behavior, and how end users perceive AI art. Do they prefer images created by humans or robots? How radical are they in their opinion?
Alex Perminov, JetStyle’s designer in the PR department, has recently researched this matter, and he gave a talk about it at a local IT conference. This article is based on his findings. In the first part, you’ll read about the results of a mini survey Alex conducted. In the second part, he shares a few recommendations about augmenting UX and UI with the help of AI.
Enjoy and share your thoughts about our common future with AI as an integral part of all visual communications.
Global Research
Global companies are collecting data about how people perceive neural art. Ipsos found out that around 50% of respondents are happy about using products created with AI. The same amount of people get nervous when they come across AI-based products and services, and the number has grown since 2021.
In research conducted by the advertising network Dentsu there’s evidence that more than 70% of people would like to know that a service is made with the help of AI.
For Alex, AI has been a valuable creative partner for more than 2 years. But does our audience feel the same about the contribution of neural networks to design?
Alex conducted his own mini-research to get to know that. He aimed to learn if it’s important for the audience to know whether the design is created by neural networks or a human. Also, he wanted to get insights about the labeling: should we put it on the images created by AI for higher transparency? Moreover, Ipsos and Dentsu conducted their research in 2022-2023, so Alex wanted to see if anything changed since that time.
Apart from that, Alex was looking for correlations between the age group and industry the respondents work in. We especially wanted to hear opinions from those outside IT: they are the potential users of the products we create at JetStyle.
The format of the research was a survey: Alex included about 36 visuals he created either himself or with the help of AI, labeled them accordingly, and asked the respondents to choose the visuals they liked more.
Alex intentionally put the labels “made by designer” / ”made by AI” on the visuals in the survey, as he needed to check a hypothesis. At the start of the research he was sure that information influenced the perception of the visuals. His hypothesis was that human empathy was a crucial factor, and people preferred ‘real’ designers to robots.
We processed about 200 responses:
60% of respondents work in IT, design, or technologies:
Among those who work outside IT, the interest in visuals created by humans and AI was almost equal.
Interviewees from the IT industry were more favorable towards AI-generated visuals.
As for labeling, our audience agrees with the findings of global studies: 50% of our respondents want to know if AI assisted in the creation of the visuals.
In the beginning, we expected people outside IT to express more skepticism about AI. However, we fully denied this hypothesis. A few core conclusions Alex made after his research:
- People are highly aware of the presence of AI in the content they consume, and they seem to tolerate it. At least, there is no vivid negativity towards AI-generated content.
- Still, it’s too early to say that the phenomenon of AI is completely known and safe. People are precautious, as there’s too much fake content that is closely associated with AI.
Another reason why AI is perceived as dangerous is that there’s no comprehensive legislation covering all possible issues. However, is it even possible? Most AI tools are based on language models and generative networks, so they are open source; anyone can download the code and re-create it.
New technologies give us new possibilities but they bind us with new responsibilities at the same time. It’s our job to stay ethical while implementing AI and stick to the original goal: creating products that benefit people.
AI for UX & UI Design
Here are a few tips, tricks, and examples of our use of AI for design tasks.
The greatest benefit lies in the way AI cuts costs on everyday design tasks. We elaborated a lot on this topic in one of our articles. Long story short, JetStyle’s designers generally manage to save 2-3 hours on sketching and making concepts that used to take 6+ hours.
AI x UX
Today UX design cannot be fully done by AI: neural networks are capable of creating impressive art, but they cannot perform complex tasks related to user experience. However, there’s still a huge variety of smaller design activities AI may help designers with. A study by NN Group shows most UX professionals regularly use AI at their work.
Neural networks help with:
- Editing text content;
- Conducting research: AI can guide you through the preparation process, as well as analysis of the results of your research;
- Designing images and creating videos;
- Brainstorming content ideas.
While modern text neural networks are very advanced, you cannot expect AI to do all the job for you. When Alex designed JetStyle’s website menu, he addressed ChatGPT for inspiration and extra opinion. He had his own vision about how the menu should be arranged, and ChatGPT gave him a fresh unbiased insight. It turned out to be a productive collaboration:
AI & UI
The situation is just the same: you may view AI only as an assistant. Existing neural networks already create impressive visuals, but the quality of interfaces is still low.
Among all the tools we tried, Galileo AI is definitely at the top. It’s a chatbot that processes your prompt, asks questions to understand the task better, and gives you an interface you can use as a draft. Here’s an example from Alex:
Here’s the interface Galileo suggested:
The interface looks primitive and boring, but it’s what you expect from a neural network. You may import it to Figma with functioning layouts and nesting components, and use it as a basis for your future interface.
To Sum up
The variety of neural tools is almost endless. They are fine, but they won’t solve your task without you. The best tools require some time for adjusting and mastering.
AI is evolving, and as tech enthusiasts, we want to believe one day it will cover most of our mundane tasks.
As for today, the biggest responsibility is still on us: we have to take into account the user experience and look for ways to improve it. AI is only a partner in this way. A disclaimer if it’s important for our readers: Alex’s speech was also created with the help of AI. ChatGPT found relevant studies and helped Alex analyze the data from his research. You’re welcome to ask any questions about collaboration with neural networks; be sure we’re going to answer them personally :)
