top of page
  • Writer's pictureJakob Nielsen

UX Roundup: Ideogram Big Improvements | UX Apprentice | High School UX | Kindergarten UX | Self-Taught UX Designers

Summary: Ideogram 1.0 ships with huge improvements in text rendering and image quality | You don’t need a university degree to do UX | Estonian high school students learn UX | Explain UX to a 5-year-old | Self-taught UX designers can shine by embracing AI’s randomness


UX Roundup for March 4, 2024. (Ideogram’s new release 1.0)

What Are the Top AI Tools Right Now?

I want to build a list of the AI tools that are used the most by UX professionals right now. (If readers like it, I may make another list next year, which will surely be different, considering how many new AI tools are released every week.)

Please help by completing this short survey asking about the top 5 AI tools you use.

The survey should take no more than a minute if all you do is enter the names of your preferred tools, but I (and the UX community) will be grateful if you are willing to spend a little more time to explain what’s good or bad about the tools. Most of the many respondents until now have provided great, insightful commentary on their use of AI for UX.

I will publish the results in a future edition of this newsletter. Completely anonymized, of course.

(I’m reposting last week’s survey to give you one more chance to contribute. I’ll close it at the end of this week. If you have already contributed, Thank you! Please don’t vote a second time.)

What are the top AI tools for UXers right now? Let’s find out! (Midjourney)

Ideogram 1.0: Text Rendering and IQ Much Better

Ideogram released its first “real” product, labeled version 1.0. (The previous releases had numbers like 0.2.) Dramatic improvements. I had started writing them off after Midjourney improved its text rendering, but Ideogram is now back in the game!

Here is a simple example. Prompt: The text "Visibility of System Status" with a big human eye. Also smaller text "Jakob Nielsen's Usability Heuristic #1". For iterations 2 to 4 below, I used the remix feature for my preferred image from each previous each round, rather than rerolling the prompt from scratch or trying alternate prompts. If we needed an image for a real campaign, the designer ought to try these other strategies as well.

Upper left: the very first result is already almost perfect, but has misspellings of both “Sytem” and “Usabilty.” Upper right: second iteration, still with misspellings. Lower left: in this third iteration, I love the AI creativity that replaced the iris of the eye with a bunch of computer icons. This came out of nowhere and must have been guided by some form of understanding of the topic I aimed to illustrate. Iteration 3 also spelled all the words of the main text correctly (still misspelling of “usabilty,” though, and a weird M in “system”). Lower right: the fourth iteration finally got all the words right, though I don’t like the hashtag after the main text. (Ideogram 1.0)

These are honestly very impressive for the current state of AI image generation. In particular, it’s otherwise very hard to get AI to understand the concept of primary and secondary text, which is fundamental to visual design hierarchy. I don’t know how well one can control other aspects of the typography, such as the typeface. If I were generating social media posts for all 10 heuristics, I would probably like the same font in all of them.

Ideogram made images with text superimposed above the big eye for these examples. This is less challenging than images where the typography is integrated with the action. For an example of this type of imagery, see the hero image at the top of this newsletter, where the text is rendered as “foam art” on a cup of cappuccino.

Prompting Ideogram 1.0 with the description of the “Visibility of System Status” heuristic, but not providing any direction for the visualization. This unleashes its native creativity. Ideogram still includes an eye (though I didn’t prompt for eyes this time), which may be the most obvious way to illustrate something that starts with the word “visibility.” There’s still a way to go before we can get good concept artwork from AI. (Ideogram)

Talk about Ideogram: they have a job opening for a design lead in Toronto.

The UX Apprentice

You don’t need a university degree for a UX career. 🚀 Kate Moran wrote a nice essay making this point recently.

And yep, this goes beyond BA/BS and MA/MS degrees, extending to Ph.D. holders. Considering the opportunity cost, especially for those not eyeing a life tethered to theoretical academic research, the opportunity cost of a Ph.D. isn't worth it. For applied research or design, learning on the job trumps classroom education.

(I’m not saying that one does not learn how to conduct research in a Ph.D. program. I’m saying that one would learn more about applied research for improving UX designs by spending the same number of years as the researcher on a design team. Too much of a Ph.D. student’s time is spent mastering the intricacies of writing academic papers, which is a useless skill in a design job.)

Classroom learning vs. on-the-job learning: for UX the latter is better. (Dall-E)

My dream scenario for mastering UX? An apprenticeship program, much like the ones in trades across Germany or my native country, Denmark. 🇩🇪🇩🇰 Here’s a quick peek at the German model, which I believe could be a game-changer for UX education (Note that the program includes 30% classroom learning, which does seem appropriate):

In Germany, apprenticeships range from 2.5 to 3.5 years, tailored to the profession and the apprentice’s background. This setup thoroughly encompasses the profession's essentials, ensuring that graduates are armed with both the practical skills and theoretical knowledge needed. 🛠️📚 If you've got some relevant qualifications already, you can fast-track your journey.

What’s unique? The German system is dual-core: about 70% hands-on work in companies, immersing apprentices in real-world experiences, and 30% in vocational schools to hammer down the theory. This mix guarantees a well-rounded education, prepping apprentices for the workforce with a solid grasp of their field’s foundations. 🏫🏭

Historically, being apprenticed to a master was how to learn a craft. UX has a dash of art and another dash of science, but it’s mainly a craft. Thus, I believe apprenticeships are the best way to learn UX. Learn from senior UX professionals on the job, and avoid being the sole UXer in a company while you’re new. (Midjourney)

UX in High School

An apprenticeship may be the best way to become a great UX professional. But there’s no reason to wait until after high school to learn about UX. Tatjana Domnina is teaching two UX courses at a high school in Estonia: “From idea to a solution,” and “Prototype and testing.”

The tiny country of Estonia, with a population of only 1.3 million people, is one of the most advanced digital societies in the world, sometimes referred to as e-Estonia. Giving their youth an early leg up in UX will further the country’s high-tech ambitions.

Location of Estonia (DallE).

Most readers are unlikely to move to Estonia, even for a better high school education for their kids. But there is no reason high schools in other countries could not launch their own UX courses. Not that most high school graduates will become UX professionals. Even my most optimistic projection for the number of UX professionals in the world taps out at 100 million by 2050, which will be 1% of the world’s population or maybe 2% of the workforce. But the 98% of high school graduates who will end up in other professions will still benefit from knowing the basics of UX because the need to connect with an audience in digital media will be ubiquitous.

Estonian high school students are learning user interface prototyping. (Midjourney)

Kindergarten UX = Explain the Field to a 5-Year-Old

Continuing our journey down the education-system ladder: from Ph.D. to BA to high school to kindergarten. I’m not actually proposing teaching UX to 5-year-olds. But can we explain ourselves to this audience? That’s a good test of

  • Really understanding the core of what we do

  • Being able to articulate this insight in simple terms

Will Grant (Senior Director at Gartner) did this exercise and posted his explanation of UX to kindergartners recently:

“User Experience (UX) is like going to a new restaurant. If you can easily read the menu, order your food, enjoy the meal, and pay your bill without any confusion or problems, that's good UX. If the menu is hard to read, the food takes too long to arrive, or it's difficult to pay, that's bad UX. In the digital world, UX is about making websites or apps as straightforward and enjoyable as possible.”

Do you have a different explanation of UX that would make sense to a 5-year-old? Post in the comments. Extra props for adhering to our user research ethos and testing your explanation with a bunch of 5-year-olds. (Your own kids don’t count; they have probably already figured out roughly what their Mom/Dad does for a living.)

Can you explain what you do to a group of kindergartners? (Midjourney)

Self-Taught UX Designers Can Shine by Embracing AI’s Randomness

I was on The Cutting Edge Show with Ansh Mehra, a leading UX influencer and educator in India. The full 45-minute recording is now available to watch on YouTube. Here are some of the topics we covered:

🔥 UX Tigers: Unleashing Creative Freedom in User Experience

🤖 The AI Wave Is Transforming UX Design for the Future

🚀 How UX Designers Can Embrace AI for Success

💡 Self-Taught Designers in the AI Age: Opportunities and Challenges

I discussed the potential of AI as a transformative tool in the field of design education, especially for self-taught designers. AI is an accessible resource that can significantly lower the barriers to entry in the design world. Even with limited resources, aspiring designers can leverage AI tools to enhance their learning and creativity. This democratization of design tools, brought about by AI, can be a game-changer for those unable to attend formal design schools or afford expensive continuing education courses.

Hands-on learning and experimentation with AI tools are the way to becoming a great UX professional. I encourage self-taught designers to actively use AI in their projects, seeing it as an assistant that can offer a range of ideas and solutions. This approach not only enhances technical skills but also fosters a deeper understanding of design principles. This hands-on experience with AI will be invaluable as these tools become increasingly integrated into professional design workflows.

AI is superb at ideation and creating alternative versions of anything you want to use in a design project. But there is a randomness to its results because you can’t predict what you’ll get from a certain prompt. Sometimes, you even get “hallucinations,” which is a polite word for things that are wrong. Or, in design, you may get suggestions that have poor usability, even if they’re not specifically false. UX designers must excel at managing this randomness, but that’s a skill that was not taught in the past because it was not something we needed to deal with. Sometimes, we get a lot of wacky and wild ideas from AI, and we have to learn how to discern what’s good and bad. Edven more important, UX designers must cultivate the skill of how to sometimes combine two things that AI came up with into a single design. Those are new skills where a self-taught UX specialist can quickly become much better than people with formal degrees that didn’t teach how to manage randomness.

Here is a comic strip I made with Umesh’s Story Illustrator GPT to illustrate the point I made on Ansh Mehra’s show:

Feel free to copy or reuse this infographic, provided you give this URL as the source.


Top Past Articles
bottom of page