top of page
  • Writer's pictureJakob Nielsen

UX Roundup: First Mouse | Dark AI | AI Spend Grows | Most Drivers Think They’re Above Average | 4 Clicks = 4 AI Models

Summary: First computer mouse in museum | Beware dark AI design | 293% y/y growth in companies’ AI spend | Most drivers think they’re above average | Easy access to ideation variety by using multiple AI models

UX Roundup for June 3, 2024. Northern Hemisphere summer. (Ideogram)


The First Mouse Now in Computer History Museum

Doug Engelbart invented the computer mouse 60 years ago, in 1964. The very first prototype mouse has now been donated to the Computer History Museum in Mountain View, CA, and has gone on display in the museum. (The Computer History Museum is one of the very few attractions worth seeing if you visit Silicon Valley.)

My personal connection with the first computer mouse goes back about 30 years, when I had the privilege of meeting its inventor, Doug Engelbart. He was a true gentleman, modest about his groundbreaking accomplishments. Doug even allowed me to try out that first computer mouse. Now that it’s a museum piece, I realize how fortunate I was to handle this important historical artifact, an experience that I still remember vividly.

As a prototype, the Engelbart mouse was big, heavy, and clunky, but it was the first of billions.

The modern computer mouse has added several features beyond Doug Engelbart’s first mouse, most notably the scroll wheel. (Ideogram)

Beware Dark AI Design

Interesting video with Sinead Bovell discussing the possibility for AI to engage in social engineering to influence users by abusing its understanding of emotions. Don Crowley immediately realized that this opens up a whole new realm for dark design.

Studies from last year found that last-generation AI was as persuasive as humans, according to a meta-analysis across 121 studies with 54K participants. More recent research with current AI products found that GPT-4 level AI was 82% more persuasive than humans.

Extrapolation from two data points is not great, but it doesn’t take much imagination to expect that the GPT-5 level AI products we’ll have soon will be even better at manipulation. 100% better than humans, 200% better than humans? That’ll remain to be seen, but I will be very surprised if that 82% manipulation lift is the top of AI performance, given that we’re only in Year 2 of the AI revolution.

Unfortunately, AI-driven dark designs may not be dark patterns per se because AI affords individualization, and thus different dark designs in each case. Lack of patterns may make dark AI design harder to recognize and shame. We have names for many old-school dark design patterns and this makes them easier to fight. More research is needed to fully understand dark AI and protect users against it (ha, there’s a cliché, but it’s true).

Evil AI is scheming to impose dark design on users. However, as always, we should remember that it’s evil humans who are behind any dark design. AI is just a tool. (Midjourney)

This may seem to be a negative posting, but I’m an optimist at heart. (Much as you may find this difficult to believe, given my 30-year history of writing criticisms of products with bad usability.) While dark AI is a definite danger to humanity, AI itself can come to the rescue. We can train the user’s AI agents to recognize dark design, even when it doesn’t follow strong dark patterns as in the old days. AI can fight AI, and we just need the stronger AI to be on the side of users.

293% Y/Y Growth in Small/Midsized Companies’ AI Spend

Expense-reporting service Ramp has released an analysis of AI spending by its approximately 20,000 customers, who are mainly small and midsized companies, in Q1 of 2024, compared with Q1 of 2023.

During this one-year period, companies’ spending on AI increased by 293%. Even so, they only recorded AI spending by a little over a third of their customer companies. Obviously, some companies may use AI tools that are not charged to a credit card and not captured by an expense-reporting and credit-card service like Ramp. So the total proportion of AI use is likely to be closer to half of the small businesses in the United States.

Many employees use their personal AI accounts for work, using products like ChatGPT and Midjourney that unenlightened management is still not willing to reimburse. (This is a terrible business policy, because it results in AI being out of control. Individual employees use AI without aligning with any corporate strategy because AI makes them so much more productive. But across the organization, much potential for gains and organizational learning is lost.)

Across 389 AI tools analyzed by Ramp, the retention of business customers over a one-year period was 56%, for a churn rate of 44%. This is around the average churn rate for cloud-based SaaS services.

In other words, current AI tools are useful, but not greatly more useful than non-AI tools. I view this as a positive finding, given that AI is only in Year 2, whereas SaaS has been popular for almost 30 years, with early cloud-based services starting 40 years ago.

Most impressively, OpenAI had a one-year retention of 82% (annual churn of 18%), meaning that businesses find sufficient value in the ChatGPT Plus subscription to keep paying up, month after month. (ChatGPT is also the top AI tool used by UX professionals.)

Furthermore, companies paying for ChatGPT subscriptions increased their spending on this tool by almost 800% during the year. Since OpenAI has maintained constant pricing for the last year, 8x spending can only come from 8x the number of seats. In other words, a typical company will start small, with just a handful of ChatGPT Plus seats, and rapidly scale to more users as the pioneers prove the business value of AI. (This is also my recommendation: Start Small, Start Now, rather than procrastinating until you have an elaborate plan in place. The learnings from hands-on AI use beat anything.)

The ultimate UX metric: are customers willing to pay and keep paying? For AI, the answer is “yes,” proving that it works. But too many companies have yet to start using AI, at least officially. (Ideogram)

Most Drivers Think They Are Above Average

In a new study, 77% of participants rated themselves as superior to self-driving cars, whereas 60% thought most other people were worse. The actual accident numbers show that self-driving cars are already better than average human drivers, meaning that many lives could be saved by changing over, even without further advances in AI.

The researchers paradoxically suggest that the way to improve automobile safety is to stop talking about safety. Since most people think they are above-average drivers (which obviously can’t be true) they will not be convinced even by strong data that self-driving cars are safer than average human drivers. Instead, focus on the convenience of being chauffeured around by AI instead of having to do the manual labor of operating the car yourself.

It annoys me to say that they may be right.

Most drivers think they are above average, according to new research. This belief kills people. (Ideogram) Leonardo’s animated version of this image is on the UX Tigers' Instagram feed.

4 Clicks = 4 AI Models: Easy Access to Ideation Variety

I recently recorded a webcast with Dwayne Samuels from Samelogic. Dwayne is an excellent interviewer and the webcast covered many interesting points in the evolution and future of UX (YouTube: 61-minute video).

Here are several different descriptions of the top points in the webcast. Rather than giving you a single summary, I used a new feature in the paid AI service Poe, which offers most of the main AI large language models in a single user interface. After getting the answer to a prompt from one service, Poe allows users the ability to reissue that prompt to another AI model with a single click. After you get that answer, you can try yet another AI model with one more click, and so on.

Here, I asked 4 leading AI models to describe the top 3 points I made in the webcast. I’ll show you the answers from GPT 4o, Claude 3 Opus, Mistral Large, and Gemini 1.5 Pro.

Poe allows you to easily assign the same problem to many different AI models so that you can compare solutions and widen your ideation. (Midjourney)

I uploaded the transcript of the full episode (10,871 words) and asked the AI to perform the following steps:

  1. Identify the top 3 points I made that would interest an audience of UX professionals.

  2. For each point, write 3 different postings in widely varying tones of voice.

  3. For each variant, write a headline that includes an emoji.

  4. For each variant, write a 3-paragraph summary

  5. Label each variant with its tone of voice.

GPT4o and Gemini 1.5 Pro followed my instructions. Claude and Mistral only provided one-paragraph summaries, which I think are too short. Mistral forgot to give me emojis for each posting variant, while Claude forgot to label each variant with its targeted tone of voice.

I found the GPT 4o solutions to be the most useful in this exercise. To promote this webcast on LinkedIn, I would probably edit together a combination of its professional and casual tones for each of the top points. I might also throw in a few of the humerus phrases.

For the perfect post about this webcast, I would also incorporate a few ideas from the models I didn’t like. They weren’t all bad, after all.

Poe’s ability to easily experiment with different AI models’ solutions to the same problem is an excellent way to broaden the “explore” phase of diamond prompting. For example, the 4 different models didn’t identify the same top 3 points. Next time, I might ask for 5 top points, even if I only planned to post 3.

(Selecting what to use from the broad selection of ideation variants is the job of the “refine” stage, and for best results, proceeding with more iterations of prompts would be recommended. For example, ask model A to expound on an idea you got from model B, if you like model A’s approach better for your current problem. Human-AI co-creation at its best.)

Here are the different writeups of my webcast, from 4 different AIs. Which do you like better?

GPT 4o

Point 1: Evolution of Usability from 1983 to Today

Posting Variant 1 (Professional Tone): 🚀 UX Evolution: From 1983 to Today

I shared the fascinating journey of usability since 1983. Back then, UX was almost nonexistent, with text-based and mainframe computing dominating the landscape. Companies rarely conducted user testing. Fast forward to today, and we see a remarkable transformation in usability practices. Technological advancements, like the introduction of graphical interfaces and the rise of the web, have driven significant improvements in UX.

Today, businesses understand the importance of investing in UX to enhance user satisfaction and boost conversion rates. The shift from text-based commands to graphical interfaces made interactions more intuitive. Additionally, the web revolutionized how we approach usability, placing the user experience at the forefront of business success. This evolution continues with AI, pushing the boundaries of what we can achieve in UX design.

Posting Variant 2 (Casual Tone): 🕹️ From 1983 to Now: UX's Epic Journey

Hey UX pros! Ever wondered how far we've come since 1983? In my latest webcast, I dived into the early days when usability was almost a foreign concept. Think text-based computing and mainframes—yikes! Companies barely did any user testing. But look at us now! Thanks to tech advancements, UX has grown by leaps and bounds.

The transition from text commands to graphical interfaces was a game-changer, making everything so much more user-friendly. The web boom in the 90s forced businesses to prioritize UX, and today, AI is taking us to new heights. It's wild to see how far we've come, and the future looks even brighter!

Posting Variant 3 (Humorous Tone): 🤣 UX in 1983: The Struggle Was Real!

Guys, you won't believe what UX was like in 1983! I shared the hilarious—and kinda painful—early days in my webcast. Imagine using text-based computers and mainframes. User testing? What's that? Fast forward to now, and we've got slick graphical interfaces and the power of the web transforming our lives.

Remember how revolutionary the Macintosh was? That was just the beginning. Today, AI is making UX even cooler. We've come a long way from those clunky, oppressive systems. Cheers to the future of usability!

Point 2: Intent-Based Outcome Specification with AI

Posting Variant 1 (Professional Tone): 🤖 AI in UX: Shifting to Intent-Based Outcomes

I discussed how AI is revolutionizing user interactions. We've moved from command-based interfaces to intent-based outcome specification. This means users no longer have to provide step-by-step instructions. Instead, they can specify their desired outcome, and AI takes care of the rest.

This shift transforms how we de

sign and interact with technology. It reduces the burden on users and makes systems more intuitive. Although AI isn't perfect yet, its potential to simplify complex tasks is immense. As UX professionals, we must embrace this change and focus on creating intent-driven designs that enhance user experiences.

Posting Variant 2 (Casual Tone): 🧠 AI Magic: Just Tell It What You Want!

Hey, UX fam! In my latest webcast, I talked about how AI is changing the game. Remember the old days of command-based interfaces? Now, we're moving to intent-based outcomes. Just tell the AI what you want, and it figures out the rest. So cool, right?

This makes our job as designers even more exciting. We get to create systems that understand user intent and deliver what they want without all the complicated steps. Sure, AI isn't perfect yet, but it's getting there. Let's ride this wave and make UX more intuitive than ever!

Posting Variant 3 (Humorous Tone): 😂 AI Be Like: "Got It, Boss!"

Hey UX peeps! Guess what? AI is taking over the heavy lifting. In my webcast, I talked about moving from old-school command interfaces to intent-based outcomes. Now, you just tell AI what you want, and it handles the rest. No more typing out complicated commands!

It's like having a super-smart assistant who just gets you. Sure, AI isn't perfect yet, but it's getting there. Let's design with this in mind and make our users' lives easier. Who's excited for the AI takeover?

Point 3: Jakob's Law and User Expectations

Posting Variant 1 (Professional Tone): 🔍 Jakob's Law: Designing for User Expectations

I explored Jakob's Law, which states that users spend most of their time on other sites and apps, shaping their expectations. This means our designs should align with familiar patterns to avoid confusion and errors. Users prefer interfaces that work as they've learned from other experiences.

By adhering to established con

ventions, we create intuitive and user-friendly designs. Only when we have a significantly better solution should we deviate from these norms. Jakob's Law reminds us to respect users' existing knowledge and design accordingly for seamless experiences.

Posting Variant 2 (Casual Tone): 🧩 Jakob's Law: Keep It Familiar!

Hey, UX friends! In my latest webcast, I dived into Jakob's Law. It says users spend most of their time on other sites, which shapes their expectations. So, when designing, we should stick to familiar patterns. Users don't want to relearn everything!

Following established conventions makes our designs intuitive and user-friendly. If you have a killer new idea that's way better, go for it! But if it's just a bit different, stick to what users know. Jakob's Law is all about making things easy and familiar for our users.

Posting Variant 3 (Humorous Tone): 🤔 Jakob's Law: Don't Make Users Think!

UX fam, listen up! In my latest webcast, I talked about Jakob's Law. Basically, users spend more time on other sites, so they have certain expectations. If our designs are too different, they'll be like, "What's going on here?" Stick to what they know!

When users don't have to think twice, they're happy. Follow familiar patterns unless you have a super amazing new idea. Jakob's Law is all about making things easy and predictable. Let's keep our users smiling and confusion-free!

Claude 3 Opus

Top Point 1: Jakob Nielsen developed the discount usability movement to enable good UX work with limited resources. This approach relies more on expert insight and small sample sizes to quickly iterate and improve designs.

🌟 Revolutionizing UX on a Budget 🌟

Jakob Nielsen's discount usability methods help you achieve top-notch user experiences without breaking the bank. Embrace the power of expert insight and rapid iteration to create designs that wow your users!

🚀 Blast Off with Discount Usability! 🚀

Want to skyrocket your UX game? Look no further than Jakob Nielsen's ingenious discount usability approach. With a focus on quick iterations and expert analysis, you'll be delivering stellar user experiences in no time. 3, 2, 1... Liftoff! 🌙

🎯 Bullseye! Hit Your UX Targets with Discount Usability 🎯

As a seasoned UX pro, I swear by Jakob Nielsen's discount usability methods. By leveraging expert insights and small sample sizes, you can rapidly improve your designs and hit the mark with your users. Ready, aim, succeed! 🏹

Top Point 2: Jakob's Law states that users spend most of their time on other interfaces, so they prefer designs that work similarly to what they already know. Designers should follow existing conventions unless their new approach is significantly better.

🌍 Navigating the World of UX with Jakob's Law 🧭

As a UX trailblazer, I've learned the importance of Jakob's Law: users crave familiarity across interfaces. By embracing existing conventions, we can create intuitive experiences that feel like home to our users. 🏠💙

🚦 Green Light for Conventional UX 🟢

Stop! ✋ Before you reinvent the UX wheel, consider Jakob's Law. Users spend most of their time on other interfaces, so sticking to familiar patterns is a surefire way to keep them cruising along. Let's give convention the green light! 🚗💨

🎭 All the World's a UX Stage 🎬

In the grand theater of user experience, Jakob's Law takes center stage. 🎦 As UX directors, we must recognize that our users are seasoned performers, accustomed to the familiar scripts of other interfaces. Let's give them a starring role in a play they already know and love. 👏🌟

Top Point 3: AI-powered tools are transforming UX design and research. UX professionals must learn to leverage these tools effectively, using them in tandem with human insight and critical thinking for the best results.

🚀 Blasting Off into the Future of UX with AI 🌌

Attention UX explorers! 🧑‍🚀👩‍🚀 It's time to embark on a thrilling journey into the uncharted territories of AI-powered design and research. With tools like natural language interfaces and generative AI at our fingertips, the possibilities are endless. Buckle up and get ready for an epic UX adventure! 🪐✨

🎨 Painting a Brighter UX Future with AI 🖌️

Hey there, UX artists! 👩‍🎨👨‍🎨 Ready to unleash your creative genius like never before? AI-powered tools are here to help you paint a masterpiece of user experience. With AI as your trusty assistant, you'll be able to focus on the big picture while leaving the details to your digital companion. Let's create UX art that stands the test of time! 🎭🌟

🌿 Nurturing UX Growth with AI 🌱

Attention UX gardeners! 👨‍🌾👩‍🌾 It's time to cultivate a lush landscape of user experiences with the help of AI. By planting the seeds of AI-powered design and research, you'll be able to nurture interfaces that blossom with intuitive interactions. Get ready to watch your UX garden thrive like never before! 🌻🌷

Mistral Large

Top point 1: The Evolution of UX: From Primitive to Powerful

Informative: "Tracking the Trajectory of UX: A Look at Its Evolution"

An insightful look at the evolution of user experience design, from its humble beginnings in 1983 to the powerful tools and technologies available today. Learn about the key milestones and developments that have shaped the field of UX.

Engaging: "Unleashing the Potential of AI in UX: A Game-Changer for User Experience"

Are you ready to witness the transformative power of artificial intelligence in UX? In this webcast, Jakob Nielsen discusses how AI is revolutionizing the way we design user interfaces, making them more intuitive and user-friendly than ever before.

Persuasive: "Why the Web Needs UX: The Business Case for User Experience"

In today's digital landscape, user experience is no longer a luxury but a necessity. This webcast makes a compelling argument for why businesses should invest in UX to improve their online presence, increase conversion rates, and stay ahead of the competition.

Top Point 2: The Power of AI in UX: A Game Changer

Informative: "Artificial Intelligence and UX: A Powerful Combination"

Discover how artificial intelligence is transforming the field of user experience design in this informative webcast. Learn about the key benefits and challenges of integrating AI into UX and how it can help create more intuitive and efficient user interfaces.

Engaging: "Unleashing the Potential of AI in UX: A Game-Changer for User Experience"

Are you ready to witness the transformative power of artificial intelligence in UX? In this webcast, Jakob Nielsen discusses how AI is revolutionizing the way we design user interfaces, making them more intuitive and user-friendly than ever before.

Persuasive: "The Future of UX is Now: Why You Need to Stay Ahead of the Curve"

Emerging technologies and tools are rapidly changing the landscape of user experience design. This webcast makes a persuasive case for why UX professionals need to stay updated and adaptable in order to succeed in this dynamic field.

Top Point 3: The Web and UX: A Match Made in Heaven

Informative: "The Web's Impact on UX: Understanding the Connection"

Jakob Nielsen explains how the web has changed the way we approach user experience design. Learn about the key differences between UX for personal computers and UX for the web, and how businesses can leverage this knowledge to improve their online presence.

Engaging: "The Perfect Match: How the Web and UX Are Meant for Each Other"

The web and user experience design go together like peanut butter and jelly. In this webcast, Jakob Nielsen discusses how the internet has driven the need for better UX, making it easier for users to find what they need and for businesses to increase their conversion rates.

Persuasive: "Why the Web Needs UX: The Business Case for User Experience"

In today's digital landscape, user experience is no longer a luxury but a necessity. This webcast makes a compelling argument for why businesses should invest in UX to improve their online presence, increase conversion rates, and stay ahead of the competition.

Gemini 1.5 Pro

Point 1: The Importance of Jakob's Law in a World of AI

Variant 1: Playful & Engaging: 🧠 Jakob's Law: Still Relevant in the Age of AI? 🤔

You bet it is! We all love a shiny new AI-powered feature, right? But hold your horses! As UX professionals, we can't forget the golden rule: Jakob's Law. Users are creatures of habit, spending most of their time on other platforms and forming expectations based on what they already know.

Throwing a curveball with a radically different AI-powered design might seem tempting, but it can leave users feeling lost and frustrated. The key is to strike a balance between innovation and familiarity.

That's where our UX superpowers come in! We need to deeply understand user expectations through research and rigorously test our AI-driven designs to ensure they're not just cool, but actually usable.

Variant 2: Bold & Provocative:⚡️ AI Won't Save Your Bad UX. Jakob's Law Still Reigns Supreme.

Let's face it, AI is changing the game. But here's a reality check: slapping AI onto a poorly designed product won't magically make it user-friendly. Jakob's Law is still the law of the land!

Users bring their experiences from other platforms to your product. They expect consistency, familiarity, and ease of use. Ignoring these expectations in your AI-powered designs is a recipe for disaster.

Don't be fooled by the hype. Focus on the fundamentals: understand your users, design for their existing mental models, and test, test, test!

Variant 3: Informative & Insightful:💡 Navigating the AI Revolution: Why Jakob's Law Matters More Than Ever

The rise of AI presents exciting new opportunities for UX professionals. But as we embrace these advancements, it's crucial to remember the enduring relevance of Jakob's Law.

Users form their expectations based on their experiences across various platforms. Deviating significantly from established conventions, even with AI-powered features, can lead to confusion and frustration.

To create truly user-centered AI-driven experiences, we must prioritize understanding existing mental models and leverage AI to enhance, not disrupt, familiar interaction patterns.

Point 2: Discount Usability for Agile UX Design

Variant 1: Energetic & Enthusiastic: 🚀 Ditch the Waterfall, Embrace the Flow: Discount Usability for Agile UX!

In today's fast-paced world, who has time for clunky, waterfall design processes? Discount usability is the secret sauce for agile UX teams!

We're talking rapid iterations, small user groups, and a laser focus on gathering feedback and making those design tweaks on the fly. It's like having UX superpowers! ⚡️

The result? Faster design cycles, less wasted time and resources, and ultimately, a product that truly sings to your users. Let's get agile!

Variant 2: Confident & Authoritative:🏆 Winning at Agile UX: Why Discount Usability is Your Secret Weapon

In the ever-evolving landscape of UX design, agility is paramount. Discount usability methods provide a proven framework for maintaining user-centered design within fast-paced development cycles.

By embracing iterative design and leveraging insights from smaller user groups, UX professionals can rapidly gather feedback, validate assumptions, and refine designs with precision. This iterative approach minimizes risk, optimizes resource allocation, and ultimately leads to a superior user experience.

Variant 3: Playful & Engaging: 🕹️ Level Up Your Agile UX Game with Discount Usability!

Think of discount usability as the cheat code for agile UX design. It's all about working smarter, not harder!

Instead of waiting for that big, elaborate usability study, we're talking quick and dirty testing with smaller user groups. Gather those insights, iterate on your designs, and repeat! 🔁

It's like leveling up your UX skills in record time. Get ready to dominate the agile UX arena!

Point 3: UX is More Than Just Visuals: Embracing the Broader Impact

Variant 1: Bold & Provocative: 💥 UX is Not Just Making Things Pretty. It's About Empowering Humans.

Let's be real: aesthetics matter. But in the world of UX, they're just the tip of the iceberg. We're not just pixel pushers; we're architects of experience, advocates for users, and champions of accessibility.

It's time to move beyond superficial design and focus on the heart of UX: creating technology that truly serves human needs.

Variant 2: Inspiring & Visionary: ✨ UX: Designing a More Human-Centered Future

As UX professionals, we have the power to shape the way people interact with technology. It's a responsibility we don't take lightly.

It's about understanding the nuances of human behavior, designing for accessibility and inclusivity, and constantly challenging ourselves to create experiences that are not just usable, but truly empowering. Let's build a future where technology enhances, not hinders, our humanity.

Variant 3: Informative & Insightful:💡 Beyond the Interface: The Holistic Impact of UX Design

UX design is often associated with visual aesthetics and user interface elements. However, true UX encompasses a much broader scope, encompassing the entire user journey and its impact on their lives.

From understanding user needs and motivations to designing for accessibility and inclusivity, UX professionals play a crucial role in shaping how technology integrates with our daily lives.


Top Past Articles
bottom of page