All Articles


Cover Image

The future is today — How can we leverage AI to improve our UX Design work

March 21, 2023

The centuries-old fear of machines taking over

Several article titles like Will AI (Artificial Intelligence) replace UX designers? and Can artificial intelligence take over the work of UX/UI Design?

When running a quick Google search about UX and AI, a number of articles pop-up and many seem to converge in the same concern: the fear of AI, this dreadful, job-stealing monster, somehow soullessly ripping UX designers of their desks and taking over their work.

A GIF with a news presenter next to a screen where it says “Are computers taking over?”. When the screen says “Yes” she goes hysterical.

Source: Giphy

This is not something new. The fear of machines replacing workers and leave them jobless dates back to the Industrial Revolution, more than two centuries ago. Upon Thomas Edison’s invention, cities started switching to electric streetlights and lamplighters saw their 500-year-old job become irrelevant. In New York they went on strike and in Belgium, fearing they would lose their jobs, lamplighters smashed electric lamps. However, this was not exclusive to lamplighters. In Britain, many skilled workers sabotaged the new machinery as a way to defend their jobs. A movement that would then be called “machine-breaking” and culminate in the Luddite rebellion where factories were burned and attacked. Curiously, “Luddite” is now a term used to describe people who dislike new technology.

“At present machinery competes against man. Under proper conditions machinery will serve man. There is no doubt at all that this is the future of machinery, and just as trees grow while the country gentleman is asleep, so while Humanity will be amusing itself, or enjoying cultivated leisure — which, and not labour, is the aim of man — or making beautiful things, or reading beautiful things, or simply contemplating the world with admiration and delight, machinery will be doing all the necessary and unpleas­ant work.”

The Soul of Man under Socialism, Oscar Wilde (1891)

Today, we’re living a new technological revolution driven by artificial intelligence and, once again, the same old fears have return. I guess we also have to thank movies like The Terminator or The Matrix where machines take over the world for that.

AI has now become a big part of several areas of our lives, and UX Design is no exception. It’s actually becoming more and more applicable to the UX design process.

However, as Oscar Wilde put it back then, rather than focusing on machines that “compete against man” we prefer to focus on machines that “serve man”. So for the sake of this article, we chose to think about how we can take advantage of AI to help improve our daily UX design work and to help us build better experiences for our users.

AI in a Nutshell

Artificial intelligence refers to computer systems and machines able to imitate intelligent human behaviour. These systems can do so because they are exposed to large amounts of training data and find patterns to make predictions. To sum it up, they can perceive, learn, predict and analyse data.

AI can be classified into 3 subcategories, based on the human characteristics it can imitate:

  • Weak or Narrow AI — A learning algorithm is designed to perform a single, specific task without human assistance. Any knowledge acquired from performing that task will not automatically be applied to other tasks. Today, all AI can be categorised as narrow.
  • Strong or General AI — A machine capable of understanding the world as well as any human, with the same capacity to learn how to carry a huge range of tasks. This type of AI is still on its early stages.
  • Super AI — Machines that can outperform humans in any task. Super AI systems are able to understand human feelings and experiences as well as show emotions, beliefs, and desires themselves. This type of AI is still hypothetical.

3 stages of AI. Narrow AI: dedicated to assist with or take over specific taks. General AI: Takes knowledge from one domain, transfers to other domain. Super AI: Machines that are an order of magnitude smarter than humans. Credit: Chris Noessel

When talking about AI we often hear two other buzzwords: machine learning and deep learning. AI uses algorithms and techniques such as machine learning and deep learning lo learn, evolve and get progressively better at assigned tasks.

  • Machine learning — Algorithms that analyse data, learn from it and then apply what they’ve learned to make informed decisions.
  • Deep learning — It’s a subset of machine learning. AI systems use deep learning to continually learn and evolve, making accurate decisions without the help from humans.
Every day examples of AI-powered systems

Over the past few months there’s been a huge buzz around AI, mainly thanks to ChatGPT and AI-generated art. However, AI-powered systems have been around for some while now, across most industries — transportation, healthcare, manufacturing, construction, and retail. Here are some examples:

  • Recommendation engines: You see this when you’re shopping at Amazon (recommendations inspired by your browsing history), choosing what to see on Netflix (recommendations based on your previous viewing data), or on your Spotify’s tailored playlist. Algorithms learn from user data to suggest relevant content.

Example of Amazon showing products recommended based on your latest shopping trends

  • Chatbots and virtual assistants: These tools simulate human conversation through auditory or textual methods. Chatbots have been widely used in retail for customer service. Popular virtual assistants include Google Assistant, Apple’s Siri, Microsoft’s Cortana and Amazon’s Alexa.

A meme with a furious woman, screaming and pointing her finger that says “Customers trying to get a chatbot to understand they want to speak to an agent”.

Human interaction with chatbots is still far from perfect

  • Self-driving vehicles: A vehicle capable of sensing the environment and operating without human involvement. As an example we have some Tesla models, Google’s Waymo, autonomous drones, boats and factory robots.

  • Computer vision: Leveraging AI, computers can analyse images and videos to detect and recognise objects and faces. Facebook recognising who’s in a photo, Google translating text with the camera of our smartphone or Faceapp, are a few examples.
Google Lens translating a sign saying Exit

Google Lens translates text in real-time from over 100 languages.

  • Generative AI: These tools can create new content, be it art, images, text or audio, based on a text input. Recently, ChatGPT and DALL-E 2 have become popular examples of generative AI.

OpenAI tweet with an image generated based on the input “A see otter in the style of ‘Girl with a Pearl Earring’ by Johannes Vermeer.”

OpenAI shares DALL-E text-to-image functionality

How can AI help creating better user experiences?

In the field of User Experience, AI is gradually making its way as a relevant auxiliary tool used during the research, design and development phases.

Collect and analyse large volumes of data

A GIF of a cat with sunglasses with charts moving around that says “Talk data to me”.

Source: Giphy

Gathering information about users and their behaviour, and interpreting insights within the data is a big part of every designer, researcher, marketer or product manager work. Drawing insights from all the data collected can be pretty challenging, not to mention time-consuming.

Although humans are pretty good at recognising patterns, if we talk about big sets of data, we’ll likely lose focus and make mistakes. AI, on the other hand, can be used to analyse hundreds, thousands or millions of data, discovering patterns and insights quicker than human researchers.

Google Analytics Intelligence is a machine learning tool that provides data reports with detailed insights. You can ask questions about your data in plain English and get fast answers. For example: “Which countries are most new users from?” or “Why did my traffic drop?”.

Google Analytics Intelligence  interface showing the answer “New users on mobile — 3,119” after the question “How many new users did we get last week on mobile?”.

Source: Best Practices: Asking questions in Google Analytics

PaveAI uses machine learning and data science combined with Google Analytics to produce data-driven reports and provide actionable recommendations.

AI’s ability to analyse large quantities of data to reveal patterns, gives designers meaningful UX research results, saving them some precious time.

More precise personalisation

A GIF with a woman saying “Please don’t put us in the same category”.

Source: Giphy

Taking advantage of user behaviour data and machine learning it’s possible to provide users with content tailored to their specific needs and interests. Netflix is a pretty good example. Their product’s interface adapts according to who’s accessing it, showing the most relevant information to who’s viewing. Personalisation is also about showing relevant content based on users’ location or time of day, for example.

We’re living in a time where users preferences and behaviours are constantly changing. It’s no longer enough to tailor content to a group of users with the same characteristics (demographic, age, background, etc.) According to the Next in Personalization 2021 Report from McKinsey, a staggering 71% of customers now expect personalised interactions and 76% are frustrated when these expectations aren’t met.

Personalisation is intrinsically based on data. To offer a truly personalised experience it’s necessary to understand users’ needs and find the most effective ways to engage them. That involves the analysis of a sheer amount of data. There’s where AI comes in. It can collect, process, and analyse huge amounts of data, providing useful insights.

While with traditional personalisation you see online ads based on what you bought last season or on what other people are buying, hyper-personalisation takes it up a notch, making it possible for brands to treat every person using their website or app as a unique entity. So instead, it will draw information from data about your prior purchases to infer what’s your preferred colour palette, height, body structure, location, shopping time, preferred payment method to build an ad just for you.

This AI-driven capability to personalise a brand’s message, content, products and services will enhance the user experience, thus ensuring better adoption and retention.

AI-powered tools to increase productivity

A GiF of a young man sitting in front of a computer and high-fiving a robot.

Source: Giphy

Resizing images, making colour corrections, cropping images, removing images background. I’m guessing you do these tasks pretty often during your design work, right?

Luckily, AI now offers several solutions to save time with these repetitive, time-consuming tasks, leaving designers with more time to focus on designing better experiences.

UX designers, researchers and content writers have plenty of tools they can incorpore in their work, to help in different stages of the design process. Let’s go over some of them.

  • — This free tool allows you to “remove backgrounds 100% automatically in 5 seconds with one click”. All that painful time spent removing backgrounds from images by hand, can now be used for more creative tasks.

A baby smiling and laying down on his crib. Half of the background is transparent.


There’re also plenty other AI background remover tools, such as Removal.aiZyroHotpotand

  • Colormind — Using deep learning, this AI-powered colour palette generator learns colour styles from photos, videos and popular art.

Example of a colour scheme generated by Colormind.

Source: Colormind

Similar to this tool, there’s Khroma, Huemint, and Coolors.

  • Let’s Enhance — An automatic AI editor capable of increasing image resolution up to 16px in seconds, without losing quality. 

A kitchen utensil on a counter next to some vegetables. Half of the image is pixelated and the other has a good resolution.

Source: Let's Enhance

  • Magician — This is a plugin for Figma powered by AI that can help you by creating icons and images from text, copywriting, and more.

The tool generating 4 different icons from the input “hand holding water”. The selected icon is dragged and dropped into the design.

Text-to-icon functionally in Magician

  • Flair — An AI-design tool for branded content. You simply drag your product photos into the canvas, visually describe the scene surrounding your product and voilà, a high-end product photography is created in minutes.

A cosmetic bottle in a jungle setting from the input “cosmetic bottle on a mound of moss surrounded by flowers against a jungle with sunlight streaming down on it"

Source: Flair

  • Fronty — With Fronty’s AI-powered tool users can generate clean HTML and CSS code from a design or mockup within minutes, which can be edited and published in real time.

On the left a website’s mockup and on the right the mockup transformed in a web page.

Source: Fronty

  • Fontjoy — Generates font combinations through deep learning.

Screenshot of Fontjoy website showing a font pairing.

Source: Fontjoy

  • IconifyAI — This tool let’s you create professional app icons with AI in seconds. You simply have to describe the objects you would like to see in your icons, select the colour that best represents your brand, the style and shape you prefer. The price can go from $10 for 15 icons, to $25 for 60 icons.
  • DALL-E 2 — AI system that can create original, realistic images and art from a description in natural language. It can be a useful tool to help designers brainstorming potencial creative ideas.

Other AI-images generator include Images AI, StableCog, Artbreeder, Leonardo AI, Midjourney AI, Runway and Deep Dream Generator.


While AI is not able to generate original content or information as a human is, it’s able to imitate human language and structure. By auto-generating strings or story text based on inputs, it exposes you to new content ideas. A precious help for anyone doing UX writing, content writing or copywriting and in need to overcome writer’s block. Quality writing can be produced faster and more efficiently.

  • UltraBrainstormer — This is a creative writing assistant that will help you with content on any given topic. You can tailor the content by selecting the content type, tone, target age group, and target industry.

Selection of content type, tone, target age group, and target industry.

UltraBrainstormer (Source: Future Tools)

  • Sudowrite — This tool claims to be the “AI writing partner you always wanted”. You can type in a concept and it will generate up to 1000 words. It analyses your characters, tone, and plot arc and generates the next 300 words in your voice. It can help you with descriptive parts of text, expanding the parts of your story that feel rushed. It also has several editing features, like rewriting sentences, give you feedback on your writing with actionable areas to improve, and help you find the perfect word.

Related words feature (Source: Sudowrite)

  • CopyAI — This tool is an AI content generator. Here you can type in what you need (emails, social media posts, long-form blog posts, etc.), give some context and instructions, choose a tone, and the AI will give you multiple options. It also features an editor that helps you rewrite paragraphs and polish up sentences.

Form to create copy with the fields: ”What are you looking to create?”, “What are the main points you want to cover?” and “Choose a tone”.

Creating copy in

  • Jasper — This generative AI platform helps teams create content for their brand “10x faster”. It claims to produce high-converting marketing copy and help you overcome writer’s block. One of it’s features is called the Boss Mode and can be used to write long content like blog posts, reports, emails and stories. The content is 100% plagiarism-free and ranked for SEO. Jasper can read and write in over 25 languages by translating using deep learning.

Other AI writing tools include AnywordRytr, and QuillBot.


There are several AI-tools that can support UX researchers when conducting usability research.

  • Attention Insight — This tool allows you to validate concepts for performance during the design stage with AI-generated attention analytics. After you upload your design files or paste a URL, it generates attention heatmaps that show which design elements captivated the users attention. You can also define an area of interest to test and get the exact percentage of attention that object receives, among other features.
  • VisualEyes — VisualEyes simulates eye-tracking studies and preference tests with a 93% accurate predictive technology. Once you upload your design, their intelligent algorithms forecast user behaviour patterns based on extensive data from large scale studies. It features attention maps, areas of interest measurement, clarity score, and clarity maps.

Example of an attention map by VisualEyes

  • Expose — Expose’s platform uses AI to predict the outcome of an eye-tracking research study, eliminating the need to recruit human participants. It’s main features include attention heatmaps and areas of interest predictions.

    A comparison of eye-tracking and (Source:

    • WEVO — A platform that blends human input and AI technology to produce quick, precise users insights at scale. Wevo analyses data and feedback from +100 users per test, automatically checks for quality response, organises, and filters the results. It delivers scores, benchmarks and synthesised insights.
    • Happy Scribe — This software allows you to transcribe files in minutes. Although you still need to proofread the output by yourself, it still saves you a lot of time. It supports dozens of languages for transcription and subtitles. Tools like this make translation cheaper and quicker while opening the possibility to offer content in multiple languages. Although a human translator still needs to check the style and flow, the AI translation can serve as a good first draft.
    • Holler — This tool allows you to create one-question surveys for customer satisfaction, employee engagement, and market research, and then find trends and patterns in the data collected with AI’s help.

    An example where you can ask the AI to make a list of the key action points that should take priority (Source: Holler’s website)

    • Olli — They present it as the “ChatGPT for data analysis”. The platform connects with the company’s data, visualises it and is able to answer questions and provide analysis on all types of data.
    • ChatGPT — It takes only a couple of minutes to find a number of articles giving ideas for using ChatGPT in user research. They include asking the AI to imitate users by generating interview answers and comments on design concepts, to write research questions, create a UX persona and more.

    📝 This article by Jeff Humble in The Fountain Institute is a pretty good help for anyone trying to figure out how to use ChatGPT in their research and design processes.

    An example of Chat GPT’s answer to the question “What’s a good user experience for a flight?”

    Faster prototyping and wireframing

    Prototyping is an essential process for any design team creating a new website or app. With prototypes, designers can explore new ideas and solutions without having to invest in development. They’re also useful to test with real users and find out any usability issues.

    Using AI-based prototyping tools, designers can get a glimpse of the end outcome at each stage, resulting in increased productivity and reduced effort. They make product development process simpler, and gives teams the opportunity to focus more on testing functional prototypes.

    While designers have been using different prototyping tools available in the market, new AI-powered UI design tools are emerging.

    With Uizard, for example, you can scan your sketches and mockups and automatically transform them in designs for a mobile app, web app or website. Creating these quick, high-fidelity prototypes can be pretty useful to test ideas with real users.

    A design idea transformed in a fully functioning UI prototype with Uizard

    Improve web accessibility

    Although countless websites are created every day, they’re not build with accessibility in mind. Summing those with the all the other existing websites that also fail accessibility standards, it’s pretty clear how people with disabilities struggle to navigate the web.

    Many AI-powered solutions are already helping making the web a more accessible place for people with disabilities. Facial recognition is an AI-based alternative to typing a password. This is clearly useful for blind users or for people with some kind of physical impairment that prevents them from using their arms. Speech recognition is a great feature for users with some physical, cognitive and learning impairments who might struggle with typing.

    Taking advantage of AI-powered testing and error fixing tools, companies can evaluate the accessibility issues on their websites, and fix issues automatically during website development.

    Deque — Deque provides axe, a digital accessibility toolkit which includes axe DevTools, axe Auditor, and axe Monitor. Their tools integrate machine learning to enable automated testing. Deque’s automated testing alone identifies 57.39% of accessibility issues. When combined with a technology called Intelligent Guided Testing (IGT) the number jumps to 80.39%, on average. Their aim is to empower the development team by providing tools that can automatically test a webpage as well as the tools to speed the testing that needs human input.

    Their ML capability is driven by the years of data on comprehensive accessibility audits for thousands of organisations, and the number of users of their tools that provide a wealth of data.

    ML — and AI as a whole — show immense potential to transform digital accessibility and any organization’s ability to create inclusive digital spaces.

    Dylan Barrell, CTO at Deque

    UserWay — UserWay offers a AI-powered accessibility widget which functions were built to ensure ADA and WCAG 2.1 compliance. The widget corrects website code to accommodate screen readers and other assistive technologies according to WCAG standards and adds customisation tools all visitors can use. Their AI is able to interpret and write accessible code similar to a team of experts, but much faster.

    They also offer a Scanner tool that reads the code on a website and evaluates its structure, navigation, links, buttons, and many other criteria to discover WCAG violations.

    UserWay Accessibility Scanner (Source: UserWay)

    AccessiBe — The company has developed an automated web accessibility solution using AI, machine learning and computer vision, to address the more complex requirements needed for screen reader and keyboard navigation accessibility.

    As an example, using image recognition, the accessWidget scans all the images on a website, and if any alternative text is missing, it automatically provides accurate and elaborate alt texts to those images.

    accessWidget using computer vision to add alternative text to images

    Other AI-powered tools for web accessibility include Max Access, AudioEye, EqualWeb, Monsido, and Silktide.

    An evolving technology with great potencial

    GIF representing the human evolution ending in a robot.

    Source: Giphy

    As AI continues to evolve it will continue to have an undeniable impact on the future of jobs. By analysing over 200.000 jobs in 29 countries, PwC estimates that up to 30% of jobs could be automated by the mid-2030s. However, while AI might take some jobs away from humans, it will also create 97 million new jobs by 2025 (as predict by the the World Economic Forum).

    While some jobs in the finance, transport and logistics sectors might be easier to automate, jobs that depend on social skills won’t be displaced so easily. UX is one of those cases, as it requires empathy to relate with users and understand their issues.

    As we could see from all of these tools, AI is already changing the design of products and systems and is certainly a great opportunity for UX designers to transform the way they work. By providing valuable insights and automating certain tasks, AI helps UX designers create more personalised and efficient experiences for their users.

    Related Articles