April 23, 2021

How Artificial Intelligence helps create a better User Experience

Insights on how AI and UX work together to improve themselves and some use-cases that mix those two technologies.

The user experience is a big part of the reason people might or not use your software application, it’s very important to be able to understand humans in order to design these experiences, as a reminder, User experiences (UX) can be molded by User Interfaces (UI), and while their objectives are fairly similar, they are not the same.

UX improving AI

A great user experience for AI will make the interaction with the machine as human as possible. At this point in time, we humans interact with machines by talking like machines, this is the user experience we’re used to by now, things like saying “Hey Siri” or “Alexa, play this song” are part of that machine-like behavior we do while communicating with them, we don’t have to say “Hey + their name” every time we want to talk to someone, but we’re ok doing it with AI.

At one point, that’s going to change, as AI developers work closely with UX designers we’re receiving more human-like behavior from AI and we will adapt to that, the less we need to be robotic while we talk the more we feel like we’re talking with an actual human, getting the general user experience of the AI to an almost perfect state.

AI improving UX

Now, I think it’s pretty clear how user experience has helped improve AI in the past couple of years, but what about the other way around? How does AI improve the creation of better experiences for end-users? Here are some examples:

AI and UX

Thumbnails and posters

Netflix has created a great algorithm that I’m sure has helped UX designers in a big way, transforming their job of producing hundreds of posters into designing one template that adapts to the viewer according to their location, age, likes, etc. Now those designers are spending less time doing the UI work and more time finding ways on feeding their algorithms different particularities so their experience gets better.

The way the algorithm works is like this: the AI is capable of cutting characters from different scenes from movies and tv shows and also selecting different backgrounds, the end product being a poster with some characters, a background and the movie or show’s logo. This first started as an A/B testing utility but has evolved into an aggressive targeting strategy.

They now receive enough information to understand the type of posters people that like to watch certain type of films and create posters that are specific to that target audience, for example if you love action films a la Michael Bay filled with explosions and adrenaline influenced cinematography, then there’s a high chance your posters are portraying that, you will start to see more action like backgrounds and scenes in your posters and thumbnails.

UX and AI


This is not the first time we’ve talked about chatbots and their particularities, but this is a great example on how users love to interact with human-like machines. UX web designers usually apply UI knowledge to generate different traffic flows on their websites according to their objective. But now, some websites are utilizing chatbots to redirect users and basically being their “website assistant” while at the same time receiving important information from the user, like their name, email and more demographics.

This varies a lot from user to user and as of today, a lot of people still like to browse through websites and have the UI take the hold of their general user experience, chatbots are very helpful when users have doubts but the general UX is still empowered by the UI for the most cases, it will be interesting how websites will work in 20 years when we have a more developed AI and users are more used to interact with bots than to navigate a website.

AI and UX

Heatmaps before production

The traditional way for an UX designer to improve their website is by checking heatmaps and general user flow on their sites. By doing this, designers have to do a general research and development phase in their process in which they analyze competition’s websites, behavioral patterns from different industries and based on that they develop their user experience map and put it on production, with this method here is when the real testing starts, you can test out your UX flow with people but unless you have a group of people that work on the industry you are targeting, production tests will be the ones that will give you better results.

Today we are seeing a different approach that lets UX designers test their flows before getting into their beta tests with real people and production. With AI having the ability to interact as humans and even have different personalities and traits, designers can do MASSIVE tests involving hundreds and hundreds of AI’s with different backgrounds, likeness and even embedded cultures. This way they get amazing heatmap results that usually would take 5 or 6 months in production, in just a couple of minutes.

It’s interesting to see how AI has been integrating itself in more and more methodologies and it’s even more interesting how it always works in a two-channel improvement deal, the service or methodology improves the AI and then AI comes back to improve them. Also as AI is becoming closer to humans, the way we interact with them is getting more human-like thus making the user experience more human than ever. Click here to learn about our AI solutions.

Case Study from Arkusnexus

3065 Beyer Blvd B-2
San Diego CA 92154 - 349

mind hub tijuana