How will AI affect UX?

“Machine intelligence¹ is the last invention that humanity will ever need to make”—James Barrat
A.I. is becoming more complex and complete each year.

Ultimately, just like most of us, UXers will lose their job, due to Artificial Intelligence (AI)…

… But, that will take some time… and like in every revolution², new jobs will appear, e.g. Machine-Human Interaction designer, AI growth mentor, among others. (UXers shouldn’t get too concern, for now).

The topic is immense and dense (both in its philosophical and technical aspect), so this article is a mere introduction / exploration of the subject, that might be followed by more articles. 
So for the sake of this article, some simplifications and omissions were made.

But then you might be asking: 
Why did you choose to write about this troublesome topic?

Well, last week after watching Ghost in the Shell (the original, obviously), I stumbled upon a book about programming Artificial Intelligence (AI). The rediscovery of the book and the movie, led to a chain of thoughts that ultimately ended up in the question:

“How will AI affect UX?”³

I immediately started to read about the subject, and found some interesting articles⁴ and TED talks⁵, that further inspired me to create this article.

For starters, and those who think that this is still a distant reality, AI is already affecting the UX of the applications in rather nice (and disturbing) way.

Example of this is Facebook, which is using AI to improve the timeline, allowing image search, video real time classification or simply translate comments. Other very known examples are Siri and Alexa learning from your habits or even Tesla using it’s entire fleet information to make decisions. Another example of AI nowadays, is also chatbots, that are transforming (slowly) how clients interact with companies and services.

But, let’s step back for a moment and understand What is AI?

What is AI?

AI stands for Artificial Intelligence, and a (simplified) definition of what it is, can be something like:

“The capability of a machine to imitate intelligent human behavior” — Merrian-Webster

However the subject is much much more complex, and is divided into the several branches⁶:

  • Reasoning, problem solving — studies how a problem is approached, in order to be solved. While early AI used step-by-step approach, nowadays, AI approach is starting to resemble humans, using intuition. (e.g. DeepMind’s AlphaGo program)
  • Knowledge representation — focus on how the information can be represented so that the system can use it to solve a complex problem. (e.g. IBM’s Watson assisting in drug discovery)
  • Planning —focus on how using the available options, the system can achieve the goals. (e.g. unmanned vehicle crash dilemma)
  • Learning — studies how computer algorithms can improve automatically through experience. (e.g. DeepMind’s learning to play Atari games — Breakout game)
  • Natural language processing — focus on how the system can communicate and understand human language, leveraging using it to learn. (e.g. IBM’s Watson won the Jeopardy TV show by learning from Wikipedia)
  • Perception —studies the ability to use input from sensors (such as cameras, microphones, tactile sensors, sonar and others) to deduce aspects of the world. (e.g. Honda Asimo robot)
  • Motion and manipulation —studies Intelligence that is required for robots to be able to handle such tasks as object manipulation and navigation, with sub-problems of localization, mapping, and path planning. (e.g. Honda Asimo robot — see it in action)
  • Social intelligence — aka as Affective computing is the study and development of systems and devices that can recognize, interpret, process, and simulate human affects. (e.g. SimSensei system to support clinicians and healthcare providers — see it in action)
  • Creativity —addresses creativity both theoretically and practically: via specific implementations of systems that generate creative outputs. (e.g. Google DeepDream)

AI is then categorized in three levels that differ between each other, on how many of the previous branches the system can master⁷:

  • Level 1 - Artificial Narrow Intelligence (ANI): aka as Weak AI, Artificial Narrow Intelligence is AI that specializes in one area. 
    (e.g. Deepmind AlphaGo program, can’t maintain a dialog in natural language)
  • Level 2 - Artificial General Intelligence (AGI): aka as Strong AI, or Human-Level AI, Artificial General Intelligence refers to a computer that is as smart as a human across the board. Creating AGI is a much harder task than creating ANI, and we’re yet to do it. 
    (e.g. Lance Bishop from Aliens)
  • Level 3 - Artificial Superintelligence (ASI): aka as We are f***ed, or an intellect that is much smarter than the best human brains in practically every field, including scientific creativity, general wisdom and social skills” — Nick Bostrom, Oxford philosopher and leading AI thinker.

We have achieved Level 1, ANI, and has it’s being used more and more. However, achieving Level 2, AGI and then Level 3, ASI, is much harder, and it will bring the so called “AI Revolution”, that will either takes us to a new paradigm of existence, or make us disappear.

“I prefer the term «Artificial Person» myself.” — Lance Bishop

Also something important to mention, is the fact that AI growth will be exponential since more intelligence generates more progress, which in this case will make the system more intelligent. 
Here’s a simple chart from explaining that:

Graphic showing how human progress might be very different from what we think

Now, getting back to the original question…

How will AI affect UX?

I’ll try to answer this in 3 snapshots of the future, being all in a not so distant future. However given the exponential growth of AI, the distance between each moment will be shorter than the one before:


  • AI is used to analyse and understand human behavior patterns across the target audience to give insights to UXers when creating new mockups.
  • AI is also used commonly through the application to simplify it’s usage, being communication in natural language (voice), common.
  • AI is a commodity.
AI, might not even exist physically, but it’s insights shall be precious.

With the valuable insights given by AI in every stage of app creation (Competitor analysis, User Research, mockups, Users tests and service design), doing better UX become more simple and affordable. This also increased users expectations toward applications and services.


  • Transmorphic applications are a reality, completely adapting to users behaviors and usage, accordingly to well define patterns from UXers.
  • AI has made construction of applications simple, given suggestions, regarding best practices related with the unique characteristics of the target audience.
  • Applications are used in through several channels and communication means by users: movement, voice, and other ways.
Transmorphic application, displaying information in known patterns.

By now, UX market has changed, AI UX Assistants become wide spread, leaving UXers to support and audit its work. Also, iterations and optimizations in the applications become faster and in real time, adapting to the individual user and his specific needs.


  • UI is no longer built, it is generated when needed.
  • Most interactions occur in most natural way to communicate, being sometimes visually, other times verbally.
  • There aren’t applications or systems, just information, and AI simply convey information in the most effective way.
Convenient and efficient conveying information, through visuals and voice

AI became our extension, eliminating apps need, by collecting and providing information from the “vast and infinite net”⁸. With no more interfaces to create, UX job role (and related UI & FE) have disappeared, and UXers are free to follow their ghost.

Final thoughts

“The best way to predict the future is to invent it” — Alan Kay

The general consensus is that AI will replace less creative jobs first, but with its exponential growth, the same is expected to happen to creative jobs. Although these thoughts might look futuristic, probably they are closer than what we want to admit.

Nevertheless, instead of fearing, we should embrace change and evolve (or get extinct), maybe become teachers, until the student (AI) surpasses the master: providing users, a natural and seamless experience, impossible to be devised by humans limited capacity.

In the next articles, we will explore in detail the impacts of AI in UX with examples, that at the moment are hard to obtain, since they are yet to be invented.

“The future is dark, the future is vile” ⁹, but so are we.