How AI Will Take Designers Job and Make the Web More Personalized

I know that is a little bit cheating topic because there are too many mainstream words as “ai”, “big data” and “algorithmic design”. But anyway for me it was fascinating to imagine how the designer job would change shortly. And I want to highlight that it is only one of my point of views. And feel free to let me know what do you think about it.

So, let’s start with microtargeting. Long story short, it’s showing to different users content that will better fit them. Last years this approach became very popular, especially in advertising. Social networks collect an enormous amount of data, and that help build detailed digital portraits for each user. Therefore you receive maximum prevalent for you promoted posts. And it looks like this approach can give hazardous and unpredictable consequences. Like “Cambridge Analytica” company what working with big data helps Brexit and Trump election happen.

By the way, Here is an excellent investigation by Das Magazin (the German language)

But in the case of advertising, we choose the particular group of user to show our ads post. With the election it is even simpler, we just should persuade the user that our candidate is better than others, what is easy as we know about user more than he knows about himself.

‍

But going deeper if we have tens or even hundreds of features for each person we will have millions slightly different digital portraits. Maybe it is no critique, but what if we would be able to give each of them unique content that ideally fits. I don’t think that it is logical to hire so many writers, and what is worse they definitely will not be able to catch such small difference in sentiments.

Right now, computers power reach a level when AI (artificial intelligence) can write a novel or even create music. So why don’t use this power to adjust content for millions of people? For example, let’s take any online magazine. In this case, not only the “recommendations” section will change. Even all texts should be adjusted to a particular person and conditions depending on grade, mood, personal interest to a theme, and any other features that we are not able to predict at this stage. Anyway, this is the task of AI to define features that are more influential for taking content in a particular case, measure people by this feature, and adjust the style of writings accordingly.

This idea looks even more impressive from the design perspective. Like the style of presentation, content makes the huge subconscious influence for the user. On top of the level, the goal of the design is to help user taking content easier and contribute to making effective actions. And of course, make business more profitable at the same time.

Now in the design world is very standard practice to create personas that describe an average type of user and have usability testing sessions. But all of this is really awkwardly. What if we would be able to create an infinite amount of personal and improve interface using information from usability training sessions automatically or even real-time.

We could change not only content and writing style of content to make it maximum efficiency for every particular user, and also adjust styles, grid and overlook of a web page.

Algorithmic design is more popular in architecture and interior design because it allows creating insane shapes by just describing it with a single formula, without drafting all details by hands. In the UI interface, the algorithmic design is not so popular. The bold old solution is responsive design, but it depends only on the device instead of a user. Also, Jon Gold did tremendous work to study algorithms working with fonts:

Taking The Robots To Design School, Part 1

The thing is that we don’t have to predict each tendency. For example: if this user is 1% more angry than average, should we make this colour darker or brighter? Solution calculated by AI using bid data of users behaviours could suggest us that blue colour is more suitable for this case, and it allows us to take more attention and confidence from such user.

This approach is not good for all websites, but it sounds weird:

“Design that made without designers for each user personally and is entirely unique to reach all goals of business”.

So next time when you’ll start thinking that this website is ideal just check if your friend sees the same.

‍