Generative Design: Artificial Intelligence (AI) in Design

For some, the future potential par excellence, for others merely a trendy phenomenon of the digital age, combined with a little Hollywood science fiction, a few buttons and gimmicks. We are talking here about artificial intelligence (AI). AI is a broad topic with many different subfields in many different industries. At Cleverclip, we are experts in the field of design, which is why we would like to focus on the topic of “generative design”. How will AI influence design processes in the future? What opportunities will arise? What challenges can we expect?

What is artificial intelligence (AI)?

Most of us have already encountered it in some form, for example in the form of chatbots in customer service or playing our favorite Alexa playlist. And we’ve all certainly insulted Siri or confessed our eternal love to her. Artificial intelligence is indeed not just entertaining science fiction talk.

AI is a branch of computer science that deals with the automation of intelligent behavior. It simulates human intelligence with machines, especially computer systems, to solve challenging problems in a simple way. Algorithms that can imitate intelligent human behavior are primarily used for this purpose.

What is generative design?

Generative design, as a sub-discipline of artificial intelligence, involves a design process in which the product is not generated directly by the designer, but with the help of a programmed algorithm. To get a better idea of this, it is best to look at a few examples.

Airbus uses generative design to develop the cabin partition wall for the A320. After specifying the dimensions and load requirements, intelligent software creates suitable design options. BMW subsidiary Designworks uses similar processes to design and shape rims and seats. And Nutella’s “Nutella Unica” campaign in 2017 used AI-powered algorithms to create over 7 million product jars, all with custom designs. The campaign went viral on social media, fans of the brand shared over 10,000 custom videos, and all 7 million jars were sold within 4 weeks. Proof enough of how much potential AI has in the design space.

Generative design: potentials

Artificial intelligence is not only interesting for global players. Especially with the advancement of technology and the ever-increasing possibilities within artificial intelligence, various, smaller design processes can be optimized.

Increasing productivity

We all know them, those nerves- and time-consuming tasks that simply have to be done. It’s not much different in the design process. AI-supported programs can remedy exactly this situation and support designers with certain routine tasks.

At the same time, special design systems, components and style guides can be generated quickly, adapted for different processes, and used across the team. For example, Disney has been using the semi-automated technology “Meander” to fill in animation frames since the Oscar-winning short film “Conquered in Flight”.

Designers and developers could define the exact content, context, and user data while a platform compiles the design using principles and patterns. This enables efficient work and gives designers more time to focus on other creative processes.

Content personalization

Customized content is one of the biggest AI potentials, especially in terms of user experience. On social networks, for example, algorithms analyze our user behavior and display a personalized feed based on it. Especially in the field of design psychology, artificial intelligence can contribute enormously to brand success. The AI can create multiple variations using a known pattern to create custom designs. I mentioned one example earlier: Nutella uses an algorithm to create individual packaging for each consumer.

The webpage builder Wix is probably supposed to enable every user to create their own stunning webpage with the help of an algorithm. For this, the website owner is asked a few simple questions to learn more about their needs. The algorithm then chooses between billions of variations to create a unique website in a few minutes. The whole concept is probably not quite there yet and has some flaws, but it sounds very promising for a start, doesn’t it?

However, we can look forward to much more here in the future: With facial recognition, for example, special designs can be created based on user characteristics.

Tests and analyses

AI-powered A/B testing allows designers to test specific products on a large number of users in a matter of seconds, allowing them to get feedback early in the design process and tweak the design if necessary.

Here, features such as facial expression recognition, gesture recognition, and eye-tracking heat maps can be used to test the design for usability.

AI, therefore, has two key potentials within the design process: On the one hand, in production: Here it supports designers with individually created templates that are optimized for the target group, especially through rapid testing procedures. On the other hand, from the user’s perspective through a personalized user experience with numerous variations, face and speech recognition, computer vision, and much more.

Generative design: challenges

Of course, we have to keep our feet on the ground. AI is exciting and offers enormous future potential, but it is not yet where it should be or where we would like it to be. The following hurdles arise:

Lack of differentiation of nuances

We laugh, we cry, we are happy, we are sad. Emotions make us human. The emotional intelligence distinguishes us. (Well, it is indeed missing in some, but you know what I mean. ;)) How should a robot imitate something like that? Understanding and analyzing emotions correctly is one of the biggest weaknesses of AI.

Create moving content

Yes, AI allows for the creation of diverse design variations tailored to the user, but is this really emotionally moving content? This ties back somewhat to the previous weakness. For just as AI is difficult to measure feelings of the user, it is from difficult to express the emotional depth in design. Especially with large data sets, a lot of work is done with abstract elements that may be appealing as an overall picture, but just can’t convey the same feelings and portion of storytelling as an independently painted painting.

The danger of bias

Researchers have also warned of the danger of so-called “algorithmic bias”. This could increase discrimination and racism in particular, as decisions are made on the basis of skin color or facial features, for example. In this context, the Council of Europe and nine non-EU member states signed the “Human Rights by Design”. This calls on designers to consider individual rights as well as dangers to society in the context of AI development.

Overcoming AI Weaknesses

So how do we properly deal with the aforementioned challenges? And how can we make the best possible use of the potential of AI? At this point, three points in particular are important:

Understanding tools and processes

Everything that is new is scary. Only when we get to know certain processes and understand how they work can we deal with them successfully and implement them. Of course, this includes a step-by-step examination of the existing types of AI and the possibilities of integrating them into one’s own design process.

Consider ethical aspects

In order not to give prejudices a chance and not to run in the wrong direction with AI, it is important to set ethical principles right from the start. Pioneers in this area would be Google and Microsoft. The global players have already established AI principles.

Consideration of external factors

Those who want to conceive successful designs with AI must not only understand how different designs can be adapted depending on certain variables, but also how products can be adapted to the different needs of users and to the changing environment. Flexibility must become a central design principle.

Is generative design putting our jobs at risk?

What’s the real answer to the crucial question? Will artificial intelligence replace humans? In our case, eliminate the designer’s area of responsibility?

Artificial intelligence can seem frightening to some people. It is therefore important to take a close look at the subject and understand what AI can and cannot do. Yes, machines can analyze large amounts of data, optimize processes, take over simple operations and develop patterns. But they cannot taste and feel, look into the future and reflect on the past. Generative design, therefore, cannot take over the designer’s job, but only shift the focus. Instead of grappling with standardized processes, it leaves him more time to devote to the creative thought process. We should not see technology as an enemy, not look forward to it with fear or even assume that it will put us out of work. We should see it as a support, as an aid to speeding up and perfecting our work.

Conclusion: The future of generative design

Meaningless science fiction? AI certainly isn’t. Generative design is a groundbreaking technology of the future and offers enormous potential, especially in the optimization and analysis of processes. Designers who see this opportunity and learn to deal with AI will be able to remain competitive in the future. Others, on the other hand, risk being overtaken by competitors.

We live in a different age, and digitization has become an important part of it. If you don’t close your eyes and are open to this change, you will see that robots and algorithms are not that scary, but support humans as friendly companions. Maybe in a different way than a golden retriever, but who knows… 😉