Design Development

Design by machine

Design by machine

Is it possible that machines will be able to design on the same, or even better, level as humans? Will people be replaced by computers (normal or quantum) in fields such as design and art? Let’s explore these questions furthermore.

What is design?

In order to answer this question I will borrow few definitions Wikipedia provides:

“Design is the creation of a plan or convention for the construction of an object or a system.”

“(noun) a specification of an object, manifested by an agent, intended to accomplish goals, in a particular environment, using a set of primitive components, satisfying a set of requirements, subject to constraints;”

“(verb, transitive) to create a design, in an environment (where the designer operates)”

That’s it. So, simply said design is the act of creation. No matter what industry or subject it relates to, until creating something new is included, we are talking about design(ing). Interesting thing is that not in any part of the definitions above are mentioned human beings as important factor in the whole process. Can we assume if a computer execute a set of specific commands and the result is creating something new that the computer is actually designer?

It may sounds ludicrously at first, but if we think about it it successfully passes the definitions above. If all the requirements are fulfilled without any fallacies than the theory is proven – the computer (or program) in example is a designer. Let’s take this question a bit further … There is a field called machine learning. The goal of it is to construct algorithms able to study and learn from statistic data and to adapt itself (algorithm or program) according to results. So, can we use machine learning to create a program able to design on its own – without any human interaction?

We humans are often afraid when it comes to advanced technology and robots. We are afraid of being replaced. In many fields this fear is justified. Simple routine tasks can be done by computers faster and better – with smaller amount of mistakes – than by humans. For example, factories with computerized assembly lines work in much faster pace along with keeping the error rate at minimum than we ever could. Today, we are starting to see robots assisting surgeons in the operating rooms. Well, they are used as another tool to increase the capabilities of surgeon. They are not operating on their own, yet.

It’s not magic

When you decide to become a designer and work in design industry, you have to learn certain rules. These rules can vary due to specific field you want to work in. Graphic design, for example, requires you to learn about perspective, rule of thirds, golden ratio, color theory, psychology, layout theory, typography, grid system and so on. Also, don’t forget learning about various design movements in the history. Next step for designer is to supply this knowledge by practice. He have to train his eye to spot the smallest details and to understand why, when and how to apply his knowledge. This can be done only by doing and there is no supplement for it.

There is a foundation that has to be learned and practice to be done. However, nothing is suggesting that only humans are capable of following this path. In theory it doesn’t even seem so hard to replace the human designer with artificial. Take the stack of theory designer has to learn. Every single piece of it is a variable, chunk of data, that can be given to computer. Once those data are stored, computer them can be accessed and used over and over again. In other words, it has the required knowledge.

What about the practice part? Here, I want to use machine learning mentioned before. We can build a database of various design examples and deliver these data to program containing both, learning algorithm and knowledge. With this foundation, program will be able to loop through the database and apply stored knowledge to every record. The more data will be in the database, the more accurate the final result will be. This “learning” can be enhanced furthermore by providing feedback and correction suggestions from human observer. Just like in school. With enough time and samples in the database, the computer can become better designer than human being.

Creativity issue

No matter how pretty, or horrifying, this may sound, one thing is still remaining. This thing is what distinguishes the best designer and artist in the world from the rest. It is creativity. Even though we will successfully teach computer the necessary knowledge and supply samples of data to improve its abilities, this will not make computer creative. For a long time I thought that creativity is something available only to humans. Recently, I was thinking about creativity and the process of innovation through and through. Now, I’m starting to change my mind.

The reason for it lie in history. Many examples of creativity and innovation in recent centuries were more or less the result of interconnection between two or more different fields (or environments). Henry Ford’s assembly line, Da Vinci’s glider, military camouflage. Assembly line was used in slaughterhouse. Glider and camouflage was taken from the nature. Most of inventions we have today was already around in some form. We just made another iteration or connect it with something else.

So, my question is following: If we store all the knowledge human kind has into computer with enough computing power and capacity, will this machine be able to produce these intersections? Will we be witnesses of some unique creative spark? If yes, can we call this a creativity?

Conclusion

Since design is based on rules I am certain that computer can become very good designer and even outrun human designer after time. The only unknown variables are the time it will take and how advanced the technology will have to be to make this reality. Will this computer able to simulate creativity and produce groundbreaking inventions? I don’t know. What do you think?

If you liked this article, then please consider subscribing.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.