The Flow Of Generative Design
Generative design mimics nature’s evolutionary approach to design. Designers or engineers input design goals into generative design software, along with parameters such as materials, manufacturing methods, and cost constraints. Then, using cloud computing, the software explores all the possible permutations of a solution, quickly generating design alternatives. It tests and learns from each iteration what works and what doesn’t.
With generative design, there is no single solution, instead there are potentially thousands of great solutions. You choose the design which best fits your needs.
The idea of complete replacement of the designer by the algorithm sounds attractive and modern, but it is wrong and weakly promising. The designer helps the team to turn the raw idea into a holistic interface with a good logical work, information architecture and visual style that solve business problems and strengthen the brand. In the course of the work, it takes a huge number of decisions, many of which can not be described with clear procedures. In addition, incoming requirements are erroneous and contradictory in separate details, so that the designer helps a manager to solve these collisions — it makes the product better. All this is much better than choosing the right template and painting it with modern stylistics.
But if we are talking about creative partnership, when the designer, in pair with algorithms, solves product problems — there are plenty of perspectives and good examples.
Roelof Pieters and Samim Winiger study the history of computers in detail, as design assistants. They distinguish three levels of development tools:
-The first generation has translated analog instruments into programs and has developed along the path of increasing capabilities.
-The second has learned to take over part of the routine operations that previously required professional expertise.
-The third should become a co-author of decisions, helping to find new interesting directions.
The authors conduct a good analogy with generative art — the designer determines the algorithms by which the work is formed, and then manually selects the most successful derivatives. If the bust is not entirely manual, the computer will also help to filter the resulting stream — the work will become even more productive and creative.
Algorithmic design should become such an “exoskeleton” for the designer, significantly increasing the number and depth of development of solutions. How can the designer and computer interact in such a node? You can see the overall process of food work:
-Explore the space of problems and take for solutions those that will give maximum value to business and users.
-Explore the solution space and choose the one that best covers the problem.
-Generate, launch and distribute a product that solves the selected problem.
To evaluate the effectiveness of the chosen solution in practice and optimize it we need markets, segments of the target audience, business models, product types, internal organisational structures at the same time and there are so many, that universal solutions for all are just utopia. So, it should rather go about own decisions of companies, confined to specific tasks. Going down to the level of specific design solutions, it builds the interface, preparing graphics and content.
Simple publishing tools like Medium, Readymag and Tilda have already reduced the amount of manual work — there are plenty of good templates with which you can create a good result without a designer. Improving the templates will make the entry threshold even lower.
Yes, this way does not create a revolutionary product. But you can free yourself some time for this. And we must admit that a huge part of everyday tasks is more than utilitarian and does not require revolutions. And if the company is mature and has a design system, connecting to its algorithms than it will allow you to do more with less. For example, a designer and a developer describe the logic of processing incoming signals — content, context, user information and actions, and then the algorithm itself generates screens based on ready-made patterns and principles. It allows you to achieve fine tuning for a specific narrow situation or scenario without having to draw manually and develop dozens of screen states.
Creating the same type of graphics in different variations is one of the dullest parts of the designer’s work. It takes a lot of time and demotivates, while these forces could be spent on more meaningful product work.
And absolutely black magic takes place in the direction of neural networks. One of the best examples, the Prisma application edits photos under the work of famous artists. Will this work be selected by illustrators? Those who have a good style are unlikely. But this, again, will reduce the threshold of the entrance to receive good-quality illustrations for an article or a site where
But this, again, will reduce the threshold of the entrance to receive good-quality illustrations for an article or a site where absolutely unique approach is not required. Farewell, sad stock photo! And for complex stylistics it will help to get a quick sketch in the spirit of “what if we try to paint a building or a cat in our style.” You can do storyboards and describe scenarios in the form of comics (photo easily turns into a sketch). We think this list of applications will expand soon.
The story is beautiful, but we must understand that the algorithms are built according to clearly described rules, even if they are great with the help of machine learning. The strength of the designer is that he can break these rules and ask new one, so in a year something quite different will be considered as a beautiful. Of course, not everyone in our profession is strong and the algorithms will easily replace lazy ones. But for those who know how to both comply and break the rules, a whole new set of tools and opportunities will open up.
Stay on track with us!