What does chess have to do with design? When and how will machines replace designers? And what is Elon Musk's role in all of this? Get a cup of coffee and read on if you find any of these questions intriguing.

Photoshop, Sketch, Figma. Waterfall and scrum. Design thinking, design system. We are surrounded by many terms, tools, and methodologies in our work, which is going through a series of revolutions again and again. Even if the market for tools and philosophy grows every year, one thing remains constant: the goal we face as designers. Essentially, the goal is to visually design a solution that is then implemented by a development team and becomes a real product. The rest of it is just orbiting slogans that aim to get us to this point.

Due to the development of technology and the growing popularity of digital products, the solutions we design are becoming more comprehensive and sophisticated. Designers as individuals are increasingly giving way to teams of designers who must learn the rules and forms of collaboration. The tool market comes to the rescue, as it recognizes the direction of design development in advance and offers a wide range of convenient ways of working together in real time.

Nevertheless, when the scale of a project grows, the number of views and processes expand by another batch and a new member joins the team, then we are faced with the common feeling of design overwhelm. I call it the design-hydra. One process creates a dozen variations and states, some views need to be simplified, some shifted to a different phase, and the upcoming release calls for reorganization of the project. The project is out of control. One designer’s solution turns out to be incompatible with another created in a different place and time.

Chaos prevails. Now what?

Design and systems

Let's start with some philosophy. In order for our design process to not buckle under the weight of product expectations, we must ensure that our role is clearly understood and that the form of work organization does not weaken along with the duration of the project. A solid foundation is needed.

Here is where we come across the first term that will appear several more times in this story: systemic thinking. This enigmatic term refers to an approach to work based on a relatively simple principle - everything is connected to everything else. The Butterfly Effect. Using system thinking, we assume that every element we design has a relationship with another element or component.

We are applying the color to the component according to the parent style set. We are designing a process based on an analogous piece of code in another part of the product, and changing just one element will lead to a series of more or less expected reactions and consequences. When a button is pressed in one place, a lamp will be lit in another.

All things are connected. Even though it may appear inconspicuous, this is not something we can omit or ignore. The dependency system is simply there. Our screens and interface components circulate it, waiting for us to overlook a connection and add unnecessary components or lines of code to increase the technical debt of the product. As our awareness of this presence increases, we will be able to exert greater control over the entire design process. Our decisions will have inevitable consequences. Only we can determine whether these consequences will come as a surprise or as foreseeable.

Putting this into practice means we should analyze all, or at least the vast majority, of our design decisions if we wish to think systemically. Try to gain a comprehensive picture of the entire product by studying it from a bird's eye view. Search for patterns and examine processes and interface elements that have already been developed.

Of course, this is not an easy task.

There is no way for anyone to know all the nooks and crannies of a product's interface. This is where the recent design system comes in, a tool whose basic assumption is to document and close your decisions into a framework that is difficult to breach. A source of truth like this helps to systematize the project and its components, as well as being a reference point and a handy way to find patterns and existing solutions quickly. As a result, consistency is maintained in the product without forcing us to accumulate this knowledge in our heads, which could quickly lead to what is known as the bus factor.

A product that is based on the design system is conducive to systemic thinking, though the two terms are not interchangeable. Can a product be based on a design system and yet not think systemically at the same time? Of course.

Take a look at any application on your phone whose design differs from market quality standards. Whenever chaos creeps into a design and components start to become inconsistent, it is a sign that system thinking was missing from the design process. This design of the system can be enhanced by adding any number of new components, since we, the designers, still decide about the limitations.

It is up to us whether or not, when designing a new process, we take into account conditions and dependencies that arose while designing a different, similar set of screens and interactions. It's up to us if we want to design a new type of button. We will have to determine if the change in error state presentation will be preceded by an analysis of old instances of this state. We are responsible for the outcomes of our design decisions.

Can such thinking and analyzing the product in terms of its dependencies be considered flawless? No, of course not. The first and most important disadvantage of this method of work is its time commitment.

If we assume that everything we design needs to be verified and analyzed beforehand, it will soon become apparent that the design process will take much longer than it should.

A whole new set of challenges, related methods, and solutions arise here. Each decision is weighted with an estimated value. We will soon discover that there are tasks and choices to which we can devote 20% of our energy, others 50%, while some extremely important ones require so much analysis and anticipation that we must devote the most time to them because their implementation determines whether our product design will be high quality or not.

It is only when we learn to manage our time and intuitively understand that the level of our commitment should be properly dosed that we can go through the design process unscathed and stop getting design-overwhelmed by the complexity of large products.

Instant design and microdecisions

Making the aforementioned big decisions and choices is part of a designer's job. Nonetheless, there is another set that is easiest for me to categorize as microdecisions. These are the little moments when we make a design choice, sometimes without even realizing it. Our daily lives are filled with questions to which choosing one of the possible routes is the answer. In turn, how well and accurately we make these choices will assess the quality of our work.

As a passionate chess player, I was unable to ignore the comparison I found last year in an industry article that discussed the trend of modern product design. Despite the name sounding Japanese, Hikaru Nakamura is an American grandmaster of chess who, as his name suggests, has mastered the rules very well. His skillset perfectly resonates with the world of design, but for better understanding, let's stay on the chessboard for a while.

A traditional game of chess entails a slow process of analyzing variants and predicting about a dozen moves forward, for which players have plenty of time (games can last up to 8 hours). While Nakamura is a good player, there are a dozen more who would rank higher at this distance. When we move on to the second variant of the game, which is commonly referred to as blitz or blitz chess, things start to change.

Instant games require players to condense all of the processes that take place within their heads into a few minutes. Despite the same level of complexity of items, the decision-making process differs. There is no time for long calculations. Currently, what counts is a quick assessment of the situation, instant choices and, above all, intuition.

The shorter the time taken to make a decision, the higher the ranking will place Nakamura. When conditions are more restrictive, his approach to solving chess problems becomes more unrivaled. You can see that the final outcome of a game is not determined solely by the skill of the players. A crucial aspect of this game is the circumstances and framework that accompany it and impose certain limitations.

What does all of this have to do with design? Here's how: both of these areas rely on intuition. When time is relentlessly passing, this skill proves to be decisive. The process of designing large products is almost always driven by a deadline and the need to make quick decisions, whether they are large or, more often, small.

It is relatively easy to master the rules governing how experiences and interactions are designed. There is too much material available for learning the basics of design due to the increasingly progressive democratization of knowledge. Only when this knowledge is put into practice can one realize that it is not knowledge that is most important.

A good designer is not one who just knows how to fix problems. A good designer is one who has the right answer the quickest. Usually, calm problem-solving is reserved only for wishful thinking.

Every day, we designers make dozens of micro-decisions. When we do so, we trust our instincts unwaveringly and do not wait for data to support them. Using the visual aspect of interface design is the easiest way to illustrate this. We don't think so deeply about choosing one color or one style over another, increasing negative space, or changing form or shape. We just make a decision.

The same applies to designing a comprehensive user experience. Whether the confirmation method is a new screen or a modal window, or the dropdown is changed to a radio button, our design intuition plays a role in all of these decisions.

What is the best way to achieve this level? And how do we develop an appropriate "intuition"? First of all, the quotation marks here are not by accident. We are not talking about some supernatural, extra-conscious ability here. It is rather a down-to-earth and methodical set of experiences and observations that help us make the right decisions despite a lack of data. It is worth emphasizing this lack of data. In an ideal world, our decisions would be based on numbers, quantitative research, analytics, or any other type of stats about the product and its users. What if we need to make a design decision, but we don't have any such numbers? Here we come back to the intuition placed in quotation marks.

The experience we gain from working with real products is one of the most important values that will help us build it. During this process, we learn to identify and catalog the problems that are frequently associated with the design process. It is in this phase that we build a set of solutions to alleviate the problem or provide a quick fix. If the decision once again turns out to be the right one, it will reinforce our professional self-confidence.

Knowledge of trends, product sense, the ability to analyze your own mistakes, have a keen eye for details, and a sense of observation are some of the traits that ultimately build up a professional instinct.

It is such a set of experiences that allows us to remember the value of story points for the solution we suggested and which was priced by the development team. Thus, in the next project, we would either repeat them (if it were possible to do so quickly) or already know that we needed to look for something else, since this method of solving the problem was negative-verified during the estimation stage.

Time is money. If you are a freelancer, you can take the previous sentence literally. The faster you can solve a problem, the more money you will earn.

Robot holding a pen

Design systems are the present of designing large digital products. However, the question arises: what next? Where will this still quiet and creeping revolution lead us?

Well, there are many indications that before we know it, the product design environment will be going through a transformation that the industry has not yet experienced. The reason is simple - using the possibilities of automating the design process can bring clear benefits that can be measured by opening your wallet and looking at your watch. And if something is cheaper and faster, while at the same time not reducing the quality of the product, it will not take long to become a standard in the manufacturing process. Teams that miss this moment will stay stranded on the platform watching the lights of a departing train.

Alright, but what form of automation are we talking about? Can we start to think about it realistically today, or are we still stuck in the stage of waiting for progress in the development of AI in our industry? The answer, as befits the degree of complexity of the issue, must be complex.

Let's move overseas for a moment. Elon Musk, a man of many businesses, once headed a project called OpenAI. The name of the project explains quite well what this research studio does on a daily basis. It creates and develops a generally available (well, almost) machine learning language, GPT-3, the main distinguishing feature of which is distinguishing the context between words and processing a huge amount of data for this purpose. As a result, this algorithm is able to create works that may prove indistinguishable from those created by the human hand.

Sounds like science fiction? Well, if you think so, I encourage you to read on. To bring the topic down to earth, let's look at an industry that exhibits some set of similarities to design. Here we are able to distinguish between a range of tasks that can only currently be performed by humans and those that could be automated, since their complexity is so small that today they are performed by interns in their first week on the job. I'm talking about journalism. There is no doubt that gaining social trust and opinion-forming are those aspects of a journalist's work that cannot be reduced to an algorithm. What matters here is a person of flesh and blood. But what about a whole range of other content, the creation of which is mainly based on proper research and the ability to reformat data into live text?

In September 2020, an article appeared in The Guardian that began with the sentence "I'm not a human. I'm a robot." Then the rest of the article was just normal. The content was in no way different from what would have been written by a human being, perhaps apart from the fact that the text described the philosophical dilemmas and thoughts of the robot author.

That same year, Liam Porr, a student at the University of California, conducted a small experiment. He used GPT-3 to create a microblog. He provided titles that the algorithm supplemented with full content, and they then found their way to well-known American Internet forums. In one of them (Hacker News), Liam's (robot) article topped the most popular content, gaining widespread recognition by the community in the form of likes and subscriptions. Nobody took into account that the content came from an artificial pen that selected words using a clever algorithm.

This is what it looks like in journalism today. Is the design industry so different from writing articles? Will we not find among our tasks those that could be easily replaced by using machine learning? Are you sure we need a human to draw a product screen based on an already existing set of views?

If you are interested in what the landscape might look like in 3 years' time, I recommend entering two words into a YouTube search: ”GPT-3” and “Figma”. You will come across a plug-in prototype that is based on a simple dialog box into which we write our intentions. For example, "category list screen divided into two columns and a place for an advertising banner". The plugin connects to our design system, detecting a set of components and patterns that match the query, and then creates a ready view that can be prepared for implementation after small modifications. Today it's a prototype using the non-public beta version of GPT-3, but it doesn't take much effort to imagine how the concept will evolve in the next few years.

An intriguing and frightening summary of this thread can be found in a press release from August 2020, in which Microsoft announced that almost 100 editorial jobs in the US and UK will be reduced and the writing of their content would now be dealt with by algorithms. Well, welcome to the future.

Enemy at the Gates

The revolution is already in the vestibule, just waiting to break down the door. For now, it only subtly suggests that it is time to stop devoting precious human time to activities that are not as complex as our energy. However, it may soon turn out that the portion of the cake being given to algorithms will be much larger.

However, a fundamental question arises - is this wrong?

There are many indications that no, it is not wrong at all, at least not for everyone. It will be most beneficial for digital products, the production costs of which will become lower, thus increasing the potential for development and innovation. Because isn't that what it is all about in the end? That our work evolves towards increasing the quality and complexity of the solutions we design. Optimizing and automating the repetitive parts of our work will free up space and time to look for new ways to make life easier for our users and translate into the value of our products. And it is only up to us whether or not we are ready to go along with this revolution towards the future of digital product design.