Data is important in the modern world. There is no doubt that it adds to our intelligence and helps us design our products and businesses better. But there is a cause of concern. Will artificial intelligence, which is way better at organizing, analyzing, and interpreting data in fruitful ways, replace designers by taking control over matters of design? Will design become too AI-centric before we even know it? This article makes two arguments pertaining to these questions. First, that design is ultimately always for human beings, and therefore, must always focus on human interests. Second, AI is indeed more effective in working with data. We must, hence, facilitate AI instead of fearing it.
Last year, Adobe demonstrated the use of AI in its tools by introducing ‘Sensei’ to the world. Sensei does wonderful things like scanning rough sketches made by designers and finding photos that match what is depicted or sorting faces according to the direction in which they are looking. The objective of including Sensei in Adobe’s software, the spokesperson for the company said, was to let creative minds focus on what they like doing most rather than make them spend time on technicalities. The speaker was right. In today’s day and age, with the technology available to us, it is not necessary for designers to waste time and energy on menial tasks. Those tasks should be automated so that designers don’t have to break their creative flow and can spend more time and energy being productive. It is for this purpose, that AI must be used – to take care of fundamentally human interests.
Designing AI for Human Interests
- Understanding the Need: Economics works predominantly according to demand and supply. Which of these is more important is difficult to say. Could supply sustain itself without demand? Alternately, would demand even exist if its not recognized and catered to? The task in the digital data-driven age is precisely to determine demands accurately so that they can be catered to in effective and profitable ways. AI makes it possible to collect and interpret data in real time, and to adapt and produce product experiences that are in-line with user demands. For this, however, it is crucial to understand the need.
- Performance Imperative: Once the need has been established, designers can focus on fulfilling it. In this step, it is critical that the need be translated to a performance imperative – what must be delivered as a feature or as an experience of the product that targets and satisfies the need comprehensively. If this imperative is met, then users are bound to have a positive experience of the product, and to use it more often to pursue their own interests.
- Design Solution: We know what the user needs. We know what function our product must have to meet the performance imperative. It is time then to DESIGN! The most effective design solutions are always functional. Products must have a use if they are to be used. But apart from the function that a product performs, its sustainability is also important. In a data-driven world this means the design must also incorporate sustainable modes of data collection that are not intrusive, but inviting and effective. Additionally, the product can also be aesthetically appealing.
- Evaluation & Improvement: Lastly, designers have plenty of opportunities to collect feedback from users and machines in the form of data that gives crucial clues into the functional effectiveness and aesthetic and performative appeal of their designs. Thorough analysis of these clues is not only useful to improve present design, but also to design better AI for human purposes.
Three Simple Tips
AI is still only developing, and designers are only still learning how it can be used to better design and improve user experience. The key to the door that hides these secrets is data. Hence, the three simple tips to design for AI and human interests are also data-centric.
- Focus on user data: Let’s start with the basic question – data on what? There is so much data available now that choosing what matters is a tough task. It is important, therefore, to set the context right and focus on user specific data. What does that mean? If we are to go back to the example of Adobe’s Sensei, then this artificial intelligence was designed to eliminate technical hassles – a very specific user need. This need is both general and specific at the same time. On the one hand, all professionals want to automate menial tasks. On the other hand, designers, specifically, want images that match their sketches. This insight can only be developed by profiling users on the product effectively and understanding their pain points while using it. There are many impediments to design. But data that reveals user needs in terms of product functionality for improved performance should be top priority.
- Mutually Beneficial Relationships: There has never been a time, when designers are so reliant on users for data. I mean, designers always liked feedback, no doubt! But, now they must pay attention to all feedback – even the ones that they do not like – if they are to be good designers. If they don’t, their competition will, and their design will falter. One could argue that the success of design in this scenario is dependent on the mutually beneficial data-relations that designers build with users. The design must not be too intrusive because that might kill user confidence. It must not be too laid back because that might not get enough data. It must elicit curiosity and be effective in establishing trust and loyalty in users.
- Replicate Humans: Artificial intelligence is often associated with the horror dystopian fantasy of a hostile world takeover and domination by machines that lack humanity. It is this myth associated with artificial intelligence that intelligent AI design must dispel. Imagine if artificial intelligence was decidedly human – artificial human intelligence. The simple appearance of ‘human’ in AI is enough to calm our senses and make us feel more comfortable with the concept of machines doing perfectly normal human things, in a perfectly efficient manner in comparison to, humans! What we get by replicating humans in artificial intelligence is a sense of emotionality and feeling in machines that they so need for their own sake – to be effective at gaining intelligence – and ours – to feel at ease with machines. Humanness helps AI connect with users better.
Designing AI for UX
The pervasiveness of data in the modern world makes it clear that all human endeavors – especially the ones of commerce – are inevitably linked to effective collection and interpretation of data. Artificial intelligence is also, without doubt, better at working with data. The two statements when read together imply that businesses and professionals cannot do without either. The question worth answering then is – how do we move forward in designing AI for better user experience?
In this article, we’ve established the process of designing AI to fulfill human needs. It is important to understand the needs of the user, identify performance imperatives of the product in-design, deliver solutions that meet needs, and to improve continuously with feedback for better user experience. It is important to stay focused on user data to build mutually beneficial relations between users and AI. Lastly, it is important to make AI more human-like to make machines seem less intimidating and bring them closer to human users.