How AI Transforms Design Practices

While people who work in design and design-related roles such as user research and human-computer interaction (HCI) may be aware of Artificial Intelligence(AI), the speed that this technology is being adopted calls for some consideration. About 20% of organizations are considered AI pioneers. Having achieved successes in initial AI experiments, these firms are now rapidly infusing all areas in their enterprise with AI. What does this mean for design practice?

You may have heard that AI automates repetitive, mundane and often time-consuming tasks but it may be harder to see what that means in the context of creative design roles. For people who know a bit more about AI and Machine Learning (a specific form of AI) they may believe that AI could commoditize the practice of design and lead to products that lack creative processes and outcomes. To understand the impacts of AI on design practice, first consider what AI is.  

What exactly is AI?

When I give talks about AI and User Experience(UX) I explain that most of AI boils down to one word, prediction. Prediction means using machines to process data and create new information. With this one concept, it’s possible to frame all other applications of AI. Various forms of a prediction can be; 

  • Classification – predict how to associate data with categories defined by a human
  • Clustering – predict how to group data into categories defined by the machine
  • Recommendations – predict a choice for a user based on assumed or actual preferences 
  • Generative design – using existing content to predict or generate variations of content 

You may have already interacted with a design system such as Wix, an example of artificial design intelligence (ADI) that automates generation of variant webpages. There are numerous examples of AI assistant tools applied to ideation and creation in design. 

What are the benefits of AI? 

AI equips design teams with tools to take the concept of user-centered design, to the most granular level, design for individuals versus designs for usually poorly-segmented groups of users. One example is how Netflix combines user behavior modeling with recommendations to offer each viewer a personalized viewing experience. 

What does this mean for Designers?

Designers need to learn about AI/ML, what it can do and how to strategically apply it to improve user experiences or to improve their design practice. While designers may not need to become data scientists (in the same ways that they don’t need to be coders), they can work to understand the applications, risks and nuances of AI as the first step to benefiting from its power. 

Process Automation: AI enables automation of problem solving tasks such as collecting and analyzing large amounts of data to discover invisible patterns without limitations of volume and speed. With this knowledge, designers can then pivot to focus on higher value tasks.  

Problem Identification: With AI taking on some repetitive tasks, this enables designers to focus more on “sensemaking” — understanding what problems need to be solved. 

Problem Framing: With AI, designers can shift to focus more on problem framing, clarifying the underlying questions and assumptions that the solution needs to respond to. They can also learn to architect frameworks to explore design alternatives and leverage machines to run through alternatives, freeing their time from some of the more mundane aspects of design. 

Human Element: AI is weak in areas that require knowledge of culture, emotion, context and identity, and ethics, to name a few. This is where designers’ impact would be well spent. Designers can have some time freed up from repetitive tasks that pave that way for more time to explore how the problem definition and solution variants could be optimized against these needs. 

Risk Mitigation: AI is a tool that opens the door for designers to create new experiences and to achieve process efficiencies. However, it also raises new risks around security, privacy, fairness and equity. A designer would lean heavily into the work of subject-matter experts in these areas to optimize solutions that consider the full impact of AI on a user’s experience. 

Is there anything to fear with the emergence of AI in design practice and allied roles? According to a report by Oxford University and Deloitte, 800,000 jobs were lost in the UK due to automation between 2001–2015. But according to the same report, 3.5 million new jobs were created. AI is already having an impact on the practice of design, and will likely lead to transformational shifts in how design work is done. 

More information: 

Basu, Ritupriya. (2019, Dec. 6) Algorithms Are a Designer’s New BFF – Here’s Proof. Adobe XD Ideas. https://xd.adobe.com/ideas/principles/emerging-technology/automation-ai-wont-replace-designers/

Irfan, Mirza. Artificial Intelligence and the Future of Web Design. Usability Geek. https://usabilitygeek.com/artificial-intelligence-and-the-future-of-web-design/

Basu, Ritupriya. (2019, Dec. 6) Algorithms Are a Designer’s New BFF – Here’s Proof. Adobe XD Ideas. https://xd.adobe.com/ideas/principles/emerging-technology/automation-ai-wont-replace-designers/

Lant, Karla. (2018) Art by computers: how artificial intelligence will shape the future of design. 99 Designs. https://99designs.com/blog/design-history-movements/artificial-intelligence/

Deep, Akash. (2019, May 15) Real-World Applications of AI in Design. Hackernoon. https://hackernoon.com/real-world-applications-of-ai-in-design-85c3fc728a36

Insall, J.. Borrtharkur, A. From Brawn to Brains: The Impact of Technology on Jobs in the UK. Deloitte Insight. https://www2.deloitte.com/content/dam/Deloitte/uk/Documents/Growth/deloitte-uk-insights-from-brawns-to-brain.pdf

Verganti, R., Vendraminelli, L., Iansiti, M. Working Paper: Design in the Age of AI. 

Białek, Bartek. (2019, March 26) Artificial Intelligence Disrupts UX and Product Design Like No Other Industry. https://www.netguru.com/blog/artificial-intelligence-disrupts-ux-and-product-design-like-no-other-industry

Guszcza, Jim. (2018, January 22) AI Needs Human-Centered Design. Wired Magazine. https://www.wired.com/brandlab/2018/05/ai-needs-human-centered-design/

Koponen, Jarno M. (2019, February 9) UI + AI: Combine user experience design with machine learning to build smarter products. Venture Beat. https://venturebeat.com/2019/02/09/ui-ai-combine-user-experience-design-with-machine-learning-to-build-smarter-products/

Clark, Josh. (2019, Nov 5) AI is Your New Design Material. Presentation from Amuse UX Conference. https://www.youtube.com/watch?v=Tgzu351uDIc

Continuous User Engagement

seek_understand

Why do user research projects have finite starting points and end points when users interact with your product or feature continuously? It’s a rare organization that has the resources or ability to constantly observe their users experiences, however, taking some ques from ethnography, and from some of the best market researchers, it is possible to establish a continuous user engagement effort. In his brilliant book The Tipping Point, Malcolm Gladwell provided a case study of a marketing maven who did just that. Constantly monitor the pulse of users, with a goal of understanding how their needs and interests change, even ever so subtly with time. One practical way to do this is by establishing relationships with several lead users and continuously engaging with them, and this is key – do it in their environment. Not with a problem solving agenda, that could lead you to short-sighted solutions, but to understand the context that drives their needs. Pick one or two lead users and develop a monthly dialogue with them, in their environment as much as possible. This also resonates with one of Stephen Covey’s 7 Habits of Highly Effective People, seek first to understand. For practicing user experience professionals or user researchers, be wary of project sponsors that try to limit or block you access to users. It’s like a doctor being hindered from fully examining their patient.

Trust the Process, for real

post-its

I had two incidents come to mind that prompted me to reflect on the theme “trust the process”. See my previous post on working across different personalities. In one case, several designers I worked with were quick to jump to a UI solution, without thinking through user goals, workflows and exploring several design ideas. Strong personalities, so I caved. I started developing mock-ups of the idea, and the more I worked on them the more apparent this oversight became. In the second incident, I was working with a technology development manager that wanted a solution to something in a few weeks, in a space so complex that it really would be better off going through a more extensive design process. Looking back, I am not clear there was anything could do to influence this second case for a more favorable timeline. For the first case, I decided to go back to basics. User roles, goals, workflows and then sketches of the UI mapped to workflows. After all, Design Is Basic. Note to self, “trust the (UX) process” for real.

Deciding for Decisive

Decisive_heathI’ve been a fan of the Heath Brothers book ‘Made to Stick’ so I was pleased to find that they came out with the book ‘Decisive: How to Make Better Choices in Work and in Life’. I have a small set of books that I plan to re-read every year and this one made the cut.  As a designer, understanding how people decide is critical. Making decisions on design without a process can lead to poor outcomes. Without giving up too much of this book, I have to cite two of my  key take aways; (1) Ask yourself, ‘What would have to be true for an option you are considering to be the right one? ‘ and (2) ‘Multi-track’, don’t limit yourself to a few options, ‘think and not or’. Brilliant book for many reasons. Highly recommended read for anyone interested in the mechanics of decision making.

Lean UX, my take-aways

Capture

Lean UX is the intersection of three principles: Design Thinking, Agile Software Development and Lean Start Up (auths: Jeff Gothelf, Josh Seiden)
I had a number of take-aways, of which i’ll share a few:
1) Cultivate an experimentation mindset
2) Place greater emphasis on ‘making’ versus ‘analysis’
3) Work more collaboratively, think of UX as design facilitators versus design heros
4) Set up infrastructure to enable continuous and collaborative user research
5) Drive for continuous improvement; frame and address your UX debt

Hit the Ground Running

IMG_0113

This was another very actionable insight that I gleaned from the book QE. Whether your research project is a few days, or involves tracking users over several years, it is imperative for researchers to get going. Spending too much time at the beginning limits the time you have to spend in the field. After your plan is developed, give yourself a goal to accomplish 3 key research tasks in your first 3 weeks in the field. One of the tools that may help to organize your research is a user research database. This can take on many formats but I personally have found spreadsheets to be both lightweight and effective in this regard. On a new research project, I would organize a user research workbook (one page per user) with sheets to include things like user characteristics, preferences, dislikes, tasks and any answers to specific questions. Documenting and organizing each user from the beginning is helpful, it’s very painful to do field work and then spend days on end trying to decode your notes well after the fact. As your research progresses, you can fine-tune the database to add more details or organize the details in a way that will help to uncover patterns or simplify the process of aggregating and analyzing your data.

Inside Looking Out

many_flowers

I found two ideas in the third chapter of “Quick Ethnography” to be particularly useful. First, “seek to understand people from the inside looking out”. It can be really easy to use our individual perspective to construct or deconstruct how or why a user perceives a product of feature to be. However, using our perspective instead of the users could lead to very wrong conclusions. This leads to the second important point, “anticipate your best data by appearing a little stupid”. As UX practitioners, we must strive to push aside the expert in us in order to learn from the user expert. This way we can collect the best data to draw accurate conclusions and make effective recommendations on future designs.

Start With Ethnography

quick_ethnographyI recently launched into the book “Quick Ethnography” by W. Penn Handwerker and have been pleasantly surprised at how much the practice of user experience research and design is grounded in traditional ethnographic practice. Practitioners of UX come from many backgrounds and the methods we use more or less stem from this social science. For example, the notion of creating user personas, fundamentally is the outcome of a paired down ethnographic process. I’m hoping to glean some insights about the basic principles of traditional ethnography to better understand and ultimately improve my practice.