Design & Artificial Intelligence: Complementary Approaches to Expand Human Intelligence
In the last ten years, design thinking and artificial intelligence (AI) have become exceedingly popular tools in the business world. However, to most, their impacts have been seen as entirely separate. AI belongs to tech companies. Design is for creatives.
This is a problem precisely because design thinking and AI are actually complementary tools to enhance human potential for innovation. What might happen to the design consultancy when it becomes supercharged with machine learning? What might happen to the tech company when it leverages the creative, user-centered ideation of the design process?
It is important to see that design thinking and AI leverage two important ends of the knowledge spectrum: the generalist and the specialist.
Design thinking celebrates the power of sheer generalist thinking and collaboration to generate new, and novel ideas.
Artificial Intelligence leverages sheer specialized, computational power to execute on defined tasks.
In this article, we will discuss how design thinking and artificial intelligence complement each other for innovation by:
- Exploring the shared history of design thinking and AI
- Examining both design thinking and AI through the lens of expanding human intelligence
- Now What? What combined AI — design thinking might be able to do
A History of Design Thinking and Artificial Intelligence
AI became an academic field in 1956 at a conference held at Dartmouth College. Although this was the first recognized scientific gathering on the subject, research in artificial intelligence had been occurring for years, dating even before Robert Turing’s 1950 speculative report on the potential for creating machines that could think.
Post-1956, the field progressed through a series of booms and busts, influenced by scientific progress (or lack-there-of), hardware advances, and economic backing. It was only until the early 90s that computer hardware and machine learning had developed enough for early adopters in industry to begin using machine learning algorithms for various tasks.
In 1997, the IBM supercomputer — Deep Blue — defeated the world champion in chess — Gary Kasparov — in a six game head-to-head match.
Since then, artificial intelligence has spread to nearly every industry, leveraging the age of big data, Internet of Things and improved hardware. AI is being taught at University programs all around the country.
Similar to AI, the history of design thinking also began in 1956, with Buckminster Fuller’s Comprehensive Anticipatory Design Science curriculum at MIT’s Creative Engineering Laboratory. The lab aimed to bring the expertise of engineers, industrial designers, material scientists, and chemists to create innovation.
While Fuller’s curriculum stressed a design process for scientific fields, the idea of design in a meta-sense (as an idea or open-framework) was not conceived until 1969. One of the first publications on “design,” Herbert Simon’s The Sciences of the Artificial, was an artificial intelligence book that was attempting to crack design in order for machines to solve problems. While I cannot argue that AI created design, we can clearly see that the development of both these fields were inextricably tied at their origins.
By the 1980s, systems engineering design methods had been widely established, particularly in Germany and Japan.
In 1991, Dave Kelly founded IDEO, the most famous design consultancy there is. IDEO famously introduced the first and most widely used “design process.”
By 2001 (similar to the boom in AI), design thinking had generalized from a field dedicated to only product design, to a field that could be applied to anything (business, biology, etc.).
By 2005, universities around the country (starting with Stanford) begin to teach design thinking as a process for problem solving.
Design Thinking and Artificial Intelligence as a Means of Expanding Human Intelligence
While design thinking and AI have clearly developed pairwise in their histories, their practical counterpoint doesn’t seem obvious at first look. In order for this to make sense, we need to look at each in terms of what type of intelligence each leverages. Design thinking leverages generalist ideation. Artificial intelligence leverages specialized computational crunching.
Artificial Intelligence is a broad field. Today, the most widely utilized part of AI is machine learning, which involves training an algorithm with data to be able to classify information (such as if a cancerous tumor is benign or malignant) or to predict trends in data (such as what the stocks might look like in a year’s time). Feed an algorithm specific data with a specific answer in mind, and the machine will do a superhuman job at providing accurate responses. In fact, a longitudinal study by psychologist William Grove proved that statistical (or machine learning) models almost universally outperform or, at the very least, match human experts at defined tasks. For example, the 2017 Annual Report of aiindex demonstrates that, within the last five years, machine learning models have exceeded humans at written-text recognition and audio transcription, a seemingly impossible task ten years ago.
While the current state-of-the-art for artificial intelligence is machine learning, the holy grail of AI is achieving general artificial intelligence. General AI might be defined as machine intelligence that could successfully perform anything that a human can. Humans can learn new skills and adapt to new situations without needing to be taught, unlike modern machine learning algorithms, which must be trained on specific tasks. The idea of creating general artificial intelligence opens up an entirely new realm of possibilities, which is beyond the scope of this article. Furthermore, by any expert’s opinion, general artificial intelligence is a long way off, which is where design thinking enters.
Similar to machine learning, design thinking, as a tool, has outperformed standard human practice since its adoption into the business world. In fact, in 2015, the Design-Value Index reported that “design-driven companies have outperformed the S&P index by 219% over [the last] 10 years.” 
Unlike machine learning, which leverages computational power on specific tasks, design thinking excels at generalist thinking. In short, design thinking is characterized by a group of multidisciplinary human experts (a conglomerate of generalist thinking) conducting user-focused research, wild idea generation, and solution iteration. In other words, rather than one human or a group of similar human generalists working together, a group of very different generalists come together in a creative process to assemble a super-generalist intelligence. The very nature of design thinking leverages humans affinity towards generalism and brings out the best in it.
Where machine learning generally surpasses humans at task execution (as long as the task is defined enough), the design process generally surpasses humans at idea generation. Similarly, specialists are equipped to tackle specific problems, generalists are equipped to bring foreign ideas together into new solutions.
The following table further attempts to illustrate the differences between these two tools.
The Combined Power of Design Thinking and Artificial Intelligence
What might happen when design thinking and AI are leveraged off each other?
One idea, which has been discussed on the web is design thinking for artificial intelligence. Use human-powered design thinking to generate new, user-centered applications for machine learning and define better parameters needed for the machine to execute upon. Deloitte recently published an opinion piece, stating that AI will not reach its potential unless good designers guide the algorithm execution .
New Creative Ideas:
As an example, Autodesk has deployed Project DreamCatcher, a generative 3D modelling software, where humans generate constraints and the computer designs to those constraints. Although the following isn’t necessarily user-centered, it provides a concrete example of what a machine generated output might look like when given good constraints from humans.
A better example of user-centered, Google recently launched its People + AI research Initiative, dedicated to creating a user-friendly approach to AI . The initiative researches better software to interpret algorithms, open-sourcing for non-scientific users to access algorithms, and many other applications.
Better Design Research:
Machine Learning has increasingly surpassed humans at detecting and predicting off of data.
Machine learning offers designers an expanded toolkit to research new problems. For example, rather than interviewing users on what they think of public parking, designers could scour the internet using text and speech recognition algorithms to automatically render user opinions in various regions. Rather than gathering information from a small pool of people to intuitively guess whether someone would want a parking spot at 3 pm in Toronto, a machine learning algorithm could predict much more accurately and sample a much larger population.
So what might all of this culminate to? Historically, design thinking and AI have developed together, yet have been seen as separate. At this stage in the development of artificial intelligence, the two need to come together to enhance humans overall expanded intelligence. Tech companies leveraging machine learning need design thinking as it will vastly improve their algorithms. Design-based companies need to integrate machine learning into their work. Doing so will unleash a new wave of better research and new creative potential. The design consultancy needs to change. The data consultancy needs to change. Doing so will lead to a more innovative future.
About the Author:
Macklin Fluehr is an engineer, machine learning specialist, and designer aspiring to transform sustainability with a cross-disciplinary approach.
 Smith et al. The History of Artificial Intelligence. University of Washington. December 2006. Web. https://courses.cs.washington.edu/courses/csep590/06au/projects/history-ai.pdf
 Anderson, Robert. Twenty Years on from Deep Blue vs Kasparov: how a chess match started the big data revolution. Theconversation.com. 11 May 2017. Web. http://theconversation.com/twenty-years-on-from-deep-blue-vs-kasparov-how-a-chess-match-started-the-big-data-revolution-76882
 Brown, Tim. Design Thinking. Harvard Business Review. June 2008. Reprint R0806E. Web. http://sitn.hms.harvard.edu/flash/2017/history-artificial-intelligence/
 Van Tyne, Sean. Design Thinking: A Brief History. Seanvantyne.com. 12 February 2017. Web. http://seanvantyne.com/2017/02/12/design-thinking-brief-history/
 LeGassick, Calvin. Artificial Intelligence Index: 2017 Annual Report. Aiindex.org. Web. http://cdn.aiindex.org/2017-report.pdf
 McAfee & Brynjolfsson. Machine, Platform, Crowd. London: 2017. Print. Pg 41.
 Adobe. Your Business has a New Secret Weapon: How you can use design as a competitive advantage. Adobe.com. 2015. Web.
 Perez, Carlos. The Alien Style of Deep Learning Generative Design. Medium.com. 24 December 2016. Web.
 Schwab, Katherine. Google’s Rules for Designers Working with AI. fastcodesign.com. 11 July 2017. Web. https://www.fastcodesign.com/90132700/googles-rules-for-designing-ai-that-isnt-evil
 Deloitte. Why AI Needs a Dose of Design Thinking. Deloitte/Wall Street Journal. 2 August 2017. Web. http://deloitte.wsj.com/cio/2017/08/02/why-ai-needs-a-dose-of-design-thinking/
 Liikkanen, Lassi A. Applications of Machine Learning for Designers. Smashingmagazine.com. 24 April 2017. Web.