AI and Design: Ingenium Ex Machina
We recently spoke to FastCo Design about how design jobs would evolve as technologies like AI, VR and automation enter the mainstream. The article included excerpts from Rob Girling’s exploration on how Artificial Intelligence will impact design.
Artificial Intelligence, a friend or a foe? Would it replace or augment human capabilities? In the last few months, I have been thinking a lot about how AI will impact our future as individuals, as designers and society. A very simple way to examine if AI algorithms will displace designers (or any profession) is to understand the degree to which that occupation can be easily codified or reduced to a set of reliable patterns, steps or models. Generally speaking, the more repetitive, and pattern-like your occupation is, the higher the probability that it would be replaced in the next 10–15 years by a robot or, more likely, software AI. I’m convinced most of the design profession is safe from computerized job displacement for decades.
The fundamental nature of design is applied creativity. So if we define creativity as the ability to create ideas or artifacts that are novel and valuable, then I think most would agree that this ability is fundamentally hard to codify. In addition, humans determine value in a multiple, different ways that vary across individuals, cultures and societies. As a result, many arguments about creativity are rooted in disagreements about value. When you add to that the common design skills of empathy (social perceptiveness), the need for persuasion and negotiation skills (interdisciplinary collaboration) and the subjective skills of being able to create desirability and beauty — I’m convinced most of the design profession is safe from computerized job displacement for decades. Furthermore, design jobs that employ all of the above skills are likely to grow as a proportion of design jobs overall. While humans will still be very much in charge of creative selection, the tools will massively amplify a designer’s ability to experiment, explore and refine, at a very high fidelity.
The novel component to creativity, especially in the visual realm is something that algorithms are not only very well suited to yet. But they are getting increasingly better. In the next 10 years, all visual design jobs will start to be augmented by algorithmic visual approaches. Architects, interior designers, industrial designers, fashion designers, web designers, art directors will be presented with tools that will be able to create hundreds of novel design variations, based on simple high level requirements and inputs. All an AI-powered tool will need is a high level template, or style definition. I will be able to task it with “a craftsman style, 2500 sq. ft., three-bedroom home, optimized for maximum interior sunlight on this location” and get a dozen of designs on the spot. Early versions of these algorithmic procedurally generated tools are already in use by game designers.
As we saw 30 years ago with the arrival of desktop publishing — the democratization of design through AI-enabled tools will make design far more accessible to semi-professionals and amateurs. Much of the production ‘craft’ of our discipline will be codified into patterns, with computers optimizing the execution based on context. Unfortunately, this means that the demand for professionals that focus exclusively on the craft and the execution will decrease. On the plus side though, and this is what I am looking forward to, this also means an unprecedented, massively amplified ability to experiment. And as any designer will tell you —we crave and thrive on experimentation.
What would make some professions less susceptible to automation and will design be one of these? Read on here.