Design X AI - 06

CHOICE ARCHITECTURE

A heavily debated topic surrounding generative AI is whether it actually helps designers in the creative process. While I do not know the answer to this question (if there even is a right answer), it is important to consider the set of circumstances being evaluated. The most prolifically advertised use of generative AI in the creative process is its ability to generate countless iterations of semi-polished concepts as a point of “inspiration.” There is little to no cost involved in doing this, so the underlying thinking is that there is no harm in throwing a bunch of stuff on the wall from Midjourney or ChatGPT to see if anything sticks. If nothing does, well, nothing is lost. The costs in this process are, like many issues with AI, more insidious. 

For non-designers, the gratification of creating a shiny new render from their prompt feels as if they have become empowered to be an effective designer themselves. Without the knowledge or value of the human-centered design process, these experiences lead someone with a vague idea straight to the tippity top of the Dunning-Kruger peak. They may become emboldened, smirking with their arms crossed, backed by half-baked AI-generated content that could never be manufactured, fulfills no meaningful criteria, solves no particular problem, and looks aesthetically unrefined.

Designers have long had to advocate for themselves and for the importance of the design process to achieve measurably successful outcomes. While many executives and business leaders have varied in their skepticism of the process, they have had little other choice but to trust their design team. Now, with their hunger to use AI to cut costs and delivery time, that ten-person design team and human-centered double-diamond design process is looking prime for the chopping block.

For the designers themselves, does having fifty AI-generated images actually spur more inspiration and creative output? The jury is not out yet, and it surely varies between individuals, teams, and projects, but leading research from Cornell University has shown that exposure to AI-generated content early in the design process reduces divergent thinking.This is where choice architecture comes into play. Choice architecture is defined as “the way in which options and information are presented to influence decision-making without limiting freedom of choice.” There is no doubt that having polished imagery and aesthetic visual presentation influences opinions of desirability, viability, and feasibility in the design process. It is a big part of why we as designers make our slide decks, renders, and personal branding look so good: because it works to create buy-in and excitement around a concept.

Even though designers use these strategies to communicate ideas and create buy-in, we are also susceptible to them ourselves, perhaps even more so than the average stakeholder. So when I prompt generative AI to create polished “inspiration” imagery for my project, using my design brief, it becomes a lot easier to adopt the shiny outcome as my own idea. The illusion of an easy path to get to a thing blinds us from the non-linear path required to get to the right thing.

———————————————

DESIGN X AI is an 8 part blog series written 100% by my fallible human brain. I’m a designer, not an AI expert.