Robots will take your design job.

The graphic design process is satisficing, not developing the ideal solution, but the best solution within constraints. What happens to the process if one of the constraints is eliminating the graphic designer?

Every generation the graphic design field changes in substantive ways. I was taught how to make comps with Pantone sheets and rub-down type, and before that pen, ink and rules were the tools of the day. Today comps are digital. It is an iterative evolution of tools. The core of design practice was unaltered. The changes occurring now in and around graphic design are not generational-iterative changes, they are fundamental and revolutionary. These changes affect who practices and what is regarded as graphic design activity. Past changes were mostly internal, new challenges are coming from external sources whose roots can be traced to several sources but coalesce into one idea; the commoditization of graphic design. With commoditization comes reducing or eliminating the role of the professional graphic designer.

A quick side note: The current focus of these innovations(?) is mostly logo design. The easiest artifact to commodify, but it is the first not the only.

Graphic design services are converted into goods in two main ways; DIY and crowdsourcing. One allows for the traditional solution-based graphic design and the other is client decision-based graphic design. One is a rudimentary tool and the other is the false choice engine. One leads to robot graphic designers.

The DIY path may seem subversive to graphic design, but is basically innocuous. A tool is only useful if you know how to hold it. The democratization of design tools began with the proliferation of the personal computer. In theory access to tools is a relatively noble pursuit. Without training though, the things they make are often not good. Access to tools has not been the end of design or most professions. Home Depot is full of hammers yet we still hire carpenters. You still need ideas and training, The tool is not enough. Consider web design and the rise of website builders like Wix and Squarespace. These robust and intuitive tools require some thought, time and consideration and have not ended web design as a profession. The democratization of drawing tools has got gotten more aggressive in targeting graphic design specifically. For example, Assembly app uses the tagline “graphic design for everyone.” It is a rudimentary software drawing tool, a dumbed down Adobe Illustrator. It is fun to use and intuitive, but basic and limited. The tagline is curious. You could attach those words to any drawing tool, a pencil even, and not be incorrect. Regardless of their accessibility and ease of use, tools like Assembly need someone to guide them. They need to be applied, they need an idea.

The newer more insidious path doesn’t require an idea or skill, just an opinion. Crowdsourcing takes advantage of a false choice engine and recontextualizes how people understand graphic design artifacts. (This is not yet another indictment of crowdsourcing, as much as it deserves it, it is a warning that crowdsourcing is just the beginning.) At a site like 99designs, multiple designers submit spec designs for a client to peruse and choose from. I contend it’s a false choice because none of them are particularly good. You the pick the best of a bad lot, and end up with the least worse choice. The only designer who gets paid is the one whose design was chosen. Crowdsourcing is a bad model even if it weren’t unethical. It’s a strategy you use with obstinate children. You reframe the problem so the child gets to make a choice. “Suzie, do you want to wear your red or blue shoes to the store?” You don’t ask if she wants to go, you ask what she wants to wear when she goes. The same psychology is applied here. The consumer applies their taste and judgment with a choice and then that choice is falsely equated with a good solution. This service teaches clients that they are the experts, not the designers. Mulitiple choice is how the design process works in this system and personal taste leads to the best solutions. It is scary in its own right, but how does this lead to robots? I promised you robots.

How could it be more cynical and ruthlessly efficient? Well, what if you undercut crowdsourcing by eliminating the crowd? Services like MarkMaker, do just that. Algorithms and software ‘design’ logos for you. Pick a broad industry, food and beverage for example, and enter your business name and a dozen or so logos are presented to you. They typeset your name alongside a questionably relevant graphic element, a banana for example. You pick the examples you like and are offered refinements based on your preferences. Repeat until you settle on something acceptable. Our friend the false choice engine. It functions the same as crowdsourcing but without the humans on the other end. The current generations of this software produce the worst output, essentially a clipart recombinator. It is at a nascent stage, but it won’t be long before the quality of its output is equal to that of crowdsourcing sites, and we know they are viable business entities. How can crowdsourcing compete with faster and cheaper robots? Once they have gobbled up the crowdsourcing market how much of the ‘real’ design market might they consume?

But surely graphic design is more than logos and letterhead? Yes it is, but, in the new paradigm without the professional, the unbiased outside arbiter of what is appropriate and effective, all bets are off. Why can’t a system peruse a thousand posters and produce an aggregate of the most common elements and use that as a template? As long as the result is reasonable and someone accepts it, it doesn’t need to be good or appropriate.

This encroachment cuts to the heart of what graphic design practice actually is. Can it be distilled down so easily? How much can you offload and outsource, how much of the decision and form making can you erode before graphic design practice is a different thing or no thing? Perhaps graphic design is just a collection of conventions and best practices that can be reengineered with a robust system and a large enough sample size.

How do designers defend themselves? We haven’t been able to muster much defense against the onslaught of crowdsourcing, so I am not overly optimistic. As it has been for the past few generations, our fight is not really with these interlopers but with the public. We are partly culpable as we struggle with the same well-worn issues. For many graphic design is undervalued and is it inaccessible. Undervaluation is a hard nut to crack, but accessibility seems like something we could design our way out of with a few clever ideas.

The last bastion of hope is the idea. Graphic designers are clever and creative and the robots will never be that (probably?). The real question is not whether robots will be clever, it is will they need to be? If people are willing and have been trained to accept the stuff that gets churned out ‘real’ graphic design becomes a niche market like the letterpress is today. “We went all out for the wedding and got invitations designed by a real designer, they did sketches and everything!” I suppose there is a version of this future where the designers are the ones with the best robots and the output is integrated into our process and workflow. Perhaps our future lies in art directing robots. There is also a version where the profession of graphic design is archaic. I am gonna make sure I have enough pencils.


Update: It is already happening: Design Algorithms