Craft Is Dead, Long Live Craft | Art & Science Insights

For most of the entire history of art, we have relied on the notion that you must first master craft before you can be an artist.

And what kind of art we produce, has innately been tied to the technologies of the time available to artists (yes, a paintbrush is a technology, as is a pencil, a spray can, and of course a camera and Photoshop).

Five hundred years ago, you learned your craft as an apprentice to a master; someone who had themselves spent years honing their skills to be the best in the land.

The historical belief was that you needed to learn the fundamentals—the craft—before you could make art. I would argue the same holds in most if not all creative-class professions from biologists to engineers to lawyers and graphic designers. There is a foundation of knowledge required for any individual to be competent within their domain.

The first year in art school is literally referred to as the “foundation year” and was (at least historically) required as a prerequisite before a student could determine which artistic discipline they wanted to pursue.

But this article isn’t about Art. It’s about craft. And design. And what AI is doing to a designer or creative’s relationship to craft.

The moment

The other day one of my designers posted an image (a piece of stock) that wasn’t the right framing for an ad he had to make. He needed more than the frame of the photograph afforded. He asked AI to generate more — to show him the image beyond the frame. And it did. In seconds.

The results weren’t perfect, but they were good enough.

In reflecting on this I was tossed back into the late nineties when I first learned Photoshop (in art school). Many of Photoshop’s “innovations” in the last 20+ years have related to being able to “fix” photos — to touch them up or mask something out, to generate just a bit more of a background, etc. This piece of software, like many, now has so many features that most users, even the pros, only use a fraction of them. But the ramp-up time to learn to use the software is extensive. It has its own interface paradigms to master, there are features by the dozen, and key concepts to understand before you can even begin to be considered competent at using it as a tool.

Just like a hammer in the hand of an untrained person, Photoshop itself has no value. But put a hammer in the hand of a skilled carpenter and you can get a house built. Photoshop in the hands of a skilled illustrator, designer or artist is powerful. But Photoshop on its own… is nothing.

Is AI a hammer?

What my designer did with AI (to address a technical limitation) in a couple of minutes would have taken hours at best, and doing it manually may very well not even have been adequate, and certainly wouldn’t have been cost-effective (in all knowledge work time is money).

So what is the implication here?

It removed the need for my designer to know his craft—or rather what we typically considered his craft to be. He knows Photoshop, but will he ever need to use it in its historical context and application again? With every passing day this answer trends closer to “no”.

AI-enhanced Photoshop is a better hammer. It’s a hammer that barely needs you to lift a finger. But like all tools, without vision, it’s just a freakin’ hammer!

New tools, new craft

The media and press at large like to create anxiety around the impact and implications of new technologies. This has been happening since the dawn of tech, and I don’t just mean the internet-based era we’re living through—the invention of the washing machines was also met with “but what about the people who currently scrub their hands raw doing this manually?” Technology is a tool. A tool extends and augments human capability.

Photoshop (and other digital tools) changed photography, film making and art in general. These new tools required skills that hadn’t been needed prior. But our ability to adapt to these new tools enabled all kinds of artistic expressions and commercial applications.

AI-enhanced Photoshop and other generative applications of AI are having the same consequences. I may never again need to hone my “cloning” skills in Photoshop to clean up an old photograph, or remove an ill-placed object from an image. I won’t need a mouse and super fine motor skills to get a great output from a tool like this.

What I will need, however, is to know how to “talk” to these tools.

As any contractor or skilled labourer can tell you, there’s an ineffective way to hold a hammer. And there’s an ineffective way to interact with AI.

The new skillset

What I think we’re starting to see emerge is the delineation between those who know how to “speak” to the AI, vs those who don’t.

The skill that all future designers are going to need to master is vision and concepting, and then the “craft” will be in how they are able to materialize those ideas into reality by learning how to ask for them to get the best output from our AI tools.

To those worried about the machines taking your job… if your job is entirely based on craft and not the idea behind your efforts, then your concern is probably valid.

But to the young creatives in the room who are full of ideas and optimism—rejoice! The timeline and investment to “hone your craft” just became much much shorter and smaller, respectively, provided you learn to speak AI.