Creative automation isn’t just for self-service in consumer goods, or mechanical pickers in warehouses, or even sophisticated decision-making robots hard at work on product assembly lines. Robotics, machine learning, and artificial intelligence are now accelerating into creative workspaces, with a footprint that is growing rapidly over time.
When one mentions automation today, it might strike fear into the hearts of creatives and design professionals.
Am I being replaced?
They worry that they could face replacement by a faceless algorithm. But this fear is misplaced: automation’s influence on design applications is potentially highly beneficial to many creative professionals, enabling them to focus on what they do best. Relieved from mundane and technical grunt work by automation, they are free to focus their talents on creative work, designing what is new and revolutionary.
Design software is itself a form of creative automation: nearly all designers use software such as Adobe Illustrator, Photoshop, or InDesign, and few perceive this as “automation” or any form of threat: after all, it just is a tool in their hands. Yet when you really think about it, the tool is making design decisions under the hood. Who is kerning the fonts, or managing ligatures? The software.
High-end design tools like InDesign include complex algorithms for things like optical kerning, letting an algorithm do what was once the art of a typesetter. And they can be extended.
The extensibility of desktop design applications
Adobe InDesign is one of the most extensible programs ever created. This wasn’t by accident: great automation options were built into the application’s core architecture from its inception. As our co-founder Max Dunn put it, InDesign is one of the greatest programs in the world for good reason. Its capabilities make it suitable for use across a wide range of domains, where its famous output quality has long been its calling card.
Max has led several automation projects since the 1990s, a time when the most powerful computers offered less power than today’s smartphones do. When the challenge was to create provider directories for the insurance industry, finding a way to automate this data-driven process efficiently and flexibly was a daunting task. At that time desktop software did not perform well, nor was it easily extensible. Server-based tools such as Xyvision, 3B2, or Miles33 offered decent output, with fine-grained control, but required “operators” with a programming mindset to set up jobs.
In those days pre-InDesign, software aspired to automate the recurring production of relatively inflexible document types. These included financial statements, real estate marketing collateral, product catalogs, phonebooks, and more. The advent of InDesign proved to be a key development in the march towards server-based print automation tools that offered greater flexibility while tackling document challenges at an industrial level. The few potentially suitable solutions then on the market were largely unaffordable for startups, but it soon became clear to Max that InDesign outran similar products like FrameMaker, PageMaker, and QuarkXPress, none of which offered its combination of flexibility, power and quality of typographical/graphic output.
FrameMaker was more flow-centric and used a simple scripting language, which made it ideal for long documents with consistent formatting such as that found in directories. QuarkXPress had stronger graphics capabilities with a more page-centric model. But InDesign was different. It proved to be a much more extensible product, thanks to its easy automation via simple scripting, and soon (especially once the server version arrived) the sky was the limit when it came to process automation. Competing programs, with their hard-coded UIs and platform limitations (such as the Mac-only nature of QuarkXPress), made InDesign a wide-open and vastly more extensible option. This extreme versatility remains true today.
Automating desktop applications to offload tedious operations without sacrificing creativity
This aspect includes things like rendering multiple PDF outputs with varying job options tailored to specific printers. Automation also expedites tasks like outputting series of different resolutions of image files for different media, a use case often found in sales collateral.
But sophisticated creative automation doesn’t translate into sacrificing design aesthetics or data-driven flexibility. Automated cruise booklet production for Royal Caribbean Cruise Lines is an excellent example of this, where Silicon Publishing really pushed InDesign Server to its limits.
Personalized Cruise Booklets
Their challenge centered on Royal Caribbean’s requirement to produce about 10,000 cruise booklets per day. These booklets weren’t really uniform beyond shared elements of layout and design. Instead, they use relational data including each passenger’s name, itinerary, and other information that required more personalization. This data was also flexible, being subject to late-breaking updates on the customer’s end, like a name change, or changes from the cruise line’s side, such a revised date for a given port or call. The end result was complete automation of personalized documents, each of which looked hand-crafted.
Designers were certainly involved, every step of the way. They built the templates, established the object styles, the layouts, the rules for how text flowed and copy fit, and generally fully defined the output. But they did so by defining design rules and pagination logic for thousands of unique output documents rather than toiling over each individual document one by one.
Instant Celtic Art
InDesign offers virtually endless automation possibilities, and there is certainly no need to limit it to use for commercial purposes. Instead, its scriptable automation is now enabling truly creative pathways. One example of this is Olav Martin Kvern’s free script that generates beautiful Celtic knot patterns. Such automated art could be used for anything from coloring books to video games, or even simply beautifying your desktop. While it generates truly beautiful creative output, there is still a creative hand involved.
Automation scales to your needs
The desktop and server versions of InDesign are virtually identical, with the only real difference being the automation that is usually reserved to the server variant, in support of large scale (and often lights-out) document production runs.
Creative automation is also important when facing the challenges of selling an array of products through a network comprised of multiple distributors, each of which need their own uniquely-branded version of a monthly catalog. Automation supports varying and seasonal levels of throughput, driven by information and trends. Whether prices or product availability change, distributors come or go, or other drastic changes impact the company’s business model, server-based automation offers the flexibility and power to adapt more efficiently than desktop-based options.
Finally, creative automation gives designers more time to try new ideas, research what the competition is doing, and attend to other important tasks. It can turn a daunting process into a “set it and forget it” situation, or one that requires only minor post-processing.
Artificial intelligence powers hybrid workflows, and so much more
Automating tedious design tasks is only a small part of the package, as artificial intelligence isn’t solely the realm of governments or leading-edge game developers anymore: it’s advancing the state of the art in design tools, as well. Creative automation also opens up ways to save time, money and energy, while enabling greater human productivity and innovation, which are among the most valuable elements of any workflow.
Adobe Sensei is an AI engine hosted in the cloud and it interacts with desktop programs like Photoshop. Sensei has a host of new filters still in beta stage enabling AI to intelligently recognize and manipulate the way image subjects appear and interact with neighboring content. For instance, the Neural Filters include options to change a person’s apparent age, or modify facial expressions. Image cleanup, like removal of dust and scratches, and colorization of monochrome images can also be done by AI rather than relying upon human-based skills or working within 9-to-5 time constraints.
What separates AI from machine learning is that these changes are meant to simulate an experienced and intelligent human designer, whereas machine learning would attempt to predict what the designer might do next, based on historical actions. While this prospect can seem a bit frightening in terms of both capabilities and job prospects for technically-inclined creative professionals, it actually opens up a whole new world to them.
Instead of editing, that time could be spent focusing on the creative aspects of design work. If you’re pleased with the AI’s results, or only need to perform minor tweaks after the fact, Sensei and other AI-powered tools in the pipeline could prove to be massive time-savers.
So, where is creative automation headed? Human designers aren’t going to suffer the same fate as some manufacturing and retail jobs where it may be desirable to replace a human with an algorithm. Brands still need the human touch that only a real designer can offer. Beyond saving time, automation will open career opportunities leading into more creative roles that focus primarily on design direction and advancement, as such visionary work is unlikely to be delegated to AI any time soon.