AI to Transform Graphic Design

Apple’s 1984 Macintosh revolutionized graphic design; however, it pales in comparison to the forthcoming surge of websites that will generate their own designs

In the past, graphic design necessitated physical labor. You sat at a workstation or a light table to create letterheads, business cards, brochures, magazines, books, and posters. On a printing press, you either assemble metal type or cut and adhere paper. In a dark room, you manually developed 35mm film using chemicals.

Apple’s Macintosh, which was introduced in 1984, revolutionized the industry. Designers could make adjustments with a single click using layout software like Aldus PageMaker and its successors.

The desktop publishing revolution was the transition of graphic design from the workstation to the computer screen. Design work transitioned from the laborious domain of hands-on creativity to the more abstract but liberating digital realm, where the outcomes of decisions are immediately visible. However, each decision is insignificant, as it can be reversed with a single command.

As artificial intelligence and machine learning once again transform the graphic design industry, we are currently on the verge of another revolution.

According to the slogan of one initiative, the objective is to create “websites that generate themselves.” The software will assess your text content, line of business, and imagery, generating completed pages without your intervention.

The initial arrival of these automated tools will be on the web. Still, print design will also transform as design-software manufacturers integrate machine learning into their layout tools and applications.

Nevertheless, the current state of graphic design needs to catch up to the grand vision, despite the considerable discussion surrounding AI-driven design.

The variety of products that are currently available will disappoint users who are hoping for extraordinary results from AI genies. Indeed, this is a disappointment; however, it also allows us to contemplate the types of design tasks we would like machines to perform on our behalf and the responsibilities that should be reserved for humans.

The Grid guarantees that an AI named Molly will be responsible for the design of your website. Molly is eccentric but will only charge you a little, neglect a deadline, or cave into your requests for a more prominent logo. Molly can implement a straightforward five-color palette on your website in over 200,000 configurations.

Since its crowdfunding campaign in 2014, The Grid has been promoting “AI websites that design themselves,” making it one of the earliest entries into the artificial intelligence web design marketplace.

When The Grid’s promotional video was circulated among the web design community, 37Signals/Basecamp founder Jason Fried tweeted, “Conceptually, it feels like a natural, obvious progression that is just waiting to happen.”

The Grid requested $96 from its “founding members” but failed to deliver the product for years, and reviews have been decidedly divided. “The Grid Sucks,” an hour-long complaint by early beta user DigitalDan, will likely be the following video that YouTube queues up for you if you view the company’s video today.

While Molly may represent an AI innovation, her primary responsibilities include the automatic cropping of photographs and the generation of color palettes.

Reddit is replete with dissatisfied users who assert that The Grid’s AI capabilities, which are frequently lauded, result in websites that need to be more customizable or more accessible to replicate. Other reviewers are similarly critical (“shoddy and expensive,” according to CMSWire).

I was directed to a “The Five Minute Website on The Grid” demo video by the automated email replies, which were signed “Love, The Grid.”

The company’s contact form is the sole method of getting in touch. (The duration of the video is 56 minutes.) The demo’s software is similar to WordPress, Squarespace, Weebly, and Wix, content management systems that eliminated most of the web design coding a long time ago.

Another prominent website builder, Wix, also provides an AI solution called Wix ADI (Artificial Design Intelligence). The head of Wix ADI, Nitzan Achsaf, asserts that the platform can independently develop a website by utilizing the content you submit to propose “billions of beautiful design options.”

The program performs all the reformatting after you select the option you prefer. The company’s materials present it as a straightforward and rapid process; however, it is more akin to an improved version of Wix than a revolutionary tool for creating websites automatically.

Firedrop is yet another web design tool that integrates AI and ML. Sacha, a chatbot that represents the AI of Firedrop, leads you through the site-building process by posing a number of questions and making recommendations and suggestions.

Firedrop was initially introduced in March 2015 as a drag-and-drop website builder. However, it transitioned into a design tool that utilized AI by the end of the year, coinciding with The Grid’s growth.

According to CEO Marc Crouch, Firedrop’s chatbot aims to replicate the experience of collaborating with a professional web designer.

These web design tools may assist algorithms, machine learning, and artificial intelligence but still require manual operation. You enter data, review the options, and select the most suitable option from the predetermined templates. This can significantly assist, especially for modest business owners or novice designers.

Firedrop’s target market is small enterprises, according to Crouch. (Clutch found that 29% of the small businesses it surveyed in 2016 did not have a website; this number fell to 29% in a March 2017 follow-up.)

Template selectors and drag-and-drop interfaces have already alleviated the burden on these businesses’ efforts to establish an online presence.

New AI-based tools promise to simplify the lives of consumers and ensure that the resulting websites do not resemble cookie-cutter replicas. However, the results of these tools, which often appear to have been generated from pre-formatted elements using algorithms and routines, still resemble those carved out of preset templates.

Doug Bartow, principal and design director at the id29 studio in New York, likens the use of templates in any capacity to visiting a copy center. He asserts that templates are functional; however, the layouts become “fairly neutered” rather than distinctive.

According to Bartow, templates also complicate fine-tuning, as they are predetermined and cannot be altered or customized. Nevertheless, Bartow recognizes the potential of the upcoming software surge, whether for web or print design and is amenable to incorporating it—provided that he can continue to produce his most exceptional design work and remain “original and distinctive.”

Adobe is an indispensable software toolmaker for professional designers and media producers, and it has been actively engaged in developing artificial intelligence and machine learning.

Adobe CTO Abhay Parasnis said the company has amassed its data and experience in the ultra-high-end market over the years. It has incorporated the resultant expertise into Adobe Sensei, an artificial intelligence and machine-learning framework that operates in the background of Adobe’s tools.

Currently, the Face-Aware Liquify feature is operational in Photoshop Creative Cloud and Photoshop Fix. This feature employs AI-driven face recognition to enable users to select and modify human faces in photographs.

Parasnis thinks these features, as well as those that are imminent, will empower both professional and non-professional designers by facilitating “an increase in productivity and creativity.”

However, Adobe’s current products, such as The Grid’s site builder, offer only a glimpse of the potential direction of the AI revolution in design. To obtain a more comprehensive perspective, it is necessary to transition from commercially available tools to the research realm.

DesignScape, a University of Toronto initiative, has garnered significant attention since its YouTube debut in 2015 due to the companionship it appears to provide to non-designers.

DesignScape only guarantees to perform some of the tasks for you, unlike Molly from The Grid or Sacha from Firedrop. However, it does not abandon you on your own. Instead, it functions more as a teacher than an assistant, encouraging you to consider alternative and superior solutions. Two of the project’s co-authors and DesignScape developer, Peter O’Donovan, are working for Adobe.
Consequently, its capabilities may soon be integrated into Adobe’s signature products. Adobe Illustrator’s Quick Layout, demonstrated at Adobe MAX 2016, is still in the process of development and automatically adjusts to the items inserted into the composition, as noted by an Adobe spokesperson. This feature is similar to DesignScape.

AI experts anticipate that the field will proceed in the same direction as DesignScape as the rudimentary tools of today mature. During the forthcoming desktop publishing revolution, users will shift their focus from manual labor to allowing the software to generate ideas and plans.

In this scenario, design work will be more akin to curation and management. Our tools will suggest designs, and we will determine which ones are effective. However, what is the extent to which the designs that result will be beneficial and effective?

Nina Stössinger, a senior typeface designer at Frere-Jones Type, anticipates that artificial intelligence will be capable of “replicating patterns and logic” frequently employed in design work.

Some outcomes may be predictable, while others may be layouts we could design independently. Stössinger regards it as a beneficial tool if an application such as DesignScape could rapidly generate warm-up and immediate ideas.

She also expresses concern that such assistance could snare designers, preventing them from customizing their work or thinking outside the box.

According to Paula Scher, a partner in the New York office of Pentagram since 1991, professional graphic designers will continue to desire and require the ability to “make aesthetic decisions about the retouching and the typography” in the presence of AI and ML assistance.

Scher emphasizes that designers will continue to be responsible for “fine-tuning” even if the future of graphic design software involves more curation than creation. However, Scher warns that “entry-level jobs may be lost” as the software becomes more sophisticated and widespread.

Of course, professional graphic designers have been concerned about being supplanted since the initial versions of Quark and PageMaker were launched.

John Maeda, the Head of Computational Design & Inclusion at Automattic, the parent company of WordPress, is a veteran who does not subscribe to fretting. Maeda asserts that he has been preparing for this most recent surge of change since the 1990s:

I know it is an impossible task; however, I cannot give up easily. I am more of a combatant than a warrior. Additionally, I am enthusiastic about the forthcoming obstacles in the design field.

If and when professional graphic designers relinquish some of their responsibilities to machines or are excised from the creative process entirely, they may embrace the change as an opportunity to disconnect from the computer, whether to work by hand or take a vacation from the screen.

If the AI algorithms of this second desktop revolution can save human designers time and give them more time for reflection and creativity, it will be met with applause.

However, individuals who desire the software to operate in a manner that is identical to that of a professional designer should exercise caution when making their wishes.

According to Paula Scher, “If you can locate a machine that thinks like me, do not get it.” “It would be exceedingly irritating.”