The Next Frontier is Now: What to Know About AI in Branding & Packaging Design
by Theresa Christine Johnson on 04/03/2023 | 14 Minute Read
After everything you’ve read in the news (or in our Managing Editor Bill’s Friday Wrap-Ups), perhaps you’d like to stick your head in the sand and let all this talk of Artificial Intelligence die down.
Well, sorry, friend—AI is here, whether you like it or not. Sometimes we fear what we don’t understand, so it’s time to get up to speed on everything AI in design. What projects have we already seen that use AI? How can designers utilize it? What is this weird, strange sci-fi future of ours going to look like? And everyone’s most pressing question: is it coming for your job?
Artificial Intelligence, Then and Now
While artificial intelligence may seem like something out of a sci-fi movie set in the year 2158, in reality, it’s been around for decades. Many regard the first example of AI as Logic Theorist, a program presented in 1956 “designed to mimic the problem-solving skills of a human.” Since then, AI has become part of our everyday life—chatting with a virtual assistant online or with Siri, for instance, or using facial recognition to unlock your phone. In the design world, software like Adobe Photoshop harnesses the power of AI using Adobe Sensei, making features like object selection practically seamless.
And while those tools make the news, projects have, too. There’s Jessica Walsh, &Walsh founder and creative director, who used AI to help model and influencer Isodope promote nuclear power. Or there’s graphic designer Richard Turley who walked It’s Nice That readers through the steps he took to make a soda brand using AI. On our end, designer Silas Amos dreamed up a furry fragrance brand for pop legend Kylie Minogue. Pentagram partner Natasha Jen used a customized algorithm for the School of Visual Arts Senior Library 2018. Even Heinz joined the fun, tasking artificial intelligence to “draw ketchup.”x
Several AI design tools exist (here are six to get you started), and you’ve possibly heard of the popular DALL-E—that’s the one used in both the Heinz and Jessica Walsh projects mentioned above. That, Midjourney, and the just-launched Adobe Firefly generate images from whatever text you put in (aka your prompt). Want some minimalist vodka bottle options or boxes for a kids’ cereal with a unicorn mascot? Consider it done (and in only a matter of seconds). Perhaps you need ten different name ideas for a new office supply brand and don’t feel like hitting your head against the computer for half an hour. Chatgpt, or the recently announced Google Bard, generates a body of text based on the prompt you enter.
Maybe you haven’t jumped into using any of these tools yet—or perhaps you’re like Evelio Mattos and instead have chosen to go headfirst into the deep end. Evelio is the creative director at IDP Direct and host of the Packaging Unboxd podcast. He also teaches a weekly class on LinkedIn through his page Packaging AI, and he recently released the ultimate beginner's guide to designing packaging with Midjourney.
“I guess I’m afraid,” Evelio admitted, explaining why he wrote his guide and is excited to share Midjourney with others. “I feel like I better jump in first before I get pushed in. It’s what's happening at the moment, and it's moving so fast. But this thing is moving forward with or without me.”
Evelio’s work involves concept development, and he said it’s taken him about fifteen years to build up his experience to the point he can sketch out a concept in a half hour or less. But with technology like Midjourney, he can get them in moments.
Gerardo Herrera hasn’t shied away from the topic of AI used in design, and he views it as a vital topic for any of his students. The Head of Brand Experience at Design Studio Nuovo, Gerardo also works at Art Center College of Design as the Director of the Online Master of Design in Brand Design and Strategy Degree and Director of Packaging Design at the undergraduate level.
“I'm trying to get my students to be at the forefront of this. I know what it was like to be at the forefront of mobile wireless technology, working with some really smart engineers and imagining the future,” Gerardo said. He explained that in 1996 Nokia hired him for their brand experience design team, and, essentially, he got to develop the world of mobile technology. On a more personal level, he considers himself an early adopter of new tech—one of the reasons he’s excited about where we are today with artificial intelligence.
“I want my students with that same opportunity to go, ‘Don't fear it. How do you utilize it?’”
But Isn’t AI Coming for My Design Job?! (and Other Concerns)
Before we go any further, we really should address the question everyone is asking—that is, of course, will AI replace designers? If a computer can do it quicker and cheaper, surely brands will stop hiring design professionals and turn to AI instead, leaving you to kiss your job goodbye.
Jasmine Oh, a product designer at Meta, wrote in 2019 that “Yes, AI will replace designers.” But before you race to Google alternate career paths or cry into your coffee, Jasmine’s piece highlighted AI's strengths and limitations, like its inability to understand emotional nuances, lack of moral and social consciousness, and data bias.
“While AI will replace designers, it will replace the designers of today, not the designers of tomorrow,” she wrote. “AI will become a design partner and tool that designers can use to meet ever-evolving workplace demands.”
Rudy Sanchez, a contributor for Dieline, further emphasized that automation transforms the nature of a job rather than eliminating said job. “ATMs have offloaded basic transactions from human tellers to machines, freeing up tellers to focus on more complex services not possible on a touchscreen,” he wrote. “Robots have also taken over some dangerous warehouse or factory jobs, reducing employee risk. Similarly, computers radically changed how graphic designers created but didn’t send them packing."
“It’s unlikely AI like DALL-E will entirely replace human designers," he added. "Instead, it’s more likely to follow in the footsteps of technological advancements and serve as another skill in a creative professional’s tool chest.”
Both Evelio and Gerardo also insist that AI is a tool, not a replacement for the job itself. Brands might think they can go the AI way of CNET or Buzzfeed, type some text into Midjourney, and call it a day, but the reality will look quite different (CNET can attest to this since their AI-generated articles came riddled with errors).
“You know, with Squarespace today, you can design a logo for yourself,” Gerardo explained. “You can say it's taking jobs away. Ultimately, as people succeed in their businesses, they’ll realize and think, ‘it's time for me to change, and I think I need to bring in somebody who really knows what they're doing.’”
Less clear is how the ethical issues tied to artificial intelligence and copyright infringement will play out. Artists are understandably upset because their work has been used to train text-to-image AI systems without consent or remuneration. Hollie Mengert, an illustrator and character designer at Disney, unknowingly had her illustration style cloned as an AI experiment. DeviantArt launched an art generator that trained on images uploaded by DeviantArt users—again, unknowingly to the artists. In response, Getty Images and some individual artists have sued AI art generators Stability AI, Midjourney, and DeviantArt for copyright infringement “for using billions of images downloaded from the internet without artists’ consent,” but we have yet to see how that plays out.
The ethics and legalities here are murky, and they’re different in the United States versus other places like Korea, Japan, China, Australia, Singapore, or Thailand. “I see people on both sides of this extremely confident in their positions, but the reality is nobody knows,” technologist Andy Baio told The Verge. “And anyone who says they know confidently how this will play out in court is wrong.”
There are two concerns here: the use of materials that train machine learning systems and the rights of AI-generated work. Concerning the input, some programs have scoured hundreds of domains, including WordPress, Blogspot, Shutterstock, and more. But just like we here at Dieline can’t take any image from the web we want (we need permission or copyright first), you’d think the same would apply to machine learning.
James Vincent wrote for The Verge, “The justification used by AI researchers, startups, and multibillion-dollar tech companies alike is that using these images is covered (in the US, at least) by fair use doctrine, which aims to encourage the use of copyright-protected work to promote freedom of expression.”
As for who owns work produced by AI, is it the original artist? How much fine-tuning an output does someone have to do for it to become theirs? How much fine-tuning an output does someone have to do for it to not simply be computer-generated—therefore ineligible for copyright protection according to The Compendium II of Copyright Practices from the US Copyright Office?
Kristina Kashtanova’s comic book created with Midjourney, for example, received the first known US copyright registration for latent diffusion AI art. It seems that the US Copyright Office backtracked on their decision, though, later requesting more details about her process to further analyze her involvement. As of this writing, the work is still under review.
“Currently, the United States is a world leader in AI but could quickly fall behind if our policies unnecessarily hinder or limit access to data,” Karen Robinson, vice president, associate general counsel, litigation and intellectual property at Adobe, wrote on their site. “Setting up a more definite legal or regulatory framework for this type of activity would enable AI to continue to flourish in the United States.” She also urged policymakers to find a balanced approach to copyright rules—“one that protects intellectual property rights without impeding the development of AI tools and programs.”
Something Evelio has seen is people using text-to-image generators to create something “in the style of” another artist. Style can be evidence of infringement, but not the only piece of evidence—so you can’t really have a copyright on a style of illustration or design. Evelio encouraged designers using AI to avoid inputs done in a specific artist’s style, unless they genuinely intend to work with that artist later in the process.
“I wouldn't create something in the style of a designer and share it with a client unless the idea is to bring that designer in,” he said. “Even reaching out to that designer beforehand and saying, ‘Hey, we're thinking about working with you on this project. We're going to create some stuff in Midjourney based on your particular styles, and we're happy to share it with you. We’ll let you know if the client decides to go with you. If not, then we won't use this.’ I think having open conversations is the important part.”
Also, as of last month, the US Copyright Office (USCO) has also declared that any image using AI programs like Midjourney cannot be copyrighted in the US.
How Designers Can Harness Artificial Intelligence
One of the reasons Evelio decided to publish his guide to Midjourney was because he was receiving heaps of questions and DMs about AI, and he felt there needed to be some type of resource for those just starting. It was clearly something people wanted to learn more about. There is a learning curve, though, so his guide goes step-by-step to help readers get set up with Midjourney, shares screenshots of what users can expect to see, and—perhaps most important—gives thorough advice on writing prompts.
Beyond that, it’s up to designers, agencies, and brands to see how this best fits into their workflow. AI is ever-evolving and constantly learning, but currently, what Evelio finds most useful is the initial phase of brainstorming and creating a mood board. Once the strategy gets wrapped up and it’s time to go into that first visualization of the brief, designers can use AI to build a mood board that’s more curated for a client and better communicates the feel of a brand.
Too often, Evelio sees the same images used in mood boards—they’re the top results from Pinterest, iconic brands like Apple, or coveted packaging like the Acne Studios underwear box. It’s easy for a team to get attached to an already successful design and be content with having their logo slapped on the packaging. Evelio believes AI can help avoid derivative work in the early phases, encouraging designers and brands to use a mood board for what it’s truly intended.
“We’ve kind of moved away from what a mood board is,” he said. “It’s really about the mood and what the vibe feels like. Often, people use those boards to say, ‘I like this color, I like this logo treatment, I like this texture,’ and they’re Frankenstein-ing all these components together from existing things.”
The platform SHO.AI takes AI to help entrepreneurs and businesses create and refine their brand presence. Essentially, they fill in the gaps, allowing the focus to go elsewhere rather than on mundane tasks like tagging assets to keep them organized.
“Our AI-driven platform helps clients quickly develop and operate their brand without the traditional process' use of resources and expertise,” said Sho Rus, CEO and founder of SHO.AI. “Additionally, we can help them generate on-brand assets like websites, social assets, and business playbooks and provide insights into their brand performance.”
When clients use SHO.AI, they’re using technology that has an identity associated with it—the client’s identity. The AI discovers this brand DNA by asking questions similar to what a brand strategist might ask, like what benefits do you provide customers and what are the company values? Then clients can receive curated options for business cards, social media content, or newsletters in a snap, all of which can be further edited and tweaked by the client.
“Much like the steering wheel on a Tesla car, we make sure that humans always can be a part of that loop,” Sho said. “So they're constantly providing feedback to the AI. Also, the first thing we do is ask the human questions, so we're really solving that alignment issue between who the AI thinks it is and who the human team is, what they're building, and what their intentions and values are when building this business. That coordination is so critical for it to be successful.”
The possibilities certainly seem exciting, but Gerardo said it’s vital to learn the ins and outs of design before designers—or his students—use it for work. Simply entering something into ChatGPT or Midjourney and taking the AI at its word is the same as Googling something and trusting the first thing you find. A skilled designer will be aware of trends, art history, typefaces, materials, and more so that they know how to create a solid input—and so that they know how to take the output and turn it into something tangible.
Currently, the Art Center syllabus is clear on which projects they can and can’t utilize the tech. What's more, Gerardo only brings artificial intelligence into the classroom with his more advanced students. “You need to first and foremost hone your particular point of view,” he said. “You need to work on your ability to craft and write and give us your articulation of how you see the world.”
Ultimately, AI is powerful, but it’s no substitute for the knowledge and intuition of a skilled, trained designer. Gerardo wants his students to have AI in their toolbelt but not rely on it as the only thing they can create with. Using AI requires designers to start with an idea in the first place, and the quality of what’s generated depends on the quality of what they write.
“One of the things that’s really interesting as an artist is that you’re not always typing or voicing your ideas,” Gerardo said. “You're just kind of going, you're doing things, and you're letting the flow take you. Now you're engaging another part of your brain. With this, you’re still taking those ideas, but you’re putting them into text. It forces you to be really good at exercising that creative muscle on the visualization, and they can visually articulate the future that the mind starts to dream up.
“So if you’re lazy, yeah, go for it. Put in a lazy prompt, and you’ll get an average idea. But for those pushed by the tool and their creative nature to become better at what they do, the sky’s the limit.”
What the Future Could Look Like
No one can predict what will come next, but ignoring AI is not an option. It’s here, and either people are talking about it, or they’re thinking about it and too scared to talk about it. And while there’s no doubt that what AI generators like Midjourney and DALL-E can do is impressive, we still have a ways to go until we reach dream scenarios.
For one, copyright laws will become solidified, and artists’ works won’t get violated solely for training machines. Gerardo mentioned the model of the music industry and how when someone samples a song, that artist gets compensated. He hopes to see AI adopt something like this as well.
Additionally, AI will continue to become more capable and could speed up certain processes. Gerardo hopes he can one day enter a prompt that leads to a visualization and then to dieline specifications sent straight to a Kongsberg cutting table. Sho is working to make a world where AI can replace redundant tasks, freeing up branding and marketing teams to focus on more creative or strategic tasks. Evelio sees how AI could someday aid designers throughout the design process, offering suggestions for more effective or sustainable solutions.
And while these ideas might sound somewhat fantastical right now, that might not be because of AI’s capabilities and more because of what artificial intelligence could give us: the invaluable resource of time. More time to create, more time to innovate, and more time for non-work things as well, like family and friends.
And that makes the science fiction future we’re heading towards seem not so bad.