How Will Creatives Work with Artificial Intelligence?
Accelerating creativity in the age of AI.
Adobe Sensei in action.
Machine learning, deep learning, and neural networks — these phrases are confusing if not downright daunting to many of us, yet all are just names for the same thing: artificial intelligence (AI), which, if you haven’t been paying attention, is already integrated into our everyday lives. Using a voice assistant on your phone? Do you consider adding the Amazon suggestions to your shopping cart? Have you messaged a company looking for support and got a reply back before your next breath? All AI.
At Cannes Lions 2018, we hosted a panel with Jamie Myrold (Adobe VP of design), Mario Klingemann (artist and “Neurographer”), and Natasha Jen (partner at Pentagram) who discussed an important question to the festival’s audience: How are creative people in the advertising industry going to interact with AI in their work?
Jamie Myrold (Adobe), Natasha Jen (Pentagram) & Mario Klingemann (artist)
AI: Speed, ideas, and help getting started
“What we know is it will increase efficiency, by a lot. Historically, the tedious things we manually had to do will likely be replaced with some sort of AI-assisted functions,” says Natasha Jen. “The real question is what are we making. The ‘what’ is the heart of creativity.”
Mario has always been interested in what makes something interesting for him, or an audience, and for a long time this wasn’t a subject to tackle with computers. Yet now, with machine learning, he says we can “detect patterns in complex data, where we [humans] might not have words for it, but the machine can see these things. And we have more and more data so the machines can pull truths out of it.”
Mario goes on to say that “we cannot create something from nothing.” When we make things, we get our creative ideas through external impulses. “Now with machine learning, we can have the machine give us these impulses” based on data, what the machine thinks might be relevant to us, “help us — augment our creativity.”
If AI helps everyone, how good will the work be?
It’s not easy to come up with something new, says Jamie. “Finding that original nugget, structuring many parts of an aesthetic experience, is still difficult.” Creativity will always have that human element, it will require our expertise, our practice, and our exploration. So, she questions, does AI have the potential of democratizing creativity, “taking away some of the oomph of it?”
It depends on how you define creativity, says Mario. He defines it as something when someone solves a problem in a way he wouldn’t have expected. Although some people, he concedes, define creativity as ‘creation’ itself.
“Nowadays it’s deemed creative if you use an Instagram filter,” says Mario, who thinks AI will definitely help people feel good about creating because it can help raise the quality level they can produce. Maybe the more interesting aspect of this, Mario points out, is that the better everyone gets at something, then it becomes the new normal. “If everyone is creating beautiful things, there will still have to be something that is better than that.”
Indeed, we can now all create something pleasing with Instagram, Snapchat and a bevy of mobile creative apps — but what does that say about originality?
“Being able to create something, doesn’t equate to being able to create something that can inspire and motivate,” says Jen. “We’re at a place in the discussion of AI in creativity where we need to discuss evaluating the quality of the work.” And at the end of the day, that’s a human’s job.
Jen sees how AI can help in brand identity, for example. She thinks there is big potential for AI-assisted tools that can help creatives build an “identity system” more effectively — thinking beyond just a company logo and type or font — into how a system would work across many touchpoints. “This is something I’m looking forward to…but then it comes down to that decision, ‘what’ are we making here, I still (have to have) the ownership over that,” she says.
Who authored this design?
Jamie points out that designers and creatives often look at other work, to get examples and inspiration, and put their own twist on it for new work. In this new world, where the machine can provide input on a design, who gets the credit for the work?
“It gets into a blurry area now,” says Natasha. Most designs we see in the world now aren’t AI-driven, yet they are created with very impressive tools. “But [people] wouldn’t claim tools are the authors.” Still, we’re starting to see AI out there in media, such as face-swapping on images that are totally realistic. “And that’s where we begin to question not so much who is the author here, but rather who’s responsible, which puts authorship in a totally different light.”
Mario had a video of his work on the Cannes Lions stage, which showed how the AI algorithm he created, searches, finds and creates its interpretation of the drawn faces on the fingers. Who is the actual creator here?
Image from Mario Klingemann’s video showing art created by AI; it’s interpretation of the two fingers with drawn faces
An Adobe AI-powered creative assistant in action
The panelists agreed that AI will make creative work easier by helping with ideas and speeding up production, while the questions around the quality of the work and who in the end is actually responsible for the work may be the bigger questions the creative industry has to solve.
For our part, here is how we see Adobe Sensei-powered (AI) solutions helping:
– When repetitive, time-consuming processes are handled, creatives have more time to be creative.
– When information is hard to find because there’s too much of it to sort through or it takes too long to extract it.
– When trends, anomalies, or significant events are brought to your attention — and you can get data-driven recommendations on what works or ways to optimize — you can better prioritize decision-making.
Watch what we mean, below.
Read more about Adobe at Cannes Lions in our series of articles from the festival.
Additionally, if you are interested in more takes on humans and machines, read our Human and Machine collection of articles.