AI Moves Up the Value Chain: From Productivity Tool to Ideation Partner
Advances in AI can make it not only your assistant and teacher — but also a muse.
by Scott Prevost
posted on 01-06-2020
AI is fueling a paradigm shift in software and, more broadly, in how businesses of all kinds deliver experiences. Most people think of AI as a productivity tool. It can make you more efficient by automating tasks that used to require human attention. But AI is now being used to augment human intelligence in all kinds of other ways. It’s a symbiotic relationship: a machine makes a human being smarter, and the human provides feedback that makes the machine smarter. Ultimately, humans and machines are stronger together than apart.
Self-driving cars are a great example. As long as I’m on the freeway, my car basically drives for me. But I still have to be paying attention. I’m a safer driver because even if I get distracted for a second, my car’s going to see if the car in front of me has slammed on its breaks. If I’m switching lanes, my car won’t let me do it if there’s someone in my blind spot. And the feedback I give the car as I make slight steering adjustments is making the AI better as well.
In the realm of creativity, AI complements human intelligence and creativity, opening up new avenues for creative expression. AI is all about identifying and following patterns. Human creativity is all about breaking existing patterns and designing new ones. When you bring those two concepts together, marrying the greatest strengths of machines and humans, the stage is set for a new kind of creative process — one that empowers creative professionals to push the limits of their creative vision.
For creative professionals, the value of AI-powered automation has been that it frees them from mundane, repetitive tasks. By auto-cropping images or videos to hone in on the focal point or deliver the right aspect ratio, you can spend less time on tedious and time-consuming tasks and more time developing and executing creative ideas. That’s still true, but now we’re also using AI to augment human intelligence in ways that go beyond the tactical details. AI is starting to do things that more closely resemble human creativity.
To get more value from AI, focus on what rather than how
A decade ago, only the most digitally savvy companies had any kind of mobile strategy. Now a lot of companies have a mobile-first strategy for designing customer experiences. A similar thing is starting to happen with AI. The most forward-thinking companies are designing customer experiences with an “AI-first” mindset. In 2018, 25% of all customer interactions were automated through AI and machine learning, according to Gartner research. That number is expected to grow to 40% by 2023.
What that means is that AI is no longer just a method of automating tools for specific tasks. It’s becoming a problem-solving partner.
In the past, when you were trying to accomplish something, you’d break your goal down into steps and then decide which tasks you could automate using AI and machine learning. But in the near future, you’ll be able to begin your partnership with AI as soon as you have a goal or intention. If you get AI involved further upstream, it could actually help you figure out how to accomplish your goal.
For creative professionals, an “AI-first” approach means that in addition to being your creative assistant, AI could also be your teacher and even your muse. AI could teach you how to use a graphic design tool like Adobe Illustrator, for example. And it could even play a role in ideation, helping you explore a wider range of creative possibilities.
AI could be your assistant, teacher, or muse
In a recent study by Pfeiffer Consulting commissioned by Adobe, 74% of creative professionals said they spend more than half their time on repetitive, uncreative tasks. As your creative assistant, AI automates many of those tasks, enabling you to focus on the creative work that lets your talents shine. One example is the AI-powered Content-Aware Fill feature in Photoshop, which helps you quickly remove unwanted elements from an image.
As your teacher, AI could show you which tool to use to create a particular kind of effect. Based on the sequence of tools you’ve used, AI could infer your intent and suggest a technique or feature that might be helpful to you as a next step. It could also nudge you toward training videos and other resources that could help you develop new skills.
As your muse, AI could provide inspiration. Based on what it knows about your goals and the assets you’re working with, it could offer other images or video snippets to get your creative juices flowing. AI could be an amazing brainstorming partner. Because it can generate 1,000 variations on a concept in virtually no time, it could dramatically expand your set of options. Then when you select your preferred example, AI would make additional variations based on that choice.
AI could also generate design variations tailored to different target audiences. Maybe you need a million visual variations tailored to a million customers. Soon that kind of mass personalization will no longer be out of reach. During an ad campaign, AI could directly ingest audience feedback and then use it to modify the designs at scale.
AI could learn design aesthetics, if it had enough labeled examples. It could even assess text against “voice of the brand” guidelines, which may be explicit from rules, or inferred from approved training examples.
As the lines between images, videos, and motion graphics begin to blur, AI is becoming the foundation for new visual media. Eventually AI will help creators express their intent across different media types. Over time, imaginative systems may move beyond designing creative assets to more advanced tasks like telling stories with fully animated AI characters. For example, Fable Studio is experimenting with immersive storytelling in ways that blur the boundaries between creators and consumers — their “Wolves in the Wall” VR experience lets the audience be part of the story.
From finding something to creating something
One of the most valuable early applications of machine intelligence was in search. Computers are naturally able to survey vast quantities of content, in a superficial way, to find possible assets that answer a search query. Search is also a forgiving mode in which irrelevant answers can be ignored if some of the answers are useful.
It used to be we had one visual search algorithm: you give it an image and it finds similar images. But when we talked to people who were doing those kinds of searches, they actually had slightly different things in mind. They wanted an asset with similar content but a different color, or similar objects but arranged in a different way. Or they had an image they couldn’t use for copyright reasons, so they wanted a closely matching image that was available to license.
With advances in deep learning over the past seven years, visual search algorithms have become much more nuanced. You can now search for images by visually describing other images. The description can include keywords and phrases, visual examples, or expressions of visual style and layout — including rough sketches and crude object arrangements (like boxes with labels).
Eventually search may grow up to be more than a retrieval interface. Imagine if you could ask for a “blue dog on a beach” and the system would retrieve an image of a brown dog on a beach and make the dog blue, all in a single step, using an algorithm.
If the AI can both verbally understand a question and visually understand the image content, that combination opens the door to all kinds of advanced editing functions — like auto-captioning images from visual content, auto-generating images from verbal descriptions, and perhaps even inferring the intent and plausible next actions of people in images.
Break the rules like a jazz musician
In the Pfeiffer Consulting study, creative professionals said their main concern with AI is that it might lead to homogenized outputs — all the work will look the same, which would devalue their skills in the marketplace. The study showed most creatives aren’t worried that AI will make their jobs obsolete, but rather that AI might deliver mediocrity at scale.
In the best-case scenario, the user will function as an art director while maintaining artistic control of the whole creative process. You, the human, will provide the creative spark and then let the AI go crazy with it, helping you realize what your idea can become.
Before AI got involved, you might have had time to develop three or four concepts. But when you get the AI riffing off your ideas — like a saxophonist improvising on a musical phrase — maybe you could develop 10 concepts, including an outlier that turns out to be pure genius.
Pablo Picasso famously said, “Learn the rules like a pro, so you can break them like an artist.”
The human ultimately decides. How you interact with AI is up to you. It’s human to break the mold — a machine alone can’t do that. AI could be your assistant, teacher, and muse. Discover how it will fit into your creative process.
Learn more about how humans and machines are working together to drive better outcomes and customer experience.
*This article was originally published on MediaPost.
Topics: Digital Transformation, Creativity, Artificial Intelligence
Products: Photoshop, Illustrator, Creative Cloud