How AI Is Changing the Role of the Designer

Many fear the possibility of a machine taking over the role of a designer. We’re already seeing automation take jobs away from human workers in factories and fast food locations worldwide, but can AI replace humans in the design field?

423
Image of a woman with the words

While artificial intelligence (AI) is nothing new—the field of “AI research” formally began in 1956—it does provide a constant stream of news concerning the many AI advancements in recent years. For example, there is the ability to learn without being explicitly programmed, as is seen with machine learning.

Although this sounds good at face value, many choose the glass half-empty route and instead focus on the implication, or rather, what they think it implies: an imminent AI/robot takeover.

Yes, AI is pretty much everywhere, such as where you shop (Amazon) and watch (Netflix), but this doesn’t mean it’s a bad thing.

Or does it? Does it actually depend on profession?

After all, we’re already seeing automation take jobs away from human workers in factories and fast food locations worldwide. This begs the question, what’s next?

Many fear that #AI-driven systems will take over #Design. We already see #Automation take jobs away from a range of sectors, but can AI replace human designers? No. It's a union. || #IoTForAll @Aumcore Click To Tweet

Is AI in Design Replacing Humans?

A lot of the talk surrounding AI recently has been around machine learning and whether algorithms can shape the future of design.

A robot imagining a city.
Image Credit: Sohu

For example, as technology gets smarter by the day and gains the ability to generate templates according to what’s being inputted, such as the content and colors, many fear the possibility of a machine taking over the role of a designer.

Then there’s the introduction of AI-powered web builders, like The Grid, which promise you can let the machine do everything with algorithms that take shapes, colors and text into consideration.

The Grid: "Welcome doers & creators, this medium adapts to your message."
Image Credit: The Grid

Reality set in shortly after the first sites designed with The Grid came out.
Needless to say, it wasn’t pretty.

After all, as the principles of artificial intelligence user interface design show us, the AI of today isn’t what we originally thought it would be—creating human-like intelligence. Rather, it’s about creating tools that enhance our own intelligence and capabilities.

A Perfect Mix: The Future of Design with AI

At its heart, AI is computer programming that learns and adapts. It can’t solve every problem, but its potential to improve our lives is profound.” 

— Sundar Pichai, CEO, Google

As Sundar Pichai’s quote above about Google’s AI principles of design lends, AI is neither omniscient nor omnipotent; it has its limitations.

In terms of design, this means that AI—at least the AI of today—isn’t going to be taking any design jobs soon. Nonetheless, it’ll soon be found in the workplace.

For instance, think of what design really is. Is it just images thrown together that look pretty? No, it’s something moresomething with the ability to evoke empathy and other emotions in people.

As such, can you decide how many grids you want to use to display your images based on how many you want to display? Sure, but will it be the best design? No, that takes a “human touch” that takes the whole thing into consideration.

Lack of human touch is the biggest problem with leaving everything to algorithms and machines, even those that can learn by themselves.

While they can take a variety of inputs like colors and shapes into account to produce a design, they can’t truly understand the intended audience on a personal level, and so they can’t design something with emotion that creates a connection between brand and audience.

Because of this, the future of design with AI looks more like a partnership between man and machine, in which the machine lends its (artificial) intelligence through tools to human designers, who can then deal with higher-level tasks.

For example, a set of algorithms can do the legwork and take over more mindless tasks like looking at and rating thousands of templates against set criteria, and then the designer can spend more time on the concept as a whole and on how to tailor it according to what the client wants.

In other words, while AI can parse and analyze templates, and perform other tasks like resizing images and color-correcting photos (to a certain extent), it can’t bring its own aesthetics or determine how to create a better picture.

Thus, we return to our partnership between man and machine.

The Tools

Like we just touched on, the most probable marrying of AI and design will be through AI-based design tools that provide designers with extra help as they themselves do the designing.

Again, it won’t be an AI takeover in which robots replace designers; rather, it’ll be a blending that increases accessibility to AI-backed tools, so that designers can integrate them more and more into their workflows.

Let’s Enhance

Let's Enhance: a website that “uses cutting-edge image super-resolution technology based on deep convolutional neural networks” to increase photo or image size without losing quality.
Image Credit: Let’s Enhance

For example, consider Let’s Enhance, a website that “uses cutting-edge image super-resolution technology based on deep convolutional neural networks” to increase photo or image size without losing quality.

In other words, Let’s Enhances uses AI and machine learning to learn typical features on physical objects, and after recognizing those features on uploaded images, it can add extra details based on its own knowledge of the world.

With this ability, Let’s Enhance comes with three main functions: a JPEG noise remover and two additional processing options, Magic and Boring.

  • JPEG noise remover: If an image with a .jpg or .jpeg extension is detected, Let’s Enhance automatically applies a noise reduction system based on neural networks.
  • Magic filter: The Magic filter truly seems like magic as it does the impossible: it “hallucinates” additional details and adds them to images to improve their quality (perfect for photos and complex pictures).
  • Boring filter: The Boring filter is best for illustrations, art, logos, etc. because it can keep colors, details and edges sharp while enlarging the whole thing by up to four times, essentially allowing you to upscale and clean your designs.

Select Subject

Select Subject: a new tool in Adobe Photoshop that uses Adobe Sensei, Adobe's machine learning technology, to memorize shapes and allow users to make subject selections with a single click.
Image Credit: Select Subject

Then there’s Select Subject, a new tool in Adobe Photoshop that uses Adobe Sensei, Adobe’s machine learning technology, to memorize shapes and allow users to make subject selections with a single click.

“Photoshop, in the past, has looked at images as nothing more than a collection of pixels…It had no idea that there was a person, an animal, a tree, or any other type of object in the photo.”

— Photoshop Essentials

With Adobe Sensei and Select Subject in tow, this is a thing of the past, as users can now select prominent subjects in images without the fuss of dragging around the cursor, as is the norm with Quick Selection, the previous go-to tool for subject selection.

Prisma

Prisma:  a photo-editing application that uses neural networks and AI to transform photos into “paintings” with artistic effects that look like they were created by the likes of Picasso, Munch or even Salvador Dali.
Image Credit: Prisma

More on the design side is Prisma, a photo-editing application that uses neural networks and AI to transform photos into “paintings” with artistic effects that look like they were created by the likes of Picasso, Munch or even Salvador Dali.

Deepart

Deepart: An application (or website, actually) that leans heavily on the design and artistic side.
Image Credit: Deepart

Another application (or website, actually) that leans heavily on the design and artistic side is Deepart.

Like Prisma, Deepart turns your images into artwork, but this one also allows you to upload your own style image to further tailor what stylistic elements you can use.

As to the how, Deepart uses “a neural algorithm of artistic style” that was developed by several of its creators, which allows it to separate style elements from a piece of art.

Final Thoughts

Artificial intelligence is absolutely wondrous; there’s no doubt about that.

With that wonder comes the fear that it will be too wondrous—i.e., that it will start to outperform humans and take over their occupations.

While this is a valid concern for some, as we’ve seen with the rise in automation in various industries, this does not stand for other jobs that require more of a human touch, such as design.

Instead, in these fields, we’re seeing a carrying of man and machine that uses the best of the latter to enhance the best of the former.

Written by Matthew, a Content Writer for Aumcore.