insights / March 10, 2021

The design disruption is already happening

Stephen Saugestad

Machine Learning

One of the most famous man VS machine stories is the legend of John Henry. His story took place in the early 1800s, where men built railroads westward in Virginia. John was a steel driver who drove spikes into rail lines with a giant hammer. He was physically imposing, able to drive each spike with a single blow. But soon came the innovation of the steam powered hammering machine, said to outperform all the men in West Virginia. A challenge ensued between Henry and the machine. He worked day and night and beat the machine. It was a pyrrhic victory though, because he died from exhaustion. Hammer in hand. 

There are similar stories throughout history. Each point to a disruption in the marketplace that saw machines replacing humans. This has led a lot of conversations around AI, and whether it’s going to put us out of work. It’s true that machines and algorithms have been putting a lot of people out of work while transforming industries. But overall the impact of technology has been positive. It’s allowed us to create new jobs, economies, and realize massive efficiencies. Because of machines, we are more productive than ever before. We continue to facilitate innovations that benefit mankind. But it’s impact is not all positive. 

Will machines make designers & creators obsolete?

The pace of innovation has been increasing exponentially. It’s led to disruptions in the job market as jobs are made obsolete. It’s possible that there may never be enough jobs in a future that’s driven by AI. Plus, we haven’t quite yet figured out what a post-work economy and social structure looks like. But there’s a big distinction between the disruptions that replace humans with machines, and ones that make them more effective. This is the direction AI is headed, in the field of design and creative problem solving. 

AI in advertising

Some of the results we are seeing in our advertising work reflects the strength and effectiveness of AI. Machine generated ads are outperforming human generated ones in most valued metrics like clicks, conversions, and ad spend. It’s not surprising that AI is more effective than humans with tasks like segmenting audiences, creates large numbers of adsets and test variations, and optimizing spend. The surprising part? It is also outperforming humans with design – the visual, creative part that machines are not supposed to be good at. 

So are machines replacing designers? The answer to this question is more nuanced than is allowed in our often binary way of debating these things. It also requires more critical thinking, which is not conducive to the way we consume this kind of information these days. 

Art VS Advertising

Let’s make a distinction between “design”, and “advertising”. Digital advertising is more data analytics than emotional appeal. The human generated adset looks better to the human eye, and scores higher in the more subjective metrics we use interpersonally. But digital advertising is not a beauty contest. It’s an engagement contest. And machines are a lot better at determining what we’re likely to click on, and cutting through the biases which in turn reflect our true preferences. 

Strengths & Weaknesses

AI is not good at everything. And this can very well change as the technology improves. But for now the risk of automation between jobs and industries varies depending on whether machines are good at the things you do in your job. Machine learning struggles with the following:

But they’re really good at:

  • Crunching huge datasets
  • Creating variations
  • Examining similar variables
  • Personalization
  • Learning quickly & adjusting
  • Identifying patterns
  • Creating prediction models
  • Reducing bias (if the data is not biased)

In the context of design, the strengths and weaknesses of AI will shape its effect on the industry. Humans will do more of what we are good at, and continue refining those skills. The machines will do more of what we are not good at, making us more efficient. 

I think we’re pretty far from generating really good creative work at the click of a button. But that’s an overly simplistic way of looking at this. AI and machine learning is a tool that extends your skillset and allows you to reach outcomes much quicker and more effectively. As was the case with the steam powered nail driver, the Macintosh, or Adobe’s Creative Suite. 

The evolving role of designers

Design is the process of creating patterns, intentions, and the look and feel of objects. Humans are still better than machines at creating custom or original design. Coming up with a design solution to a problem, or an idea. Facilitate a certain outcome. Analysing those ideas and applying them in useful or innovative ways. This is the process of design that requires critical thinking, and it’s the bulwark for humans against AI. 

In our line of work, the critical thinking part is the diagnostic part of our engagement. We conduct a discovery and articulate the goals. Then we come up with a strategy for achieving those goals, execute the plan, and iterate. Even if the businesses share certain characteristics (like the industry they are in), they will not necessarily have the same goals. One may be trying to increase the number of users, while another may be trying to increase revenue. Say we are talking about hotels. One may be trying to capture more of the business market, while another may want to improve their ADR (average daily rate). These are nuances that come from the executive team through analysis, in service of the needs of the business. It is reasoning and causality, which AI is not good at. Whether it’s an ad or a user interface, it takes a designer to distill that information into something actionable. 

But these nuances aren’t always logical, or aligned with the goals of the business. For example the enforcement of brand standards. If “The” is part of the brand name, our clients will be religious about the way it’s used in the ads and creative work. We don’t usually test whether it gets more clicks or earns more loyalty over the long term, which would be tough to measure anyway. But this is a good example of a subjective and intrinsic human value that a machine wouldn’t really understand.

Curation VS Creation

Once the solution is created though, machines have a very strong advantage. They can create thousands of variations at the click of a button. They can personalize the design with data and analytics. And they can do that within the parameters of a design brief, brand standards, or performance goals. 

That’s why it may be more accurate to describe AI as “Augmented Intelligence”. They are not replacing designers; they are making them better. AI is augmenting the capabilities of designers so they can perform better and increase their output. It is meant to augment creativity, not replace it. So it’s not that the machines are replacing the humans. Rather the nature of the designer’s role is just changing. In the future, the designer’s work will entail more curation then creation. 

There is always a lot of consternation when a new technology comes on the market. This happens when more sophisticated versions of design software become available, for example. But it’s not an outcome we should be afraid of. In fact we should be rooting for it, because it’s making better designers.