Why AI is not just another tool

Design lecturer and Visual System Designer Dr. Martin Lorenz explains why we need a new value system for creating with AI.

Edited by Anne Kaiser, deputy editor-in-chief, PAGE,
published in issue: https://page-online.de/shop/page-11-2023/

PAGE Magazine, November Issue, 2023

“I miss nuanced perspectives in the discussion about AI on both sides: AI optimists praise the innovation potential of the technology but ignore the long-term costs and consequences. Pessimists, on the other hand, disregard the pressure to remain competitive.”

At the peak of the AI hype, it seems there are only two camps: either celebrating the new technology as the future of the creative industry, or doubting it and hoping for a counter-movement that brings human-made design back to the forefront. Initially, I was excited about the possibilities that artificial intelligence offered. The accompanying increase in efficiency seemed to align perfectly with what I aim to achieve with my Flexible Visual Systems. However, a deeper exploration of the topic radically changed my assessment.

Now, I miss nuanced perspectives in the discussion about AI on both sides: AI optimists praise the innovation potential of the technology but ignore the long-term costs and consequences. Pessimists, on the other hand, disregard the pressure to remain competitive. The pace at which AI tools are developing hardly allows time for calm thought, let alone a reflective approach to this technology. Nevertheless, we, the creatives, should take this time because we shape, as the first users and directly affected, how AI will evolve in the future.

“Shouldn’t we strive for sufficiency rather than efficiency?”

Is AI fundamentally bad?

Creatives bear a particular responsibility in the context of AI. For our corporate clients, it already seems evident that the new technology can support better and faster communication, cater to trends, and keep up on social media. However, the long-term consequences of this efficiency improvement go far beyond the design industry: a brand that communicates better sells and produces more, which in turn leads to a greater burden on the environment, higher energy and resource consumption. We must not ignore this.

Even if it’s assumed that AI won’t be used to further boost consumption, the technology itself poses problems: ChatGPT is, according to a study by the UBS bank, the fastest-growing web app of all time. However, few users are likely aware that both during the training of AI models in physical data centers and in interaction – such as through prompts – large amounts of electricity and water are consumed. Researchers at Cornell University calculated that the energy required to train a medium-sized AI model causes approximately 626,000 tons of CO2 emissions (https://arxiv.org/abs/1906.02243). Extrapolated to the seemingly endless number of new AI tools added daily, the resource consumption is unimaginably high.

Therefore, we should carefully consider what we actually want to use AI for and whether the efficiency increase outweighs the global burden: Do we want to create more profit with artificial intelligence? Shouldn’t we strive for sufficiency rather than efficiency? Can’t we use this highly intelligent tool to show us the costs and consequences of our actions in all their complexity? I believe it is up to us – as experts in information processing – to communicate the risks and opportunities of the technology.

David Clark made this post in December 2023: “Made an adidas AI spec commercial on my coffee break. Many of you know I’m a prominent commercial director when not making films. Had some fun with Runway” If Clark had been able to make this ad in his coffee break, and Motion Designers would have taken weeks or months to do something similar, the fear of a decrease in Motion Design budgets or even the risk of the profession ceasing to exist is more than justified. Courts are discussing at the moment under which conditions AI-generated material can be copyright-protected. The outcome of these new laws might redefine the value we give to creativity.

Can AI replace creatives?

In addition to the ecological consequences, there are societal implications, such as the inevitable job loss due to the use of artificial intelligence. As much as techno-optimists hope to integrate AI as a tool into existing workflows, the reality is that it will replace many designers in their current professional definitions. Not to mention the countless creatives who have already been harmed by the use of their works in training large AI models.

This development is unstoppable and is becoming apparent everywhere: those who do not keep up will be left behind. The hype around AI makes prompting seem like a vital skill. In agencies, I read briefings written by AI. I see mood boards that almost look like final works but still rely on copyrighted material. But are they good? Or do we accept the AI-generated result simply because it’s quick and easy to produce?

In the worst case, AI could replace creatives in almost all digital areas. And this development won’t stop even for AI-savvy designers because even strategic alignment and audience-specific communication could be generated by AI and implemented visually. In this chain of tools, entrepreneurs could then curate concepts, strategies, texts, images, videos, and designs generated by AI themselves – perhaps with minimal adaptation by someone who knows how to use AI. Techno-pessimists see this scenario as the downfall of the creative industry, but I believe that now, more than ever, we should highlight the value of our design processes and expertise in client consulting.

“Strategies and ideas arise in conversation with other people, between the lines where crucial information is hidden.”

What can creatives do better than AI?

Of course, I’m not immune to the allure of technology either. For instance, AI is capable of creating professionally looking images. However, I haven’t seen anything that truly moves me. Humanity seems to be missing in all of it. It became interesting for me when I had the first debriefs written with ChatGPT. I was excited to be able to delegate an unloved part of my work. However, upon closer examination of the texts, I realized that neither the nuances were right nor were the priorities properly set.

Strategies and ideas arise in conversation with other people, between the lines where crucial information is hidden. Because even if our clients speak the same language as the AI, there’s a language barrier. What someone really means and wants must first be discovered through interpretation and dialogue – and even then, it’s not guaranteed that the clients’ vision aligns with the goals and values of the brand.

“Automation through AI condemns us to the level of instructive and curative design and prevents a deep engagement with the details.”

Is more efficient always better?

Anyone who writes or draws has experienced this: the admittedly sometimes exhausting design process opens up time and space for shaping and improving an idea. Charles Eames made the famous statement: “The details are not the details. They make the design.” Automation through AI condemns us to the level of instructive and curative design and prevents a deep engagement with the details. While we may speed up the process, we deprive ourselves of the time for inspiration and reflection that gives our design quality. This quality arises in problem analysis, in the third iteration loop, and in conversations with clients. And let’s be honest – why automate the part of the process that often brings the most joy?

I fear that designers who predominantly rely on AI-assisted design in the future will possess only superficial knowledge and lack a deep, embodied sense of design. Being able to operate software does not necessarily mean mastering human communication. Moreover, our ability to solve problems could be limited by dependence on artificial intelligence. Being flexible in the use of tools also makes our thinking more flexible.

“We must work with like-minded individuals from various disciplines towards common goals.”

How can creatives effectively deal with AI?

To seriously address this question, we would need to slow down the development and create a differentiated discussion. However, we cannot do this alone. We must work with like-minded individuals from various disciplines towards common goals. The great thing is: as creatives, we bring all the prerequisites to convey information in overarching systems, form communities, and create stronger tools.

Artificial intelligence is one of these tools and, if used correctly, could help us depict a planetary perspective beyond the boundaries of our own industry, show complex connections, and enable us to act more responsibly. The prerequisite is the willingness to live a holistic approach and a value system that goes beyond our intrinsic incentives. Then, it is up to us creatives to convey these values to our clients and make them the yardstick for our decisions in the creative process.

For this, we need more than logos or social media strategies generated by artificial intelligence. We need flexible visual languages or systems eloquent enough to make various content understandable for an interdisciplinary audience. And it requires the transformative competence of creatives to develop these languages.

So, don’t lose your head in the discussion about AI, think systemically and holistically! Resist, as designers, against commissions that do not align with your value system, and act with caution – and yes: this includes carefully evaluating when AI is truly necessary and when one should refrain from it from an ethical and ecological perspective. As designers, we are responsible for a large part of commercial communication in the public sphere. This gives us a certain power to change things or simply not do them anymore. I firmly believe that what distinguishes us creatives in the end from AI and makes our design good is the humane conduct.

Martin Lorenz

A couple of months have passed since the article was published. Reading it again I thought I had to add a paragraph to answer the initial question: Why is AI not just another tool?

We just have to observe how AI is talked about. We treat it as if it were a supernatural being. A higher force, superior to us. This is where the danger lies. AI doesn’t even need to get smarter. We are making ourselves dumber and dumber. The “democratization” of the tool increases its use in every part of our digital lives. We are making ourselves more and more dependent on it. The moment we outsource the thought process, which is a creative process, and we trust the machine more than we trust ourselves, we become slaves of the machine. AI is the most powerful and most opaque machine we have ever built. It can understand us much better than we’ll ever be able to understand it. In addition to the destructive incentives of our societies, we are most likely to harm ourselves and the planet we live on even more if we do not start taking the costs and consequences of our actions seriously. I strongly believe that we creatives are natural system thinkers and have therefore a chance to question our new tools in order not to become the tool. AI itself is not the problem. The problem is who develops it with which incentives.