No matter how you look at it or whether you want it or not, AI is here, and it is going to stay. In what format, it is hard to know yet, but it has certainly changed how we work.
Having used AI since before generative AI (ChatGPT and other tools where you talk in human language), I am acutely aware of how fast it is changing the way we do things and what becomes important.
In creative work, AI has helped (kind of) us simplify non-creative tasks and big data-processing jobs. I cannot tell you how happy it makes me that I don’t need to be analysing spreadsheets of data or crawling through all of my emails to find one note.
At the same time, AI has introduced a handful of persistent challenges that were not there before.
Here are my (least) favourite:
Client Expectations
Most of the brand design briefs we get from clients these days are accompanied by an AI-generated “idea” or concept of what they would like.
Almost always, those images are either super generic or not functional at all in the way brand files should be.
The brief used to be a lovely starting point. This is where we talk, refine ideas, consider options, and build from there.
The AI-generated concepts are inevitably something the client likes (why else would they include it?). And while “liking” is a metric, it is just one of so many boxes a brand must tick. The AI-generated concepts almost never do.
At this point, we are presented with the additional task of managing our client’s expectations away from what has, by that point, become a graphic they have an emotional connection with and towards something that is a functional and appropriate solution for their organisation. By this point, we have wasted a lot of energy (and probably time) trying to steer them away from the AI-induced identity confusion.
The final outcome is, in most cases, the same, but it takes a lot longer to get there.
Image Editing Blunders
You know when, in the movies, they pull up blurry camera footage and then “enhance” it to make it crisp and perfectly identify the suspect?
Yeah, that is not a real thing.
While you can certainly combine and edit images a lot more easily than before, it is still all about the information available in the image.
There is one thing about removing an element or filling the sky with pretty clouds, but enhancing a low-resolution image is a completely different task, especially when that image is going to be printed at full A4 size.
All of a sudden, you have palm trees covering the hills of the Flinders Ranges, and the image is still a low-resolution image snap that is now an IP nightmare.
Statistical Average
It started with the AI flyers; the first five minutes of them were kind of cute and a tiny bit impressive. But very quickly, they all started looking exactly the same.
AI gives you a statistical average in most creative tasks. While it is definitely possible to make it sound more like you and look more like your brand, the outcomes are still going to be heavily reliant on the statistical average of the LLM’s training.
This is where we bid farewell to cute errors and imperfections, lucky accidents, and true creativity.
Your flyer, brochure, website, and socials—all of a sudden, start looking the same.
And those that don’t have a very, very healthy amount of human contribution in them, if not completely human-produced.
So, next time you talk to your designer, writer, or any other creative professional whom you hire to help you figure it out, get AI to gather information, summarise data, and perhaps take notes on your behalf, but let the human take the lead.
And, if essential, they may allow AI to follow with guidance and clear guardrails.
G.
p.s. written completely by a human. Unique expressions and thought patterns are expected.
