The experiment
We recently tried a little AI experiment here at House of Kitch.
One of the problems we come up against in creative industries is that job titles often fail to capture the true essence of creative talent. House of Kitch is an indie communication agency specialising in deep tech and innovation. But describing the work we actually do isn’t always so straightforward.
Strategic thinking combined with creative execution across campaigns, events, multimedia, public relations and an ability to bring a network of diverse talent together - doesn’t quite roll off the tongue, does it? And that’s before you even get to talking about our incredibly skilled team as individuals.
So, we tried something a bit different with the help of AI.
Working with ChatGPT, House of Kitch has created the "Kitch-tionary” that reimagines our team roles going beyond traditional titles, showcasing everyone’s multidimensional brilliance and unique essence. Some of my favourites include:
Captenacious (noun): An on-point maven characterised by unwavering determination, resilience, and dependable excellence, consistently driving success with an unyielding focus and sharp, dynamic presence.
Elixirist (noun): A masterful storyteller who crafts narratives like potent elixirs, blending creativity and insight to inspire, captivate, and transform audiences with their magical tales.
Dexterioso (noun): A rare, multi-talented individual who seamlessly blends coolness, kindness, and curiosity by day, and spins beats by night, leaving others both amazed and slightly jealous.
The Kitch-tionary isn’t just a reflection of our team’s personalities; it’s a way for us to creatively experiment with AI. Many of our clients are at the cutting edge of AI – developing supercomputers and advanced space sensors. It makes sense for us to explore the possibilities of AI ourselves with something in our own sphere – communications.
Using ChatGPT we developed a process chart to set the task, and off we went.
The article
With the Kitch-tionary complete, we moved on to our next phase – a co-authored article – but this is where the limits (at least from a writing perspective) began to emerge.
Feeding in some informative prompts and a few quotes, we asked ChatGPT to pen an article that described the process and outcome of the Kitch-tionary. The article it wrote was serviceable, but something was off. When you took a step back you can see the formulas and algorithms it used to structure the piece.
Contrasting descriptive phrases such as “a playful yet poignant lexicon” was soon followed by another with “whimsical yet powerful” and then another with “humorous yet heartfelt”. A good style of expression if used once, but suddenly clunky when used just a few lines later.
There was also the tendency to be a bit verbose and oversell an idea. Whimsical yes, powerful and poignant… maybe not as much. There’s definitely the sense of a first-year uni student trying to add words to reach the minimum word-count in an essay. Of course, there could be a lot of time spent going back and forth with prompts or detailed editing, but at that point you might as well just write the darn thing yourself.
The discussion
This exercise prompted a discussion within House of Kitch about the role of AI, and where it sits in a communications framework. The conversation goes deeper than just purpose and productivity - what effect does it have on our humanity? To what level does it take away from the unique human perspectives and insights (or even just a personal writing style)? In looking for the most logical mode of expression, are you just sanitising the unique quirks we have as a writer, leading to hive-mind consensus of thought?
It also seems that AI can lull us into a false sense of security, a Dunning–Kruger effect where there’s the perception of competence, but none of the depth. This becomes apparent when you read something written by AI that you have a good grasp of knowledge on - you can start to see the errors. Similarly, when asked for a source on a piece of information - a fantastic piffy quote - the program admitted it made it up, a hallucination.
But for specific writing tasks, it’s certainly incredible. It can remove obstacles such as ‘blank page syndrome’ of not knowing where to start, build structures for processes, or as a search engine (of sorts) to point you in the right direction when researching a new topic. Of course, ChatGPT has proved valuable for programmers, students writing structured essays or summarising lengthy documents. And that’s before you get into wider AI use of data analysis, image generation (which I’ve used a fair bit to create design elements for Photoshop projects), or custom programmed tools used within legal firms.
The insight
AI used for writing like this is without context. It could write a pretty good press release, but whether it understands the political or corporate climate to put out that press release is another thing entirely.
AI should be viewed as a tool, or if you like – an assistant – for your work. You, as a creative, still have to be in the drivers’ seat. This is where the brilliance of a good comms operator comes in. A copywriter shouldn’t be judged on their ability to write a snappy tagline, but how they reach the core insight of an idea and speak directly to the soul of the reader. Similarly, a strategy expert is more than just someone who can lay out communications plan – it’s their ability to understand the zeitgeist of that moment and shape it.
I’m not under any illusion that a lot of writing work won’t be replaced by AI. And, for a small business needing a quick tagline or content for their social media posts, I can see its appeal.
But I truly believe that there will always be space for (human) strategic thinkers. And as we see more of the soon-to-be familiar hubristic tone of AI, if you really want to cut through with your comms, the human touch will be more important than ever.
Comments