Leaps in AI technology are taking center stage in the creative sector—and across every industry we collaborate with. The sheer volume of diverse and often conflicting perspectives on AI can make just staying informed feel downright Sisyphean. Faced with such hype, uncertainty, and complex legal and ethical dilemmas, it can be tempting simply to brace for impact.
At Praytell, we don’t think that’s enough.
We’re committed to exploring the potential of generative AI in driving creativity and actively influencing the future trajectory of AI by collaborating with our partners and peers.
To put our money where our mouth is, we recently hosted an all-agency workshop where nearly every one of our employees rolled up their sleeves and got hands-on with AI. Together, we explored how to write prompts that get the best out of AI tools, setting ourselves up to use these skills in our day-to-day work. We also invited expert journalist Ryan Broderick and creative exec Gavin Purcell to chat about AI and its uncertain role in our lives and work.
Amidst the swirl of heady enthusiasm, understandable skepticism, and “how-do-I-catch-up”-ism, we believe that it’s how we use AI that matters. By focusing on developing exceptional prompts and ways of working that create much-needed efficiencies, we think we can free up humans to do things computers can’t and use AI tools to elevate our creative process.
In the spirit of collaboration, we want to share what we’ve learned so far:
First, learn by doing.
It can be challenging to show people how to use technology like ChatGPT: the way I use it may not do much for you, and vice-versa. Still, the internet is full of people selling e-books full of ready-made prompts that will help you use AI to do all kinds of things.
But copy-pasting someone else’s prompt isn’t the way to truly familiarize yourself with this technology. It’s the equivalent of trying to learn how to dance by memorizing patterns while never actually listening to music.
The best way to understand what’s possible with a tool like ChatGPT is to start talking to it.
And the best way to talk to it is not far from how you would talk to a person.
If you’re looking for an idea for breaking the ice with ChatGPT, I recommend starting from zero and asking how it can help you. What’s something that you wish you didn’t have to do at work, and how might AI help? Go ahead and ask. You might be surprised by what it has to say.
The best way to understand what’s possible with a tool like ChatGPT is to start talking to it.
Think big, but be specific.
The good thing about working with AI is that you don’t have to worry about a machine’s labor, needs, or hurt feelings. However, you do have to worry about it containing bias, misunderstanding your prompt, or creating a “hallucination” (the word AI researchers and developers use for the common phenomenon of complete and utter nonsense dressed up like the right answer).
Even so, it is possible to write prompts that anticipate some of the common prejudices that can be part of default results.
If you ask ChatGPT for a list of the best guitarists of all time, it will probably give you 10 dudes, mostly white rock guitarists at that.
But if you write a prompt asking for 10 great guitarists, and include instructions to ensure the list is global and diverse in terms of both genre and identity, you’ll uncover a far more well-rounded result.
It’s good to remember that these machines don’t have human judgment, so we often need to introduce parameters that express our values—be that copying language from the brand’s best practices or showing off a best-in-class example. This is one of many reasons why we believe it’s always best to pair AI tech with human expertise and use it as part of your workflow rather than using it as the work.
I’ve spoken to dozens of colleagues at Praytell about how they have started to use AI in their work. Their use cases are all over the place—in an inspiring way.
Animate your curiosity and sense of play.
I’ve spoken to dozens of colleagues at Praytell about how they have started to use AI in their work, and their use cases are all over the place—in an inspiring way that helped me to realize that the uses for this tech were much broader and more varied than my own creative experiments.
One copywriter told me about putting his pre-written lines in and asking for them to be rewritten in the style of different Disney movie characters in order to draw out new adjective ideas.
A manager described using it to anonymize peer feedback and ensure their end-of-year reviews were free from bias.
Designers use image generators to test ideas and get inspiration for their own designs.
Strategists have found it helps them to unlock new storytelling solutions as a chatty and endlessly patient sounding board.
The potential of AI generators is limitless, so there’s a constant push-pull between playing with the limits—and putting in parameters that ensure helpful, ethical outcomes.
There’s a constant push-pull between playing with the limits—and putting in parameters that ensure helpful, ethical outcomes.
Amidst infinite content, true creativity stands out.
Beyond the automation that will hopefully mean never copy-pasting 2390483294 URLs into some spreadsheet again, the future of AI technology seems likely to involve a massive change in the way we consume and engage with content.
During our panel, creative exec Gavin Purcell compared the rise of “infinite content” through AI generation to the explosion of user-generated content via platforms like YouTube and TikTok. He reminded us that people were similarly skeptical about whether such creations would ever have an impact on traditional entertainment models (spoiler alert: they did).
That said, AI-generated creative content also has plenty of pitfalls.
For one, while AI results are often fun and feel magical, the novelty in most cases is in seeing your idea realized—not someone else’s. I’ve compared the experience of making stuff with generative AI tools to lucid dreaming. Just like with dreams, the results are rarely as interesting to others as they are to you.
The rise of generative AI means content will be cheaper than ever—which makes real creativity even more valuable.
In a likely near-future world where powerful generative AI tools are available on every phone and public library computer—meaning anyone can click to generate a passable blog post or love letter, converse with their favorite book character, or make digital artwork of, say, their cat in the style of Rembrandt—the results created by AI could start to look more like infinite debris than game-changing (particularly if brands like BuzzFeed use the tools to spam the internet with SEO-bait)
In other words, content will be cheaper than ever—which makes real creativity even more valuable. AI cannot replace human judgment or interpersonal connection. Developing so-called “soft” skills — those qualities that make us different from anything a machine can do — will be our focus as AI begins to automate once arduous tasks.
The power of the latest wave of AI tech is in the way it communicates, and the effectiveness — and future — of AI hinges on the way we communicate back. In order to create a better and more prosperous tomorrow together, we believe in the importance of becoming fluent in AI comms.