Hannah Dempsey, VP Marketing at Jellyfish, says a future dominated by AI is nothing to fear – so long as you don’t try to ignore it

If you attended MAD//Fest this year, you will have left with one two-letter abbreviation ringing in your ears: AI. It was everywhere. I can’t think of a session I attended where, if it wasn’t the main focus of the discussion, it was not at least mentioned in dispatches.

And when you consider the possibilities AI has opened up, rightly so. To give just one example, for some of our clients, we use Pencil Pro, which is a generative AI ad platform that uses your brand assets to create a multitude of different ads. You simply put in your brand assets, and it generates hundreds of variations of ads for all the major ad platforms like Google, Facebook and the rest. It will also optimise towards the designs that are performing best and create more ads like that.

For any brand trying to solve the issue of creative fragmentation and keep up with the myriad ad formats needed if you want to be visible across all the major platforms, Gen AI platforms like this are a no-brainer, as they work at a speed and scale no human could ever hope to emulate. 

Humans still have a role in the AI world

That’s hugely impressive, but it also gets you thinking: what does the future look like for humans in a world dominated by AI? What’s left for them to do? What value do they add? Because, make no mistake, if you think AI is making inroads into advertising, business processes, education, healthcare and everything else, you haven’t seen anything yet.

Despite this, I feel good about the future, with one proviso: we need to embrace AI. The only people with anything to fear or those who refuse to accept it and don’t lean into it. When people say they are worried about AI taking their job, I don’t think that’s quite right. It’s not AI that will take your job, but the person who can use AI effectively.

Because while we have powerful AI tools at our disposal – tools that are getting smarter by the hour – we still need humans to get the most out of them. Without great input from humans to feed into AI, you just won’t get great content out of it. Anyone who has ever asked ChatGPT to write a blog post on X will know that what you get back is rubbish. But if you have a human who can use their talent, their intuition, their experience, to guide the AI, then you see something totally different. 

Writing effective AI prompts is a valuable skill 

To bring this to life a little, I would reference a colleague of mine called Lucas Stanley, who is a cinematographer. A year or so ago, he watched the first video created by Sora, OpenAI’s video generation model. He recognised that this was going to be huge, and that he needed to get to grips with it. So he retrained himself to produce AI videos, one of which was a spec ad for eBay.

For that clip, he had the idea of showing clothes flying out of a wardrobe, not on people, but on hangers. And he told me that creating that AI prompt was a challenge, because AI uses popular associations based on datasets of human-made content. When AI thinks of clothes moving, it thinks of them as being on a person, not moving freely. Creating this video of clothes swishing and swirling around took a lot of time, a lot of experimentation and a lot of trial and error.

What he learned from the process is that writing prompts that deliver great AI outputs is a skill in itself. In fact, ‘AI Prompt Writer’ is probably one of those job titles that doesn’t really exist as yet, but will be commonplace within a couple of years.

Train yourself up

So my advice to anyone worried about the AI future is this: get yourself trained up. There’s no shortage of courses for people who really want to get to grips with AI and work with it, rather than fighting against it. There are lots of AI tools too, so play around with them, see what they can do, and learn what you have to do to get the most out of them.

You might also explore Jellyfish Training’s Generative AI courses. They’re designed to be hands-on rather than purely theoretical, with bite-sized modules on prompt engineering, fine-tuning models and real-world applications, so you can immediately put what you learn into practice without wading through jargon. That practical, project-based approach really helps turn curiosity into confidence.

The AI genie is out of the bottle, and it’s not going back in. Accept this, run with it, and you may well find your biggest fear is actually your best opportunity.